White House Report Discusses Expanding Opportunities for Middle-Income Taxpayers

This year's Economic Report of the President describes how after 5 years of grit and determined effort, the United States is better-positioned for the 21st century than any other nation on Earth. We've now experienced 4 straight years of economic growth with more than 8 million new private-sector jobs. Our unemployment rate is the lowest it's been in more than 5 years. Our deficits have been cut by more than half. For the first time in nearly 20 years, we produce more oil at home than we buy from the rest of the world. The housing market is rebounding, manufacturers are adding jobs for the first time since the 1990s, and we sell more of what we make to the rest of the world than ever before.

But in many ways, the trends that have threatened the middle class for decades have grown even starker. While those at the top are doing better than ever, average wages have barely budged. Inequality has deepened. Too many Americans are working harder and harder just to get by, and too many still aren't working at all. Our job is to reverse those trends. It is time to restore opportunity for all -- the idea that no matter who you are or how you started out, with hard work and responsibility, you can get ahead.

That's why this must be a year of action. I'm eager to work with the Congress to speed up economic growth, strengthen the middle class, and build new ladders of opportunity into the middle class. But America does not stand still, and neither will I. Wherever and whenever I can take steps without legislation to expand opportunity for more American families, I will. Because opportunity is who we are. And the defining project of our generation is to restore that promise.

Simply put, this opportunity agenda has four parts. Number one is more new jobs. Number two is training more Americans with the skills to fill those jobs. Number three is guaranteeing every child access to a world-class education. And number four is making sure hard work pays off for every American.

With the economy picking up speed, companies say they intend to hire more people this year. We should make that decision even easier for them by closing wasteful tax loopholes and lowering tax rates for businesses that create jobs here at home, and use the money we save in the process to create jobs rebuilding our roads, upgrading our ports, and unclogging our commutes. We should help America win the race for the next wave of high-tech manufacturing jobs by connecting businesses and universities in hubs for innovation. We should do more to boost exports and fund basic research. We should maintain our commitment to an all-of-the-above-energy strategy that is creating jobs and leading to a safer planet. Finally, we should heed the call of business leaders, labor leaders, faith leaders, and law enforcement, and fix our broken immigration system. Independent economists say this will grow our economy and shrink our deficits by almost $1 trillion in the next two decades. We should get it done this year.

Creating jobs is step one, but in this rapidly-changing economy, we also must make sure every American has the skills to fill those jobs. I've asked Vice President Biden to lead an across-the-board reform of America's training programs to make sure they have one mission: training Americans with the skills employers need, and matching them to good jobs that need to be filled right now. That means more on-the-job training, and more apprenticeships that set a young worker on an upward trajectory for life. It means connecting companies to community colleges that can help design training to fill their specific needs.

I'm also convinced we can help Americans return to the workforce faster by reforming unemployment insurance so that it's more effective in today's economy. But first, the Congress needs to restore the unemployment insurance it let expire at the end of last year, affecting around 2 million workers.

Of course, it's not enough to train today's workforce. We also have to prepare tomorrow's workforce, by guaranteeing every child access to a world-class education. Our high school graduation rate is higher than it's been in 30 years, and more young people are earning college degrees than ever before. The problem is we're still not reaching enough kids, and we're not reaching them in time.

That has to change. I am repeating a request I made last year asking you to help States make high-quality preschool available to every four year-old. In the meantime, I'm going to pull together a coalition of elected officials, business leaders, and philanthropists willing to help more kids access the high-quality early education they need. I'll also work to redesign high schools and partner them with colleges and employers that offer the real-world education and hands-on training that can lead directly to a job and career, and follow through on my pledge to connect 99 percent of our students to high-speed broadband over the next 4 years. With the support of the FCC, we've announced a down payment to start connecting more than 15,000 schools and 20 million students over the next 2 years, without adding a dime to the deficit, and with the help of some of America's top companies, we're going to make the most of these new connections.

My Administration is also shaking up our system of higher education, so that no middle-class family is priced out of a college education. We're offering millions the opportunity to cap their monthly student loan payments to ten percent of their income, and I will continue to look for other ways to see how we can help even more Americans who feel trapped by student loan debt.

But we know our opportunity agenda won't be complete -- and too many young people entering the workforce today will see the American Dream as an empty promise -- unless we do more to make sure hard work pays off for every single American. This year, we should do more to secure a women's right to equal pay for equal work. We should expand the Earned Income Tax Credit to help more workers without children make ends meet, and help more Americans save for retirement through the new "MyRA" plans my Administration is creating. We should protect taxpayers from ever footing the bill for a housing crisis ever again. And we will continue the work of making sure every American has access to affordable, quality health insurance that's there for them when they need it.

And we should raise a minimum wage that in real terms is worth less than it was when Ronald Reagan took office. In the year since I first asked the Congress to raise the minimum wage, six States raised theirs, and more companies like Costco see paying fair wages as one of the best ways to reduce turnover, increase productivity, and boost profits. As America's chief executive, I agree, which is why I signed an Executive Order requiring Federal contractors to pay their employees to pay their federally funded employees a fair wage of at least $10.10 an hour for new contracts. There is a bill in front of both the House and the Senate that would raise the minimum wage to $10.10 for all Americans. The Congress should pass that bill and give America a raise.

I believe this can be a breakthrough year for America. But it falls to all of us to grow the economy and create new jobs, to strengthen the middle class, and to build new ladders of opportunity for folks to work their way into the middle class. So in the coming months, let's see where we can make progress together. Let's continue to make this a year of action. Together, we can restore an economy that works for everybody, and our founding vision of opportunity for all.

The Council of Economic Advisers herewith submits its 2014 Annual Report in accordance of the Employment Act of 1946 as amended by the Full Employment and Balanced Growth Act of 1978.

As of early 2014, however, the economic landscape looks vastly different: total output has grown for 11 consecutive quarters, businesses have added 8.5 million jobs since February 2010, and a range of analysts are optimistic that the economy will further strengthen in the years ahead. Yet despite this progress, many American families are still struggling to join the middle class or to stay there, as they face the lingering after-effects of the crisis on top of a longstanding trend of widening inequality that has caused the rungs on the economic ladder to grow further apart. This fundamental issue -- making sure the economy provides opportunity for every American -- is the President's central economic focus.

To move toward the President's goal of a broader, more solid foundation for future growth, three key imperatives must be addressed. The first and most immediate imperative is to continue to restore the economy to its full potential. Although the recovery from the Great Recession is well underway, it remains incomplete. The second imperative is to expand the economy's potential. In the decades after World War II, rapid productivity growth propelled the American economy forward and grew the middle class. The historic surge of women into the labor force further transformed America's society and economic capacity. But in the last several decades, productivity growth has declined relative to the early postwar years, and looking ahead, America's workforce is now expected to grow more slowly as members of the baby-boom generation move into retirement. As a result, efforts to enhance overall productivity and the skills of American workers, and to expand the labor force, will be as critical as ever. The third imperative is to ensure that the economy provides all Americans with greater opportunity to realize their full individual potential and to experience the prosperity they work to create. A typical family's inflation-adjusted income barely budged in the years leading up to the crisis, which placed undue strain on millions of hardworking households, contributed to instability in the overall economy, and raised questions about how the American ideals of opportunity and mobility would manifest themselves in the 21st century.

 

Figure 1-1

 

Monthly Change in Private Nonfarm Payrolls, 2007-2014

 

 

 

Source: Bureau of Labor Statistics, Current Employment Statistics.

These challenges are substantial, but so is America's potential. As discussed in the remainder of this chapter and throughout this report, the President has set out an ambitious agenda to make progress on all three imperatives, both by working with Congress and by taking action on his own where possible. To return the economy to its full potential more quickly, the President has called for a range of measures including an Opportunity, Growth, and Security initiative, along with steps to pair business tax reform with a major effort to upgrade our Nation's infrastructure. To expand the economy's potential, the President continues to urge the House of Representatives to follow the Senate's lead and pass commonsense immigration reform, which would help attract a new wave of inventors and entrepreneurs to American soil. The President also wants to build on the great strides that have already been made in the technology and energy sectors, which are laying the groundwork for a more productive economy in the years ahead. And to make sure that the economy provides opportunity for every American, the President's agenda includes improved job training programs and an increase in the minimum wage, because no one working full time should have to raise their family in poverty. In addition, the President has set a goal of preschool for all, since one of the best ways to ensure that all Americans have a chance to succeed is to invest in their early childhood development. Implementation of the Affordable Care Act is another critical step in this direction, because it is helping to provide financial security for more American families and to slow the growth in health care costs that cut into workers' take-home pay.

The President is calling for significant legislative measures in these areas, but will continue to pursue progress through executive authority by, for example, streamlining the infrastructure-permitting process, creating new products to improve retirement security, and raising wages for Federal contractors. The President is also using his influence to work with businesses, universities, and nonprofits to, for instance, help low-income students succeed in college and ensure that the long-term unemployed have a fair shot at a new job. These are just a few of the President's key efforts in what he has described as "a year of action," and the President remains ready to work with anyone offering constructive ideas to move forward.

 

Figure 1-2

 

U.S. Merchandise and Overall Trade Deficits, 2000-2013

 

 

 

Source: Census Bureau, U.S. International Trade in Goods and Services; Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

 

Figure 1-3

 

Major Deficit Reduction Episodes Over a Four-Year Period

 

Since the Demobilization from WWII

 

 

 

The progress seen in 2013 is notable in part because the steep decline in the Federal budget deficit during the year was a major headwind on macroeconomic performance. Figure 1-3 shows the largest four-year reductions in the Federal budget deficit since the demobilization from World War II. Since the end of fiscal year 2009, the deficit has fallen by 5.7 percentage points of GDP, with nearly half of that reduction -- 2.7 percentage points of GDP -- coming in FY 2013 alone. Some of the deficit reduction in 2013 was a natural consequence of the gradual improvement in the economy, and a large portion of it was due to policy decisions, like the spending caps agreed to in the Budget Control Act of 2011, the increase in tax rates for top earners at the beginning of 2013, as well as the end of the temporary payroll tax holiday. The fiscal contraction would, of course, have been much worse were it not for the permanent extension of tax cuts for middle-income households. While these factors reflected a balanced approach to making progress toward fiscal sustainability, they were unnecessarily compounded with the onset of budget sequestration in March, which the Congressional Budget Office (CBO 2013a) estimated slowed real GDP growth over the four quarters of the year by 0.6 percentage point and reduced employment by the equivalent of roughly 750,000 full-time jobs.

On top of the fiscal headwinds, the economy was also forced to contend with a disruptive government shutdown for 16 days in October, as well as dangerous brinksmanship over raising the Federal debt limit. The shut-down cost the government more than $2 billion in lost productivity alone as Federal workers were furloughed for a combined 6.6 million days. In addition, families were unable to travel to national parks, oil and gas drilling permits were delayed, Small Business Administration loans were put on hold, and licenses to export high-tech products could not be granted, to name just a few effects. Consumer sentiment, as measured by the Reuters/University of Michigan Index, fell to its lowest point of the year in October, and fourth-quarter GDP growth was constrained by a large negative contribution from the Federal sector. Moreover, because Congress was slow to act to raise the debt ceiling, several large investment management firms announced that they had divested from Treasury securities maturing around the time of the potential debt-ceiling breach. Ultimately, the episode provided yet another reminder of the need for policymakers to avoid self-inflicted wounds to the economy and to stay focused on constructive steps that support growth and job creation.

Chapter 3 looks back at the American Recovery and Reinvestment Act of 2009 (the Recovery Act), along with more than a dozen subsequent jobs measures that the President signed into law, including the payroll tax cut, extensions of unemployment insurance, and tax cuts for business investment and hiring. At the macroeconomic level, CEA estimates that the Recovery Act, by itself, cumulatively saved or created about 6 million job-years, where a job-year is defined as one full-time job for one year (equivalent to an average of 1.6 million jobs a year for four years). Adding in the subsequent jobs measures, the cumulative gain in employment through the end of 2012 grows to 9 million job-years (Figure 1-5). This analysis is broadly similar to others provided by a variety of sources, including the Congressional Budget Office, and is generally consistent with a growing academic literature that is discussed in Chapter 3. In addition to the macroeconomic impact, Chapter 3 also reviews the Recovery Act's key investments in areas like education, clean energy, physical infrastructure, and technological infrastructure that will continue to pay dividends long after the Recovery Act has phased out.

Fiscal measures were only part of the overall policy response. The President acted decisively to save the American auto industry and the suppliers and economic ecosystem that depend on it. The Administration also instituted a range of measures that helped families stay in their homes, facilitated refinancing into lower interest rates, and took steps toward alleviating the housing blight that can threaten neighborhoods and restrain home values. Furthermore, the Treasury Department promoted financial stability and instituted a range of programs to help restore the flow of credit for both large and small businesses. The Federal Reserve undertook important independent actions as well, which are more fully described in Bernanke (2012, 2014).

 

Figure 1-4

 

Real GDP Per Working-Age Population

 

in 2007-2008 Banking Crisis Countries, 2007-2013

 

 

 

Source: Statistical Office of the European Communities; national sources; CEA calculations.

 

Figure 1-5

 

Quarterly Effect of the Recovery Act and Subsequent

 

Fiscal Measures on Employment, 2009-2012

 

 

 

Another positive but less widely appreciated story from the recession and recovery is the performance of America's social safety net. Though employment and incomes declined sharply as the recession unfolded, millions of Americans were kept out of poverty by tax credits and programs like Social Security, nutrition assistance, and unemployment insurance. Excluding these measures, the poverty rate would have risen 4.5 percentage points from 2007 to 2010, but in fact it only rose half a percentage point (Figure 1-6). The direct effect of the Recovery Act on incomes, not even counting its impact on jobs and the broader economy, reduced the poverty rate in 2010 by 1.7 percentage points, equal to 5.3 million people kept out of poverty. As discussed in Chapter 6, these developments represent the continuation of a longer-running trend in which essentially all of the progress made in reducing poverty has come as a direct consequence of government programs. The poverty rate excluding tax credits and public programs actually rose from 1967 to 2012, but when tax credits and programs are included, the poverty rate was cut by 38 percent (Wimer et al. 2013). Nevertheless, with 49.7 million Americans still living in poverty as of 2012, far more work remains to be done. As the Nation recently marked the 50th anniversary of President Lyndon B. Johnson's declaration of a war on poverty, Chapter 6 details lessons that have been learned, ways that antipoverty programs can be strengthened, and other policies that can help take the next step forward by achieving meaningful reductions in poverty even before tax credits and government programs kick in.

 

Figure 1-6

 

Change in Poverty Rate From 2007-2010,

 

With and Without Tax Credits and Benefits

 

 

 

Cyclical Factors

Diminished fiscal headwinds. The most predictable reason for optimism about the U.S. economy in 2014 is the waning drag from fiscal policy and reduced fiscal uncertainty. In December, a bipartisan budget agreement averted a second round of discretionary sequester cuts that were scheduled to go into effect in January and also relieved a portion of the cuts that had already taken place during the preceding year. While Congress could do substantially more to support job growth and economic opportunity, the economy is unlikely to face anything like the fiscal consolidation seen at the Federal level in 2013, with deficit reduction continuing at a much more gradual pace going forward. As part of the budget deal, Congress also agreed on discretionary funding levels for the remainder of FY 2014 and all of FY 2015, offering a way to avoid another counterproductive shutdown. Earlier this year, Congress passed appropriations bills for FY 2014 consistent with these spending levels and also extended the debt limit into 2015.

As fiscal headwinds ease at the Federal level, State and local governments are also showing encouraging signs. After shedding more than 700,000 jobs from 2009 to 2012, State and local governments added 32,000 jobs in 2013.

Improved household finances. American households saw trillions of dollars in wealth wiped out as a result of the recession, but recent data indicate that a large degree of progress has been made in the recovery. As of the third quarter of 2013, real per-capita household wealth had recouped over 80 percent of the large decline from its peak, reflecting gains in housing and stock prices, as well as the progress households have made in deleveraging. Moreover, the household debt service ratio -- the estimated required payments on mortgage and consumer debt as a share of disposable income -- was 9.9 percent in the third quarter of 2013, the lowest since the data began in 1980, and down from 13 percent in 2007. Further improvements in household finances and expanded access to credit will contribute to strengthening in consumer spending. Looking over the course of the recovery, real personal consumption expenditures have grown just 2.2 percent at an annual rate, compared with the 2.9 percent pace during the 2000s expansion period, a fact that partly reflects the lingering after-effects of the financial crisis. A noticeable pickup in consumer spending -- which comprises more than two-thirds of the U.S. economy -- would represent an important step toward turning the page on the crisis era.

While the aggregate statistics on household wealth paint a picture of improvement, too many families have not shared in the gains. For instance, middle-income households have on average a larger fraction of wealth in their homes relative to equities, and house prices -- despite recent improvements -- have not recovered as sharply as equities, which represent a larger fraction of the wealth held by upper-income households. The challenges of ensuring that more Americans share in the gains from economic growth are discussed in additional detail later in the chapter.

Structural Trends

Along with these positive near-term developments, this report also highlights three longer-term, structural trends that have emerged recently and will support growth on a sustained basis into the future.

Domestic energy boom and changes in energy use. The first major trend is the dramatic increase in domestic energy production combined with a shift in the use of energy that represents an important opportunity not just for the economy, but also for America's security and climate. Current projections indicate that the United States became the world's largest producer of oil and gas in 2013, exceeding both Russia and Saudi Arabia. As noted earlier, domestic production of crude oil rose above imports in October for the first time since 1995 (Figure 1-7), and further increases in domestic production and reduced oil imports are expected in the coming years. Moreover, natural gas production continued to rise in 2013 from the 2012 record high and is up more than 20 percent over the past five years. The power sector has undertaken a shift from coal to natural gas, which was responsible for 27 percent of our overall energy consumption in 2012, up from 24 percent in 2008. But the progress is not limited to oil and gas -- consistent with the President's "all of the above" energy strategy, great strides have also been made in renewables and energy efficiency. Wind and solar power generation have each more than doubled since the President took office, while oil consumption has fallen over this time, as stronger fuel economy standards and investments in cutting-edge technologies have led to the most fuel-efficient light vehicle fleet ever. This broad-based energy boom supports jobs directly in production and distribution, and also indirectly, by making the United States more attractive as a location for multi-national firms in energy-intensive industries like manufacturing.

 

Figure 1-7

 

Domestic Crude Oil Production and Net Imports, 2000-2013

 

 

 

Source: Energy Information Administration, Petroleum Supply Monthly.

The President recently announced new steps to further capitalize on these exciting developments in the energy sector and reduce our dependence on foreign energy sources while creating new jobs. In the 2014 State of the Union address, the President announced his intention to forge ahead with new executive actions that will improve the fuel efficiency of the nation's trucking fleet and help states and localities attract investment in new factories powered by natural gas.

Although many of the recent trends in the energy sector are positive, looking ahead over the coming decades, climate change continues to pose considerable threats to America's environment, economy, and national security. The combined effects of shifting electricity production from coal to cleaner-burning natural gas, large increases in wind and solar power generation, and ongoing progress in energy efficiency have made a large contribution to reducing national energy-related carbon dioxide emissions by more than 10 percent since 2005. In his Climate Action Plan, the President set out the concrete steps that the Administration is taking to address the costs of climate change through new actions to reduce greenhouse gas emissions and prepare for the future climate changes that are an inevitable consequence of past emissions. The President has also recently directed his Administration to continue working with states, utilities, and other stakeholders to set new standards on carbon pollution from power plants.

Health care cost slowdown. The second structural trend is the slow-down in the growth of health care costs. The growth rate of real per-capita health care expenditures from 2010 to 2012 was the lowest since the Center for Medicare and Medicaid Services data began in the 1960s (Figure 1-8), and preliminary data and projections indicate that slow growth continued into 2013. As detailed in Chapter 4, this historic slowdown in health care cost growth does not appear to be merely an after-effect of the recession. The slowdown has persisted even as the economic recovery has unfolded, and it is evident in areas like Medicare as well as in the gap between health care price inflation and overall inflation, neither of which should be sensitive to cyclical fluctuations. Chapter 4 also presents evidence that some already-implemented features of the Affordable Care Act, including reductions in overpayments to Medicare providers and private insurers as well as payment reforms that incentivize better patient outcomes, are contributing to this trend. Primarily as a consequence of slower health care cost growth, the Congressional Budget Office has marked down its forecast of spending on Medicare and Medicaid in the year 2020 by about 13 percent relative to the projection it issued in August 2010. Employers and families are also likely to see significant benefits as health care places less pressure on employers' compensation costs, and the resulting savings are passed on to workers in the form of higher wages.

 

Figure 1-8

 

Growth in Real Per Capita National Health Spending, 1961-2013

 

 

 

Source: Centers for Medicare and Medicaid Services, National Health Expenditure Accounts; Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

Expansion of innovation. The third emerging trend that presents a major opportunity for long-term growth is the rapid advance in telecommunications technology, particularly in fast and widely available wired and wireless broadband networks, and in their capacity to allow mobile devices to take advantage of cloud computing. The economic potential of these technologies and the broader context of U.S. productivity growth are discussed in greater detail in Chapter 5. From 2009 to 2012, annual investment in U.S. wireless networks grew more than 40 percent from $21 billion to $30 billion, and the United States now leads the world in the availability of advanced 4G wireless broadband Internet services. This infrastructure is at the center of a vibrant ecosystem that includes smartphone design, mobile app development, and the deployment of these technologies in sectors like business, health care, education, public safety, entertainment, and more. All told, the expansion of advanced telecommunications technology -- along with the slowdown in health care cost growth and the rise in domestic energy production -- are major reasons to be upbeat about the U.S. economy's growth prospects in the coming years.

Long-Term Fiscal Sustainability

To the extent these structural trends continue to unfold and support stronger-than-projected economic growth in the years ahead, they will help move the Federal government closer to fiscal sustainability over the medium- and long-run. The steep decline in the Federal deficit over the last several years discussed earlier has been accompanied by similar improvement in the long-term fiscal outlook. One key gauge of the long-term fiscal outlook is the fiscal gap, which represents the amount of tax increases or spending cuts as a share of GDP required in the present to stabilize the debt-to-GDP ratio over the next 75 years. While long-run fiscal projections are always subject to a wide margin of error, recent estimates of the fiscal gap are smaller than those issued just a few years ago. These improvements are thanks in large part to the aforementioned slowdown in health care cost growth, including the cost-saving measures in the Affordable Care Act together with other spending restraint and the restoration of higher tax rates on high-income households.

Continuing to Restore the Economy to its Full Potential

Despite the 8.5 million private-sector jobs added over the last 47 months and the decline in the unemployment rate to a five-year low, the economy has not fully healed from the massive blow of the Great Recession, and helping restore the economy to its full potential is the most immediate challenge policymakers face. This imperative can in large part be understood through the prism of the U.S. labor market, which is currently subject to multiple distinct but closely related challenges. First, given the magnitude of the job losses stemming from the Great Recession and the need to add jobs to support a growing working-age population, the economy continues to exhibit an absolute shortfall in the number of jobs.

Moreover, while the long-term unemployment rate has trended down, it still remains markedly elevated. As shown in Figure 1-9, the prevalence of persons unemployed for 26 weeks or less has returned to its pre-recession average (4.2 percent of the labor force in January 2014, same as the average from 2001 to 2007), but the long-term unemployment rate is more than twice what it was during the pre-crisis years (2.3 percent in January 2014 compared with 1.0 percent on average from 2001 to 2007). Reducing long-term unemployment presents a major challenge because these individuals may face stigmatization from employers or experience skill deterioration.

 

Figure 1-9

 

Unemployment Rate by Duration, 1994-2014

 

 

 

Source: Bureau of Labor Statistics, Current Population Survey; CEA calculations.

Even as the economy continues to add jobs, it will also be important to see a concurrent recovery in the volume of job-to-job mobility. The flow of workers across firms plays a critical role in the economy because job mobility enables rising productivity and wages as individuals switch to jobs for which they are better suited. However, in recent months there have been fewer than 9 million combined hires and separations a month, compared with more than 10 million a month on average from 2005 to 2007. The rate of voluntary separations -- a measure of workers' confidence in labor market conditions -- also remains below pre-recession levels.

As the unemployment rate trends down and worker mobility picks up, real wages should grow more quickly; but currently, the sluggish real wage growth seen in recent years represents another serious outstanding challenge in the labor market. Real wage growth remained positive for most of 2013, a key sign of the progress being made (Figure 1-10). However, as discussed in greater depth below, substantially faster real wage increases will be needed to make up for a decades-long trend of average wages failing to keep pace with productivity gains.

 

Figure 1-10

 

Growth in Real Average Hourly Earnings for

 

Production and Nonsupervisory Workers, 2007-2014

 

 

 

Source: Bureau of Labor Statistics, Real Earnings.

 

Figure 1-11

 

Real Median Family Income, 1980-2012

 

 

 

Source: Census Bureau, Historical Income Data.

The typical family's inflation-adjusted income has also been slow to recover. The Census Bureau reported in September that the median family earned $62,241 during 2012, little changed in real terms from the preceding year (Figure 1-11). After increasing cumulatively less than half a percent in the seven years leading up to the recession, real median family income fell markedly during the recession and its aftermath and, as of 2012, was still 8 percent below its previous peak. Going forward, this is an important indicator of the way in which the recovery makes progress for the middle class.

To speed the recovery, boost job creation, and tighten labor markets so as to put upward pressure on median wages, the President has repeatedly called for investment in America's infrastructure. This type of investment would not only help address the 11 percent average unemployment rate in the construction sector during 2013, but would also foster stronger long-run growth. In the 2014 State of the Union, the President renewed his call for using the one-time revenue associated with the transition to a reformed business tax system to finance a major modernization of U.S. transportation infrastructure. But in the absence of the necessary Congressional action on this proposal, the President will press ahead in other ways by speeding up the permitting process for new construction projects.

In addition, the President's budget includes an Opportunity, Growth, and Security initiative, which will finance additional discretionary investments in areas such as education, research, infrastructure, and national security. The $56 billion initiative is evenly split between defense and non-defense and is fully paid for with mandatory spending reforms and tax loophole closers -- and would both speed the return of the economy to its full potential and also expand that potential.

In addition to the challenges in the areas of jobs and income, the housing sector represents another key area with scope for further improvement. As shown in Figure 1-12, construction activity fell so far in the wake of the recession that, despite notable gains in recent years, the rate of permitting for new residential sites still remains well below the level suggested by demographic trends and home depreciation. To help unlock this potential, steps must be taken to bring certainty to the mortgage finance system, and also to support communities that were particularly hard hit when the housing bubble burst and are still coping with a legacy of foreclosures and blight.

Expanding the Economy's Potential

In addition to taking steps that speed the economy's return to full potential, the Administration continues to push simultaneously for measures that will expand that potential. To understand the importance of this goal, consider that an American worker in 2012 could produce more than four times as much per hour as his or her counterpart in 1948. About 10 percent of the increase is due to improvements in the composition of labor, mostly because of greater education, and 38 percent is due to increases in the amounts of capital that workers have at their disposal. Fully 52 percent is due to increases in total factor productivity, or what the Bureau of Labor Statistics calls multifactor productivity, which reflects technological change as well as the scale of markets and organization of production processes.

 

Figure 1-12

 

Building Permits for New Residential Units, 1960-2014

 

 

 

Source: Census Bureau, New Residential Construction; CEA calculations.

The growth of total factor productivity can vary widely year-to-year, but the longer-term trends can be broadly illustrated by splitting the last 60 years into three periods, as shown in Figure 1-13. First, from the 1950s through the early 1970s, total factor productivity grew at a relatively rapid 1.8 percent annual rate, fueled in part by public investments like the inter-state highway system and the commercialization of innovations from World War II like the jet engine and synthetic rubber. Then, from the mid 1970s to the mid 1990s, the rate of total factor productivity growth slowed substantially, to just 0.4 percent a year. The causes of this slowdown have been the subject of extensive academic debate, with some evidence pointing to the disruptive effect of higher and volatile oil prices. Finally, from the mid 1990s through the latest available data for 2012, total factor productivity growth picked up to a 1.1 percent-a-year rate, in part reflecting vast improvements in computer technology and software during this time. Although differences across these episodes may seem small, over time they compound to enormous differences in output and living standards.

 

Figure 1-13

 

Growth in Total Factor Productivity, 1953-2012

 

 

 

Source: Bureau of Labor Statistics, Multifactor Productivity; CEA calculations.

While the growth rate of total factor productivity is always critical to an economy's long-run potential, the projected slowdown in the growth of America's workforce due to the aging of the population places even greater emphasis on productivity going forward. Since an economy's potential output depends fundamentally on the number of workers and the average output per worker, slower productivity growth can in theory be offset by rapid population growth. And during the aforementioned 1974 to 1995 period marked by slower total factor productivity growth, the working-age (16 to 64) population continued to expand at a solid rate of more than 1 percent a year. However, the Census Bureau projects that over the 20 years from 2012 to 2032, the working-age population will grow only 0.3 percent a year, largely a consequence of the aging of the baby boomers into retirement. Thus, looking ahead, productivity-enhancing investments are likely to be as critical as ever.

The President has several key proposals that will expand the U.S. economy's long-run potential. Comprehensive immigration reform would counteract the slower growth of America's workforce by attracting highly skilled inventors and entrepreneurs to create jobs in the United States. The Congressional Budget Office (2013b) has affirmed that the bipartisan bill passed by the Senate would raise the economy's productivity and total output.

In addition, while the Opportunity, Growth and Security initiative, infrastructure investments, and business tax reform proposals discussed earlier will speed the recovery in the near term, these steps will also help make the economy more productive over the long run. Specifically, the President's framework for business tax reform would expand the economy's potential by reducing the distortions in the current system that skew investment decisions. By developing a system that is more neutral, corporate decision makers can act for business reasons, not tax reasons, creating an environment in which capital will flow to the most efficient purposes.

Expanding the economy's potential also means ensuring that the Federal budget is fiscally sustainable over the medium- and long-run. Fiscal sustainability frees up resources for productive investment and reduces borrowing from abroad that would ultimately require commensurate reductions in future national income. For these reasons, the President's budget proposals have repeatedly included additional steps toward long-run fiscal sustainability, including proposals to promote additional efficiencies in the health care system and limit tax benefits and loopholes for the highest earners. The Administration has also placed substantial emphasis on using evidence and evaluation in the budgeting process, as discussed in Chapter 7. These practices will help make the Federal Government more efficient and save taxpayer money for years to come.

Executive actions -- including establishing new, innovative manufacturing institutes and expanding the SelectUSA initiative to attract foreign investment -- will also help improve productivity and create well-paying jobs in the United States. Further, the Administration continues to negotiate new trade and investment partnerships with Europe and Asia, to help American consumers, the American businesses that sell their products overseas, and the workers they employ. The President is asking Congress for the Trade Promotion Authority he will need to make these agreements a reality.

Promoting Economic Opportunity

The third major challenge the economy faces is the need to ensure that every American has an opportunity to realize their full potential and partake in the prosperity they help create. Since the late 1970s, the United States has seen a major increase in income inequality due to a combination of technological change, globalization, changes in social norms, and institutional shifts like the erosion of the inflation-adjusted minimum wage and the decline in union membership. As shown earlier, in 2012, the median U.S. family's inflation-adjusted income was less than it was in 1997. In contrast, separate data derived from administrative tax records show that the top 1 percent of tax units -- who earned an average of more than $1 million in 2012 -- received 19.3 percent of total income (excluding income from capital gains, which can be highly volatile year-to-year), the largest share since 1928 (Figure 1-14).

 

Figure 1-14

 

Share of National Income Earned by Top 1 Percent, 1915-2012

 

 

 

Source: September 2013 update to Piketty and Saez (2003).

 

Figure 1-15

 

Growth in Productivity and Average Wage, 1947-2013

 

 

 

Source: Bureau of Labor Statistics, Productivity and Costs, Current Employment Statistics; CEA calculations.

These statistics are symptomatic of a troubling disconnect between the economy's productivity and ordinary workers' wages that has emerged over the last 40 years. As shown in Figure 1-15, real average hourly earnings for production and nonsupervisory employees roughly kept pace with productivity growth in the nonfarm business sector during the early postwar years. But starting around the 1970s, a large gap emerged between overall productivity and an ordinary worker's take-home pay. Several factors may be contributing to this gap, including the relatively rapid increase in non-wage compensation like employer-sponsored health insurance, as well as the likelihood that productivity gains may translate into higher pay differently across occupations. Moreover, the gap is substantial regardless of whether one adjusts for inflation using the Consumer Price Index, which is more indicative of the typical household's purchasing power, or using the price index for total nonfarm business output, which allows for a more direct comparison with the productivity data. Ultimately, the Figure illustrates the growing concern that too many ordinary workers are being left behind, and helps explain why the President has said that the basic bargain in America -- that those who work hard will have the chance to get ahead -- has frayed.

To address this issue and to restore a greater measure of fairness and opportunity to the economy, the President has proposed a number of important measures. In the near term, the most direct step is to raise the minimum wage, which, after adjusting for inflation, has declined by more than one-third from its peak in the 1960s and is now worth less than it was when President Ronald Reagan first took office in 1981. Along with a minimum wage increase, the President has also called for other measures that would help those striving to join the middle class, including an expansion of the Earned Income Tax Credit for workers without children. In addition to these immediate steps, the President has set out a range of ideas to invest in education and equip workers with the skills they will need to compete in the global economy for years to come. For instance, the ConnectEd program continues to move forward, putting high-speed Internet in classrooms across America, and the Administration has also secured over 150 new commitments from universities, businesses, and nonprofits to improve college opportunity and outcomes for students from low-income families.

Along with steps that create jobs, boost incomes, and invest in educational opportunities, the President is pursuing measures to ensure that families can experience a greater degree of financial security. Three million young adults under age 26 have gained coverage through their parents' plans because of the Affordable Care Act, another 4 million people enrolled in insurance plans through State and Federal marketplaces as of late February, and millions more have been determined eligible for State Medicaid programs. Additionally, in the 2014 State of the Union, the President announced the creation of MyRA, a new safe and easy-to-use savings vehicle that can help millions Americans start saving for retirement.

Looking ahead, a wide variety of indicators suggest that the economy is well situated for a pickup in growth in 2014. Following the largest four-year reduction in the Federal deficit as a share of GDP since the post-WWII demobilization, Federal fiscal policy will be much less of a drag in 2014 and thus will likely constrain overall growth by less than during the preceding years. State and local government spending appears to have turned the corner, with purchases increasing during the second and third quarters of 2013. The mid-February action of Congress to suspend the debt limit until March 2015 relaxes a situation that had been headed towards unwelcome uncertainty.

Although the economy still remains challenging for many, households -- on average -- are in an improved position to increase spending as they have further reduced their debt burden and seen a substantial increase in housing and stock-market wealth. More household wealth will facilitate an increase in spending on consumer durables such as motor vehicles, which are showing their age and due for replacement. Homebuilding, which grew rapidly last year, is likely to continue growing on a path up to levels consistent with the demographic forces of the next decade, with mortgage interest rates still below their pre-recession levels, despite a mid-2013 rise. Business fixed investment also has potential to accelerate after relatively slow growth in 2013 as aggregate demand picks up and businesses can take advantage of their sizeable cash flows.

Nevertheless, several downside risks to economic growth remain in 2014 as unforeseen events both domestically and internationally may pose a risk to the economy. Recently, for example, severe cold weather and storms in the United States and a global reduction in asset prices have contributed to some economic activity falling below trend rates of growth in the last few months.

The pace of the recovery will depend, in part, on policy choices. Additional measures that increase aggregate demand would add impetus to the economy in 2014. In particular, the Budget also includes the Opportunity, Growth, and Security initiative, which will finance additional discretionary investments in areas such as education, research, infrastructure, and national security. The $56 billion initiative is evenly split between defense and non-defense and is fully paid for with mandatory spending reforms and tax loophole closers. In addition, investments in infrastructure or extending emergency unemployment benefits would expand demand immediately while measures like business tax reform would help the economy by increasing certainty.

Growth in aggregate economic activity was fairly steady during 2013, with quarterly growth rates between 1.8 and 3.0 percent at an annual rate for the first three quarters of the year, as measured by the average of the income- and product-side of real GDP (Figure 2-1).¹ During the four quarters of the year, growth was strong in exports (4.9 percent) and in residential investment (6.6 percent), and moderate in business fixed investment (3.0 percent) and consumer spending (2.1 percent). State and local purchases edged up slightly following four years of decline, while Federal spending fell 6.2 percent.

 

Figure 2-1

 

Mean GDP Growth, 2007-2013

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

Fiscal Policy

Federal fiscal policy evolved through several near- or after-deadline Congressional actions that made fiscal policy uncertain and created a difficult planning environment for businesses and consumers.

Toward the end of 2012, policy focused on the potential negative effects of the "fiscal cliff," a confluence of expiring tax cuts and scheduled spending declines that were on track to occur simultaneously, which might have resulted in a sharp fiscal-policy tightening on January 1, 2013. The Congressional Budget Office (CBO) estimated that these policies, if allowed to occur, would have lowered real GDP growth by about 2.25 percent during the four quarters of 2013, or enough to cause a decline in real GDP. On the tax side, the 2001 tax cuts, previously extended through 2012, were expiring. Also expiring at the end of 2012 was a 2-percentage point cut in the Social Security payroll tax that was first instituted for one year in 2011, and an increase of the threshold for the Alternative Minimum Tax (AMT). On the spending side, defense and nondefense spending were each scheduled for across-the-board cuts (sequestration) of $55 billion. Medicare payments to physicians and emergency unemployment benefits were among other spending programs scheduled to be cut in January 2013.

On January 1, 2013, Congress passed the American Taxpayer Relief Act of 2012 (ATRA). The ATRA addressed the revenue side of the fiscal cliff by making permanent the middle-class tax cuts, indexing the AMT to inflation permanently, and raising revenues over 10 years by allowing high-income tax cuts to expire. The ATRA also allowed the temporary payroll tax cut to lapse. On the spending side, the ATRA extended Emergency Unemployment Compensation and delayed the Medicare physician cuts for an additional year, but the Act delayed sequestration only until March 1, 2013.

When Congress failed to reach a budget agreement by March 1, allowing the sequester to go into effect, cuts to discretionary and non-exempt mandatory programs were distributed over the remaining seven months of the fiscal year (rather than the full fiscal year in the sequester's original design). As a result, many Federal agencies furloughed civil servants, which reduced Federal compensation by $0.6 billion at an annual rate in the second quarter and by $5.5 billion at an annual rate during the third quarter of 2013 (a total of $1.5 billion not at an annual rate). The CBO projected that the sequester would cut 750,000 jobs and reduce growth during the four quarters of 2013 by a 0.6-percentage point.

The debt ceiling had technically been reached on December 31, 2012 when the Treasury Department commenced "extraordinary measures" to enable the continued financing of the government through mid-February. Around the end of February, however, Congress passed and the President signed a bill that suspended the debt limit though May 18. The next day, on May 19, the debt ceiling was reinstated at a level that reflected borrowing during the suspension period, but no more. As a result, the Treasury began applying extraordinary measures once again and, in late September, the Treasury announced that these extraordinary measures would be exhausted by October 17.

Adding to the debt-ceiling stress, more uncertainty arose in early October when the continuing resolution needed to fund the government was not extended into the new fiscal year beginning on October 1. As a result, the U.S. Government went into a partial shutdown. About 850,000 Federal civilian employees were initially put on temporary leave, but many civilian Defense Department employees were recalled during the second week of the shutdown. An agreement for a continuing resolution to end the shutdown and extend the debt ceiling was reached on October 16, and the Federal government returned to normal operations the next day. The Bureau of Economic Analysis (BEA) has estimated that the shutdown was directly responsible for a 0.3 percentage point reduction in the annualized GDP growth rate for the fourth quarter, although this estimate does not incorporate indirect effects that operate through reductions in private activity dependent on government services, reductions in confidence, or increases in uncertainty. Confidence in government policy, as measured by the Thompson Reuters-University of Michigan Survey, fell to a level in October which was in the bottom 5 percent of the monthly series since it began in 1978.

The agreement to end the shutdown (the Continuing Appropriations Act of 2014) funded the government through January 15, 2014, and suspended the debt ceiling until February 7, 2014, after which time it was suspended again until March 2015. In mid-December, Congress passed an agreement to provide partial relief from the automatic sequestration of discretionary spending in FY 2014 and 2015, and offset those increases with increased pension contributions from new Federal civilian employees, as well as a variety of higher fees and spending reductions. The bill provided only an overall discretionary cap and, in January, Congress passed FY 2014 appropriations bills consistent with these spending levels. Notably, the bill would fully restore cuts to Head Start programs, which provide early childhood education to children from low-income families, partially restore cuts to medical research and job training programs, and finance new programs to combat sexual assault in the military.

As a result of this fiscal stringency and continued GDP growth, the Federal deficit-to-GDP ratio fell 2.7 percentage points to 4.1 percent in FY 2013 and ranks among one of the largest year-over-year declines ever (Figure 2-2). The deficit-to-GDP ratio in FY 2009 was elevated by the steep recession as well as the fiscal stimulus to combat that recession (See Chapter 3). Since then, the four-year decline in the deficit-to-GDP ratio of 5.7 percentage points was the largest since the demobilization at the end of World War II. Overall fiscal support substantially raised the level of output and employment since 2009, as discussed in Chapter 3. But the reduction in the deficit, especially in 2013, has acted as a drag on growth rates. One reason for the fiscal drag was the winding down of various countercyclical fiscal policies taken during the recession. Fiscal drag is likely to moderate substantially in FY 2014, with a projected further 0.4 percentage point decline to 3.7 percent in the deficit to GDP ratio under the President's policies.

 

Figure 2-2

 

Federal Budget Deficit, 1950-2015

 

 

 

Monetary Policy

In 2013, the Federal Open Market Committee (FOMC) continued to provide substantial policy accommodation. With its usual tool -- the federal funds rate -- near its effective lower bound, the Committee employed both forward guidance for the federal funds rate and additional purchases of longer-term securities.

The FOMC made clear its intention to keep the target range for the federal funds rate "exceptionally low" and maintained throughout the year the forward guidance it issued in December 2012 indicating that the Committee will maintain the current level of the federal funds rate at least "as long as the unemployment rate remains above 6.5 percent, inflation between one and two years ahead is projected to be no more than half a percentage point above the Committee's 2 percent longer-run goal, and longer-term inflation expectations continue to be well anchored." Moreover, in its December 2013 statement, the FOMC added to its forward guidance, stating that "The Committee now anticipates, based on its assessment of these factors (labor market conditions, inflation, and inflation expectations), that it likely will be appropriate to maintain the current target range [of 0 to 1/4 percent] for the federal funds rate well past the time that the unemployment rate declines below 6-1/2 percent (emphasis added), especially if projected inflation continues to run below the Committee's 2 percent longer-run goal." This additional information was intended to provide greater clarity on the Committee's policy intentions once the unemployment threshold is crossed.

With regard to asset purchases during 2013, the Federal Reserve continued expanding its holding of mortgage-backed securities at a rate of $40 billion a month and longer-term Treasury securities at a pace of $45 billion a month in an attempt to "support a stronger economic recovery and to help ensure that inflation, over time, is at the rate most consistent with its dual mandate," to achieve "maximum employment and price stability." In the period leading up to the June FOMC meeting, financial market participants interpreted some Federal Reserve communications as implying an earlier-than-expected reduction in the pace of purchases. This interpretation contributed to an increase in market volatility and a marked rise in longer-term Treasury yields over the summer that were only partly reversed in the fall, as the Federal Reserve continued to purchase assets at an unchanged pace. However, at its December meeting the FOMC decided to begin reducing the pace of its purchases in January, cutting the monthly increase in its holdings by $10 billion to $75 billion. In addition, the Committee indicated that if incoming information broadly supports its expectation of ongoing improvement in labor market conditions and inflation moving back toward its longer-run objective, it will likely reduce the pace of asset purchases in further measured steps at future meetings. This tapering of asset purchases was continued in January 2014 as announced in the FOMC meeting that month.

Financial Markets

Financial developments over the course of the year reflected the evolving economic outlook as well as Federal Reserve communications. In the spring and the summer, speculation about a possible reduction in the pace of Federal Reserve asset purchases contributed to a sizeable increase in longer-term interest rates (Figure 2-3).

Yields on 10-year Treasury notes were 1.7 percent at the start of May before rising to 2.6 percent in July, and yields continued to rise to about 2.9 percent just before the September FOMC meeting. In response to the Committee's decision to leave the pace of purchases unchanged in September, the 10-year yield retraced part of the summer increase, dropping to 2.6 percent for the month of October. In addition, Federal Reserve communications appeared to lead investors to push back their expectations for the timing of the first increase in the federal funds rate during the fall. Toward the end of the year, however, the better-than-expected readings on payroll employment and on other economic indicators, followed by the FOMC's decision to reduce the pace of its asset purchases, boosted longer-term Treasury yields in the final weeks of 2013. The 10-year Treasury yield closed 2013 at roughly 3 percent. Short-term rates (such as the rate on federal funds, and the 91-day Treasury bill rate) were more stable throughout the year -- remaining under 0.2 percent -- although expectations of future short-term rates fluctuated.

 

Figure 2-3

 

Interest Rates, 2010-2014

 

 

 

In October, brinksmanship over the debt ceiling -- which was expected to be hit soon after October 17 -- and the two-week government shutdown weighed heavily on financial markets. Through September and early October, several indicators of financial stress reflected market participants' concerns about the debt limit. As shown in Figure 2-4, yields on specific Treasury bills maturing around that time increased in anticipation of potential delayed payments.

Moreover, institutional money market funds saw a sizeable $86 billion of outflows (about 5 percent of assets) in the three-week period that ended October 16. Fidelity Investments -- the nation's largest manager of money market mutual funds -- declared publicly in early October its decision not to hold U.S. government debt set to mature around the date of the potential debt ceiling breach. Finally, interest rates on overnight repurchase agreements, or repos, collateralized by Treasury securities, a common source of funding for financial institutions, spiked in early October. With the resolution of the debt ceiling debate, all such indicators returned to normal levels.

 

Figure 2-4

 

Treasury Bills Maturing in Late October-Early November, 2013

 

 

 

Reflecting the ongoing economic recovery and the improved outlook over the course of the year, U.S. equity markets remained on a general upward path despite the increase in interest rates. The Standard and Poor's 500 rose by 30 percent in 2013, reaching a record high in nominal terms at year-end. When adjusted for GDP price inflation, however, it remained below its March 2000 peak. The Standard and Poor's edged up slightly during the first two months of 2014.

International Developments

The past year also saw the beginnings of a recovery in Europe, with real GDP edging up between 1.0 and 1.6 percent annual rate in the second, third, and fourth quarters of 2013. These were the first three consecutive quarters of positive real GDP growth for the 28-country European Union since 2011. Concerns about the stability of the European monetary union (the 17-country "euro area") that surfaced in 2011 and 2012 have subsided. In the euro area, the unemployment rate stabilized at a record high of 12.1 percent from April to September before ticking down to 12.0 percent in the fourth quarter. Euro area inflation was subdued, declining to only 0.8 percent during the 12 months of 2013 from 2.2 percent a year earlier. The recent low rate of inflation has fueled concerns about possible deflation. The European Central Bank policy target for price stability is "below, but close to, 2 percent." The Central Bank's Outright Monetary Transactions program, first announced in August 2012, has helped bring a measure of stability to European sovereign debt markets, with Italy and Spain's 10-year yields ending the year right around a manageable 4 percent. During the year, euro area states made substantial progress to centralize and harmonize bank supervision and regulation at the euro level.

There were notable developments in several European countries as well. In the runup to the euro area crisis, countries including Greece, Spain and Portugal saw a large runup in their current account deficits to finance private and public borrowing that supported consumption and investment. In the wake of the euro area crisis, these countries have adjusted, largely eliminating their current account deficits through reductions in unit labor costs and improved price competiveness, as shown in Figure 2-5. Nevertheless, unemployment rates remain particularly high in these countries.

Japan's real GDP grew a solid 2.7 percent during the four quarters of 2013 following a 0.4 percent decline during 2012. Japan's core consumer price index (that is, excluding food and energy) turned positive, 0.7 percent during the 12 months of 2013, up from a 0.6 percent decline during 2012. This follows in the wake of the election of Shinzo Abe in December 2012, the appointment of a new governor of the Bank of Japan in March, and the announcement in April that the Bank intended to double the monetary base by the end of 2014. Under this policy, bond purchases amount to about $80 billion a month (basically, the same pace as the Federal Reserve but in a smaller economy). Expansionary monetary policy was part of a three-prong strategy that initially included fiscal stimulus and structural reforms meant to support positive growth and to keep Japan from slipping back into a period of deflation.

China's real GDP grew 7.7 percent during the four quarters of 2013, slightly below the year-earlier pace, but noticeably slower than 10 percent and 9 percent growth rates during 2010 and 2011, respectively. Xi Jinping assumed the presidency in March and presided over the Third Plenary Session of the Communist Party, which resulted in a raft of economic reform proposals. China's interbank lending rates have spiked on several occasions this year. During these episodes, the People's Bank of China was slow to inject liquidity, which many interpreted as a warning to banks that have increased off-balance sheet commitments to bypass administrative and regulatory controls and expand lending.

 

Figure 2-5

 

Current Account Balance by Country, 2000-2013

 

 

 

Among other emerging market economies, the pace of real GDP growth fell in Indonesia, Malaysia, Mexico, South Africa and Thailand. But growth also increased in a few, such as Brazil, India, and Turkey. Low interest rates in the United States since the recession coupled with higher investment return prospects in emerging market economies prompted an increase in capital flows toward emerging markets. As interest rates in the United States began to rise and growth prospects abroad waned, however, investors started adjusting their portfolios, which in some cases had adverse effects on emerging-market currencies and interest rates. Foreign mutual funds withdrew $53 billion from emerging markets between mid-May and August, leading to sharp drops in a number of currencies and emerging market equity indexes and causing central banks in several affected countries (India, Indonesia, Turkey, Brazil, and Pakistan) to raise domestic policy interest rates. Nevertheless, even with the withdrawals, investment holdings remained well above the levels of just a few years ago as shown in Figure 2-6. In some instances, currencies and bond markets have retraced their earlier losses, especially as global investors have increasingly differentiated debt by country according to the underlying economic fundamentals of each country's economy.

 

Figure 2-6

 

Cumulative Flows into Mutual and Exchange-Traded Funds Investing in

 

Emerging Markets, 2010-2014

 

 

 

Developments in 2013 and the Near-Term Outlook

Consumer Spending

Real consumer spending grew about 2 percent during each of the past three years. With consumer spending constituting 68 percent of GDP, that stability explains much of the stability of the growth of aggregate demand during those three years. Yet the stability of consumption growth during 2013 results from several offsetting developments. The termination of the temporary 2-percentage point cut in payroll taxes reduced disposable income during 2013 by $115 billion relative to 2012. This subtracted about 0.9 percent from disposable income, and held down consumption growth by about half a percent. Higher taxes on high-income households from the American Taxpayer Relief Act likely had little impact on spending due to their smaller aggregate size and the relatively low marginal propensities to consume for high-income households. Also, by reducing the medium- and long-term budget deficit, the higher tax rates on high-income households will contribute to stronger and more-sustainable growth over time.

Strong gains in aggregate household net worth -- both an increase in assets and a decline in the debt burden -- have supported aggregate consumer spending. Debt service (that is, required minimum payments on household debt) has fallen from 13 percent of disposable income at the end of 2008 to 10 percent by the third quarter of 2013 (the latest data available, Figure 2-7). Some of the decline in debt service is due to declines in interest rates on mortgages and consumer credit, but some of the decline is due to declines in the ratio of household debt to income, a process called deleveraging. Debt has fallen from about 1.3 times annual income in 2008 to 1.1 times annual income by the third quarter of 2013 -- with most of the decline in this ratio due to rising nominal incomes, although nominal debt has edged down 5 percent. Together, these declines in household debt and debt-service relative to income show that the household sector as a whole has progressed in reducing these burdens. Although these figures are relevant for projecting aggregate output and consumption growth, they do not reflect the change in debt and debt service for moderate-income and median-income households who, in many cases, continued to face challenges in 2013.

Overall wealth also grew in 2013, as shown in Figure 2-8. Although these wealth increases were in all categories of holdings, wealth likely increased substantially more for high-income households (which have a larger share of their wealth in equities) than for the typical household (which has more of their wealth in housing that appreciated more slowly than equities in 2013). As a result, this suggests that wealth inequality continued to grow as middle-class families faced persistent economic challenges. While gains in stock-market wealth have been happening since the trough of the recession in 2009, those increases were particularly sharp during 2013, when the Wilshire 5000 stock market index increased 31 percent. During the four quarters of 2013, stock market wealth is estimated to have increased by an amount equivalent to 39 percent of annual disposable income. Housing wealth (net of mortgage liability) also increased notably during the year. Housing prices, as measured by the CoreLogic National House Price Index, hit bottom around March 2011 and have increased 11 percent during the 12 months of 2013. As a result, net housing wealth is on track to increase by another 13 percent of annual disposable income in 2013.

The increases in stock market and housing assets point to an increase in the ratio of net worth to income amounting to 52 percent of annual disposable income. An increase in wealth raises annual consumer spending by about 3 percent of that increase. As a result, the expansion of wealth alone could support a rise in consumption of 1.7 percent of disposable income, or more than enough to offset the rise in taxes in 2013.

 

Figure 2-7

 

Household Deleveraging, 1990-2013

 

 

 

Source: Federal Reserve Board, Financial Accounts of the United States.

 

Figure 2-8

 

Consumption and Wealth Relative to

 

Disposable Personal Income (DPI), 1952-2013

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts; Federal Reserve Board, Financial Accounts of the United States: CEA calculations.

Looking ahead, consumer spending in 2014 is likely to grow faster than its 2-percent rate during the past three years. The rise in wealth and the progress in deleveraging have created a more-stable platform on which to base the growth of consumer spending. The rapid growth of consumer durables during 2013 (5.6 percent) is likely to continue or increase further. The average age of light motor vehicles on the road has risen to 11.4 years and it appears likely that some pent-up demand remains for motor vehicles and other durables whose purchases have been delayed during the recession and the slow recovery.

Business Investment

Business Fixed Investment. Real business fixed investment grew moderately, 3.0 percent during the four quarters of 2013, down from a 5.0 percent increase during 2012. The slower pace of business investment during 2013 was concentrated in structures and equipment investment, while investment in intellectual property products grew faster in 2013 than the year earlier. Investment in nonresidential structures declined 0.2 percent following robust growth of 9.2 percent during 2012. Investment in equipment slowed to 3.8 percent, following a 4.5 percent increase in 2012. In contrast, investment in intellectual property products picked up to 4.0 percent during 2013 from 2.9 percent in 2012. (In July 2013, as part of a comprehensive revision to the National Income and Product Accounts, the Bureau of Economic Analysis revised its classifications for business fixed investment to include 1) Research and Development and 2) Entertainment, Literary, and Artistic originals in a new category of Intellectual Property Products, which also includes software investment. See Box 2-1 on the July 2013 benchmark of the National Income and Product Accounts.)

Within equipment investment, major components such as information processing equipment and transportation equipment posted less robust growth in 2013 than in 2012, offsetting stronger growth in industrial equipment investment. Within investment in information processing equipment, declines were posted in investment in computers and photocopy equipment. Within transportation equipment, growth was not as fast as 2012 for investment in autos, aircraft, and ships.

Real investment in nonresidential structures edged down 0.2 percent during the four quarters of 2013, down from growth of 9.3 percent in 2012. Solid growth in petroleum and natural gas drilling was offset by declines in the construction of manufacturing structures and power and communication facilities.

Real GDP Before and After

 

the 2013 Comprehensive Revision, 2007-2013

 

 

 

Funds for investment were also easily available from internal sources such as undistributed profits and depreciation. For the nonfinancial business sector, the sum of these sources, known as cash flow in national income accounting, was 10.1 percent of GDP in the first three quarters of 2013, well above the historical average of 8.7 percent. Historically, nonfinancial corporate investment averages 103 percent of cash flow, with the sector as a whole borrowing from banks and the public for the rest. In contrast, during the first three quarters of 2013, investment was only 90 percent of cash flow. The cash flow that was not available for investment appears to have been spent on share repurchases, a way of returning funds to shareholders that is similar to dividends, but more volatile.

With interest rates low and internal funds readily available, the growth rate of investment might be attributable to low expectations of output growth. In a relationship known as "the accelerator," the growth of investment is related to the change in growth (that is, the acceleration) of output, as shown in Figure 2-9. For example, when output accelerated in 2010 (that is, when output growth increased from negative in 2009 to positive in 2010), investment increased very fast so that the capital stock could service the new level of demand. But when business output growth settled down to an annual rate of roughly 3 percent during the three years through 2013, investment did not need to grow so fast, and indeed it has slowed, as shown in Figure 2-9.

 

Figure 2-9

 

Business Investment and the Acceleration

 

of Business Output, 1965-2013

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts.

Inventory Investment. Inventory investment made a substantial contribution to real GDP growth during the four quarters of 2013 when it accounted for 0.8 percentage point of the 2.5 percent total growth. An increase in agricultural inventory investment accounts for 0.3 percentage point of that overall 0.8 percentage-point contribution and reflects the rebound to a strong harvest following a severe drought in 2012. In the manufacturing and trade sector, the buildup of inventories through the year was no faster than sales, so that by December, inventory stocks were at a 1.30 months' supply, roughly the same level as at year-end 2012.

State and Local Governments

Although State and local governments continued to experience fiscal pressure in 2013, the four-year contraction in the sector -- measured in terms of both purchases (consumption and investment) and employment -- finally appears to have ended. State and local purchases, which had generally declined for 13 quarters through the first quarter of 2013, ended the year at a higher level than in the first quarter, marking its first increase over three quarters since 2009. The cumulative decline in State and local purchases during this recovery contrasts with the usual experience during recoveries (Figure 2-10). In a typical recovery, growth in State and local government bolsters the economic recovery. In contrast, declines in State and local government have been a headwind to private-sector growth and hiring during the first four years of this recovery.

 

Figure 2-10

 

Real State and Local Government Purchases During Recoveries

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts; National Bureau of Economic Research; CEA calculations.

Similar to the 2013 pickup in spending, State and local employment has begun to show signs of life, adding 32,000 jobs during the 12 months of 2013, after shedding almost 700,000 jobs from the end of the recession through year-end 2012.

Despite these positive signals during 2013, major obstacles to growth remain: in particular, the burden of unfunded pension obligations of State and local governments. In its benchmark revision to the National Income and Product Accounts of the United States in July 2013, the Commerce Department, in cooperation with the Federal Reserve, began to measure State and local defined-benefit plans on an accrual basis rather than a cash basis, thereby tracking funded and unfunded pension liabilities. As can be seen in Figure 2-11, the size of these liabilities relative to State and local receipts ballooned immediately after the recession and remains elevated at a level that is currently at about 60 percent of a year's revenue for the sector. Adding in State and local bond liabilities does not change the shape of the plot shown in the figure, although they elevate the level of the liabilities-to-receipts ratio to about two hundred percent of a year's revenue.

 

Figure 2-11

 

State and Local pension Fund Liabilities, 1952-2013

 

 

 

Source: Federal Reserve Board, Financial Accounts of the United States.

International Trade

In 2013, U.S. exports of goods and services to the world averaged nearly $189 billion a month and imports averaged nearly $229 billion a month (Figure 2-12). Exports accounted for 13.5 percent of U.S. production (GDP) in 2013, the same as in 2011 and 2012.

The U.S. trade deficit, the excess of the Nation's imports over its exports, averaged nearly $40 billion a month in 2013. Import demand fell during the recession and, as a result, the trade deficit fell from $66 billion in July 2008 to $25 billion in May 2009. Exports fell too because of recession-related declines in domestic demand abroad (see Figure 2-13), but the recession was not as severe in many parts of the global economy as in the United States. Since May 2009, growth rates of exports and imports have each been averaging about 0.8 percent a month.

Figure 2-13 suggests that slower economic growth among our main trading partners dampens U.S. export growth. In recent years, the top five destinations for U.S. exports, in order from highest to lowest typically were: Canada, the European Union, Mexico, China, and Japan. While growth generally slowed in all these trading partners, it actually turned to recession for a time in our No.2 (European Union) and No. 5 (Japan) export recipients, and their recoveries look to be gradual. In the European Union, real GDP fell 0.7 percent during the four quarters of 2012, then grew 1.1 percent during 2013, and is forecasted to grow 1.4 percent during 2014 (European Commission 2013). Japan's real GDP fell 0.4 percent during the four quarters of 2012, but grew 2.7 percent during 2013, but is projected to edge up only 0.6 percent in 2014 (OECD 2013).

 

Figure 2-12

 

Trade in Goods and Services, 2007-2013

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts.

 

Figure 2-13

 

U.S. Exports Growth, 2009-2013

 

 

 

 

Figure 2-14

 

Current Account Balance, 1985-2013

 

 

 

Source: Bureau of Economic Analysis, National Income and Product Accounts.

The trade balance is the major component of the current account balance. Other components of the current account balance include net income on overseas assets and unilateral transfers such as foreign aid and remittances. The United States has run a current account deficit in all but two quarters since 1985; however, the trend from 1990 through the mid-2000s of ever-increasing deficits appears to have reversed. Figure 2-14 shows the current account balance as a percentage of GDP since 1985. Since peaking at more than 6 percent of GDP in the fourth quarter of 2005, the current account balance has fallen as a share of GDP by more than 3 percentage points. The sharpest decrease occurred during the recession of 2008-09, and although there have been some periods of increase since then, the current account deficit recently reached a 15-year low in the third quarter of 2013 of 2.3 percent. An important driver of the decrease in current account deficit in recent years is the increased domestic production of oil and gas, and the associated reduced demand for imported oil, a shift discussed in more detail later in the chapter. Removing oil, which depends on prices that are set on world markets, the U.S. current account deficit is substantially smaller.

The United States has one of the most open and transparent trade and investment regimes in the world, with a trade weighted applied tariff of 1.3 percent, making it a friendly market for imports and foreign investment. A prime motivation behind U.S. trade policy initiatives is to ensure that our accommodative trade and business environment is reciprocated when U.S. actors have the same opportunities to compete in other markets that foreign exporters and investors have in the United States. U.S. trade policy also seeks to level the playing field, including by seeking to raise standards abroad so they are closer to our own in key areas such as intellectual property, labor, and environment. Box 2-2 discusses Administration trade policy initiatives.

Housing Markets

Housing activity continued its recovery in 2013 despite headwinds from mortgage interest rates that rose approximately 1 percentage point in mid-summer, continued tight credit conditions, and waning investor demand for foreclosed properties. On the production side, new housing starts for both single-family and multi-family structures continued their 2012 growth during 2013, despite relatively higher mortgage rates. For 2013 as a whole, starts were roughly 930,000 units, up from 780,000 in 2012, and up from an all-time low of 554,000 units in 2009 (Figure 2-15).

Demand for housing increased, with new and existing home sales reaching their highest levels in 2013 since the Great Recession. With the lowest level of mortgage delinquencies and foreclosure completions in five years, the composition of sales shifted markedly to non-foreclosure properties as fewer households sold homes under distressed conditions.

Supported by a tight supply of homes for sale, housing prices climbed further in 2013, according to every major measure of house prices (Figure 2-16). As of November 2013, quality-adjusted house prices -- as measured by the FHFA index -- were 7.7 percent higher than their year-ago level and 15.3 percent higher than at their trough in early 2011. Two considerations provide some context for the brisk growth in house prices in 2013. First, such behavior appears to be typical following recessions. Even though house prices bottomed out well after the end of the Great Recession, the recovery since then has, on net, been at a rate just below the average growth rate in house prices seen during the aftermath of the eight post-war recessions of the 20th century. Second, house prices at the end of 2013 appear close to their long-run relationship with rents, one measure of housing's fundamental value. During the mid-2000s, house prices increased much more rapidly than did rents before plummeting. The recent growth in house prices has left prices broadly in line or perhaps above their long-run relationship with rents, which suggests that much of these increases have been tied to improving economic fundamentals.

U.S. Good and Services Imports and Exports, 2012

 

 

 

² USTR. 2013b. "Acting Deputy U.S. Trade Representative Wendy Cutler discusses Japan and the TPP at the Peterson Institute for International Economics." ("http://www.ustr.gov/about-us/press-office/blog/2013/November/Cutler-TPP-Japan-PIIE"). The TPP participants are: Australia, Brunei Darussalam, Canada, Chile, Japan, Malaysia, Mexico, New Zealand, Peru, Singapore, Vietnam, and the United States.

³ USTR. 2013b. "Ambassador Froman discusses the Transatlantic Trade and Investment Partnership at the Munich Security Conference." ("http://www.ustr.gov/about-us/press-office/blog/2013/November/Froman-Munich-Security-Conference").

Another indication that housing market activity is holding steady: households remain optimistic about home prices, according to the Reuters/Michigan Survey of Consumers. Housing affordability remains high and 77 percent of households report that it is a "good time to buy a house" (Reuters/Michigan Survey of Consumers).

 

Figure 2-15

 

Housing Starts, 1960-2013

 

 

 

Source: Census Bureau, New Residential Construction.

 

Figure 2-16

 

National House Price Indexes, 2000-2013

 

 

 

Source: Zillow; CoreLogic; FHFA; S&P/Case-Shiller.

At a more fundamental level, pent-up demand for housing due to suppressed levels of household formation since 2009 is likely to boost housing demand and to help absorb the large supply of vacant homes and homes still in the foreclosure process. During the Great Recession, the number of new households forming each year dropped to below 1 million a year and has remained low ever since. As Figure 2-17 shows, during the housing bubble of the mid-2000s more homes were built than were consistent with the underlying rate of household formation based on demographic trends that would call for about 1.6 million new housing units a year. This oversupply peaked in 2007 and -- because of low levels of home construction -- this oversupply began to fall. And by 2011, the oversupply turned into an undersupply. The increase in the stock of homes now lags behind the usual rates of household formation.

As employment prospects improve, household formation is likely to pick up. However, the extent to which the increase in the number of households translates into stronger housing demand depends critically on the easing of credit standards (that might have been over-tightened following the financial crisis), particularly for first-time homebuyers. In 2013, lending standards eased somewhat for prime residential mortgages, according to the Federal Reserve's Senior Loan Officer Opinion Survey, and this easing helped support a rise in mortgage purchase originations from the low levels seen in recent years.

Energy

In 2013, the United States continued to benefit from developments in the oil and gas sectors, as well as from growth in energy efficiency and the production and integration of renewable energy. As shown in Figure 2-18, net petroleum imports have fallen from more than 12 million barrels a day in 2005 to approximately 6.2 million barrels a day in 2013. Moreover, as shown in Figure 2-19, beginning in October 2013, domestic crude oil production exceeded crude oil imports for the first time since 1995.

Crude and refined oil products constitute the vast majority of the country's energy imports. This reduction in energy imports has multiple benefits: it has been a major driver of the improvement in the U.S. balance of trade, it reduces the vulnerability of the U.S. economy to foreign oil supply disruptions, and it supports American jobs both in energy production and in manufacturing. The dramatic increase in domestic oil and natural gas production added about 0.2 percentage point to U.S. GDP growth in both 2012 and 2013.

 

Figure 2-17

 

Cumulative Over- and Under-Building of Residential

 

and Manufactured Homes, 1996-2013

 

 

 

The ongoing trend toward reduced energy imports is driven both by roughly stable energy demand and increases in domestic energy supply. Overall, economy-wide energy use has declined 0.8 percent at an annual rate since 2007. The increase in domestic energy supply reflects major gains in unconventional oil and natural gas production. The sharp increase in unconventional domestic gas production has led to a 73 percent drop in the wholesale (Henry Hub) price of natural gas from a high of $13.42 in October 2005 to $3.68 in October 2013. The United States is now the largest producer of natural gas in the world, and the 2013 International Energy Outlook projects that the United States will remain the largest producer through 2030 (U.S. Energy Information Administration 2013). Since 2007, over 50,000 jobs have been created in oil and natural gas extraction alone, with more than 160,000 jobs being created along the oil and natural gas supply chain. Low natural gas prices also help manufacturing as discussed below, and have been an important driver in the reduction of U.S. carbon dioxide emissions as electricity production has shifted from coal to cleaner-burning natural gas. Indeed, between 2010 and 2013, the total U.S. carbon dioxide emissions from energy consumption decreased by 4.3 percent. In addition to providing cost savings to consumers today, this reduction in greenhouse gas emissions will benefit future generations.

 

Figure 2-18

 

Petroleum Net Imports, 1980-2015

 

 

 

 

Figure 2-19

 

Monthly Crude Oil Production and Net Imports, 1990-2013

 

 

 

Source: Energy Information Administration, Petroleum Supply Monthly.

 

Figure 2-20

 

Wind and Solar Energy Production, 2000-2013

 

 

 

Source: Energy Information Administration, Monthly Energy Review, Short-Term Energy Outlook.

The other part of the energy supply story, shown in Figure 2-20, is the dramatic growth in wind and solar electricity production, which have each more than doubled since President Obama took office. In 2012, a record 13 gigawatts of new wind power capacity was installed, roughly double the amount of newly installed capacity in 2011. More than 5 gigawatts were installed in December 2012 alone as firms scrambled to take advantage of the expiring 2.3 cent per kilowatt-hour production tax credit (Congress later extended the tax credit for 2013). These 13 gigawatts of new wind capacity represented the largest share of additions to total U.S. electric generation capacity in 2012.

In addition to increased domestic supply, energy imports have declined because of reduced energy demand across all the main energy sectors. As shown in Figure 2-21, gasoline demand per capita rose through the early 2000s and plateaued in the mid-2000s before dropping substantially during the recession. As the economy has recovered, however, gasoline demand per capita has continued to fall. Some of this continued decline in gasoline demand stems from the relatively high real gasoline prices shown in Figure 2-21, but that is only a partial explanation. Increasing fuel efficiency brought about by Federal fuel efficiency standards also played a role; and, in 2012, the Administration finalized fuel economy standards that, together with the Administration's first round of standards, will nearly double the fuel economy of light-duty vehicles to the equivalent of 54.5 miles per gallon by the 2025 model year from 2010 levels. Further, beginning in model year 2014, medium- and heavy-duty trucks must meet new energy efficiency standards as well, which will increase their fuel efficiency by 10 to 20 percent by 2018.

 

Figure 2-21

 

U.S. Per Capita Consumption of Gasoline and Real

 

Gasoline Prices, 2000-2013

 

 

 

Source: Energy Information Administration, Monthly Energy Review; Census Bureau; CEA calculations.

Despite these significant improvements in energy efficiency and reductions in energy-related carbon dioxide emissions, continued work is needed to reduce greenhouse gas emissions. In June 2013, the President laid out his Climate Action Plan (summarized in Box 2-3), which aims to reduce both greenhouse gas emissions and the impact of climate change on future generations.

The major U.S. labor market indicators continued to recover during 2013 even as the unemployment rate remained unacceptably high. As shown in Figure 2-22, the unemployment rate dropped 1.2 percentage points during the 12 months of 2013, somewhat faster than the average 0.9 percentage point annual drop during the three preceding years. Similarly, as shown in Figure 2-23, establishment employment finished its third year of growth at roughly 2.3 million a year (or about 190,000 a month)². The strength of the labor market was not matched by the growth of output, with some puzzling developments in the relation between the unemployment rate and GDP, and also the relationship between employee-hours and output (productivity).

 

Figure 2-22

 

Unemployment Rate, 1979-2014

 

 

 

Source: Bureau of Labor Statistics, Current Population Survey.

 

Figure 2-23

 

Nonfarm Payroll Employment, 2007-2014

 

 

 

Source: Bureau of Labor Statistics, Current Employment Statistics.

 

Figure 2-24

 

Unemployment Rate by Duration, 1990-2014

 

 

 

Source: Bureau of Labor Statistics, Current Population Survey

The current elevation of the unemployment rate is entirely due to long-term unemployment. In December 2013, the unemployment rate for workers unemployed 26 weeks or less fell to lower than its average in the 2001-07 period, while the unemployment rate for workers unemployed 27 weeks or more remained higher than at any time prior to the Great Recession. But the long-term unemployment rate has declined by 1.1 percentage points in the last two years, a steeper decline than the 0.5 percentage point drop in the short-term unemployment rate over that period (Figure 2-24).

 

Figure 2-25

 

Predicted vs. Actual Manufacturing

 

 

 

Source: Bureau of Labor Statistics, Current Employment Statistics; CEA calculations.

Of the 2.3 million increase in payroll employment during the 12 months of 2013, about 4 percent was in manufacturing, 7 percent was in construction, and 90 percent was in the private service-providing industries. Within the service-providing industries, the sectors showing the strongest job growth were professional and business services (29 percent of total employment growth), retail trade (15 percent) and health care (9 percent of the total).

Over the course of the recovery, manufacturing has added 622,000 jobs since its trough. Some have pointed to this growth, following a decade of job losses, as indicating a resurgence in manufacturing, while others have suggested that this rebound simply reflects the normal cyclical pattern given the depth of the recession. The Council of Economic Advisers (CEA) analysis suggests that while the overall recovery did in fact contribute to the stabilization of manufacturing job losses, the job gains are about 500,000 above and beyond what would be associated with the historical cyclical pattern (Figure 2-25).

Further evidence of the healing of the job market comes from the number of job vacancies, which increased 6 percent during the 12 months through November (the latest available at press time). There are now 2.6 unemployed workers for each job vacancy, less than half of the number following the business-cycle trough in 2009, but still in excess of the average two-to-one ratio from 2001 to 2007.

Wage Growth and Price Inflation

Hourly compensation (including non-wage benefits) increased 2.0 percent during the 12 months of 2013, the fourth consecutive year of growth at around a 2-percent rate, according the Employment Cost Index. Prices in the nonfarm business sector increased at a 1.6 percent annual rate during these four years; so from the viewpoint of a typical employer, the real product hourly compensation increased 0.4 percent at an annual rate. These four-year growth rates for real hourly compensation were less than the 1.2 percent increase in labor productivity, and as a result, the labor share of nonfarm business output (and of gross domestic income) declined.

Growth in real wages (that is, take-home wages not including benefits) of production workers picked up to 0.7 percent in 2013 from a 0.1 percent decline a year earlier. Nominal wages increased 2.2 percent in 2013 (up from a year earlier) while prices for wage earners rose 1.5 percent (down from a year earlier).

Consumer prices excluding food and energy (the core CPI) rose 1.7 percent during the 12 months of 2013, down from 1.9 percent during 2012. Overall, consumer prices rose just 1.5 percent during the year as food prices increased only 1.1 percent and energy prices inched up 0.5 percent.

Although inflation edged lower in 2013, the relative stability of inflation during the recession and slow recovery presents a puzzle. During this period, the unemployment rate has been much higher than its long-term average, and higher than the rate that is generally considered consistent with stable inflation. Under these circumstances, conventional economic theory and historical experience would have expected declining inflation and perhaps even negative inflation. In contrast, inflation has remained fairly stable since the business-cycle peak with the 12-month change in core CPI inflation never falling below 0.6 percent, raising a puzzle of missing disinflation. Standard explanations of the missing disinflation focus on anchored expectations arising from increased Federal Reserve credibility associated with targeting an inflation rate of approximately 2 percent (for example, Fuhrer and Olivei 2010, Stock and Watson 2010, Ball and Mazumder 2011).

In addition to anchored expectations, a second factor behind the lack of disinflation appears to be the unusually high fraction of the long-term unemployed in this recovery. Those unemployed for only short durations search more intensely for a new job (Krueger and Mueller 2011) and are also potentially more likely to match with a good job, which suggests that the short-term unemployed put more downward pressure on wages than those who have been unemployed for more than six months. While the relationship between the overall unemployment rate and inflation in recent years is puzzling, the relationship between short-term unemployment and inflation is less so, as discussed in Box 2-4.

Growth in Expected Average Hourly Earnings vs.

 

Short-Term Unemployment, 1976-2013

 

 

 

Core Consumer Price Inflation: Actual vs. Predicted, 2006-2013

 

 

 

Although real GDP has grown at a roughly 2 percent rate for each of the past three years, a foundation is in place for faster growth during 2014, as most components of demand point to faster growth while the supply side does not appear constraining. Although fiscal policy has generally increased the level of output, it has been a drag on GDP growth in the last several years and especially in 2013. The rate of decline in the deficit-to-GDP ratio will likely moderate in 2014 under the President's Budget policy as well as under current law, as noted earlier in this chapter. Consumer spending likely has adjusted by now to the expiration of the payroll tax cut, but it probably has not adjusted to the gains in housing and stock market wealth. End-of-2013 indicators suggest that growth among our European trading partners is looking up, suggesting stronger exports in 2014 than in 2013. While not much growth can be expected from real State and local spending, the latest quarterly data suggest that it will no longer be a substantial drag on overall growth. As discussed earlier in the chapter, firms appear ready to step up business investment if consumer spending picks up. Business investment will grow if everything else does. With the unemployment rate in January 2014 at 6.6 percent and the capacity utilization rate in manufacturing at about 77 percent, the economy has room to grow.

The Administration's economic forecast, as finalized on November 21, 2013 is presented in Table 2.1, and is the forecast that underpins the President's FY 2015 Budget. The Administration expects real GDP to accelerate from a 2.3 rate during the four quarters of 2013 to 3.3 percent during 2014. (Data released after the forecast was finalized show a slightly faster-than expected growth rate during 2013, 2.5 percent rather than 2.3 percent.) These projections, as is standard for the Administration's budget forecast, assume enactment of the President's Budget -- including the Opportunity, Growth and Security initiative.

The forecast assumed that the unemployment rate would fall 0.5 percentage point in the four quarters of 2014. Since the forecast was finalized in November the unemployment rate has fallen from 7.3 percent (as first published for October) to 6.6 percent in January 2014, considerably faster than the pace forecasted by the Administration or by the consensus of private sector forecasters. As a result, the Administration's budget forecast of an unemployment rate averaging 6.9 percent in 2014 does not reflect the latest information and an updated projection would forecast a continued decline in the unemployment rate over the course of the year. A revised Administration forecast will be released in the Mid-Session Review of the Budget over the summer.

 

Table 2-1

 

Administration Economic Forecast

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 85.]

 

 

 

Nominal interest rates are currently low due to the fact that the economy has not fully healed together with monetary policy that has kept rates low across a wide range of Treasury securities. Consistent with the forward policy guidance at the time that the forecast was made, interest rates are projected to increase for maturities that extend through periods covering dates when the unemployment rate is expected to fall below 6.5 percent. Interest rates are expected to continue to climb as the economy approaches full employment. After that point, projected real interest rates (that is, nominal rates less the projected rate of inflation) will be close to their historical average. These interest-rate paths are close to those projected by the consensus of professional economists.

Growth in GDP over the Long Term

As discussed earlier, the growth rate of the economy over the long run is determined by the growth of its supply-side components, demographics, and technological change. The growth rate that characterizes the long-run trend in real U.S. GDP -- or potential GDP -- plays an important role in guiding the Administration's long-run forecast. Through 2020, potential real GDP is projected to grow at a 2.4 percent annual rate, before slowing to 2.3 percent during the three-year period 2021-24. These growth rates are slower than in the past because of the movement of the baby-boom generation into the retirement years. These growth rates for potential real GDP are based on the assumption of no change to immigration law. If, however, immigration law were to be revised along the lines of the Border Security, Economic Opportunity, and Immigration Modernization Act (S.744) that the Senate approved in June, the growth rate of potential real GDP would be higher, because of faster growth of the working-age population and increased total factor productivity growth (Box 2-5). The Budget totals reflect the effects of immigration reform by incorporating the CBO score directly into the Budget. This CBO score incorporates both direct policy effects and the broader economic impact. In order to avoid double counting with this estimate, the economic forecast does not reflect the effects of immigration reform.

Table 2-2 shows the Administration's forecast for the contribution of each supply-side factor to the growth in potential real GDP: the working-age population, the rate of labor force participation, the employed share of the labor force, the ratio of nonfarm business employment to household employment, the length of the workweek, labor productivity, and the ratio of real GDP to nonfarm output. Each column in Table 2-2 shows the average annual growth rate for each factor over a specific period of time. The first column shows the long-run average growth rates between the business-cycle peak of 1953 and the business-cycle peak of 2007, with business-cycle peaks chosen as end points to remove the substantial fluctuations within cycles. The second column shows average growth rates between the fourth quarter of 2007 and the third quarter of 2013, a period that includes the 2007-09 recession and the recovery so far. The third column shows the Administration's projection for the entire 11-year forecast period, from the third quarter of 2013 to the fourth quarter of 2024. And the fourth column shows average projected growth rates between the fourth quarter of 2020 and the fourth quarter of 2024; that is, the last four years of the forecast interval when the economy is assumed to settle into steady-state growth.

 

Table 2-2

 

Supply-Side Components of Actual and Potential

 

Real GDP Growth, 1952-2024

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 87.]

 

 

 

The employed share of the labor force -- which is equal to one minus the unemployment rate -- is expected to increase at an average 0.2 percent a year over the next 11 years. It is expected to be unchanged after 2018 when the unemployment rate converges to the rate consistent with stable inflation. The workweek is projected to be roughly flat during the forecast period, somewhat less of a decline than its long-term historical trend of -0.3 percent. The workweek is expected to stabilize because some of the demographic forces pushing it down are largely spent, and because a longer workweek is projected to compensate for the anticipated decline in the labor force participation rate.

Labor productivity is projected to increase 2.1 percent a year over the forecast interval and 2.2 percent in the long run (line 6, columns 3 and 4), roughly the same as the average growth rate from 1953 to 2007 (line 6, column 1). The elevated rate of long-term unemployment poses some risk to the projection insofar as the human capital of workers may deteriorate with prolonged unemployment. That said, higher rates of school enrollment among young adults in recent years, as noted, should contribute to productivity growth in the coming years.

The ratio of real GDP to nonfarm business output is expected to subtract from GDP growth over the projection period (line 7, column 3), consistent with its long-run trend. The nonfarm business sector generally grows faster than government, households, and nonprofit institutions, where an accounting convention holds productivity growth to zero.

Summing the growth rates of all of its components, real GDP is projected to rise at an average 2.7 percent a year over the projection period (line 8, column 3), somewhat faster than the 2.3 percent annual growth rate for potential real GDP (line 9, column 3). Actual GDP is expected to grow faster than potential GDP primarily because of the projected rise in the employment rate (line 3, column 3) as millions of currently unemployed workers find jobs.

Real potential GDP (line 9, column 4) is projected to grow more slowly than the long-term historical growth rate of 3.3 percent a year (line 9, column 1). As discussed earlier, the projected slowdown in real potential GDP growth primarily reflects the lower projected growth rate of the working-age population and the retirement of the baby-boom cohort. If the effects of immigration reform were incorporated into this forecast, however, then it would show a higher real potential GDP growth rate.

Even with this growth, however, the economy would remain below its full potential and the unemployment rate would remain unacceptably high. Additional sound policies would speed the return of the economy to its full potential, including policies like investments in infrastructure and increasing certainty through business tax reform. Conversely, adverse policy developments in the United States or adverse shocks in the United States or abroad could impede this favorable scenario.

² The Department of Labor conducts several labor market surveys. The household survey -- conducted in cooperation with the Bureau of Census -- queries 60,000 households every month with a variety of questions including whether members of that household were working or looking for a job, and this survey is the source of the unemployment rate, among other important statistics. The Establishment (or Payroll) survey queries employers about how many workers they employed, how many hours did they work, and what they were paid. The Establishment survey is the source of the most quoted figures for job growth. The Job Openings and Labor Turnover Survey (JOLTS) (a relatively new survey, begun in 2000) also queries employers about their job openings (vacancies) as well as their hiring, quits, and layoffs.

The purpose of the Recovery Act was to provide countercyclical fiscal support for the economy as part of a suite of monetary and fiscal policies aimed at containing the already-severe recession that, had it spiraled further, could have resulted in a second Great Depression. The Act was also intended to lay the foundation for a stronger and more resilient economy in the future.

In the four years following the Recovery Act, the President built on this initial step, signing into law over a dozen fiscal measures aiming to speed job creation. These measures, which extended key elements of the Recovery Act and provided new sources of support, were motivated by a deepening understanding of the severity of the initial shocks to the economy, as well as by new challenges that subsequently arose. These additional measures nearly doubled the size and impact of the Recovery Act's fiscal support to the economy through the end of 2012.

Nearly half of the jobs measures in the Recovery Act and subsequent legislation, or $689 billion, were tax cuts -- with most of them directed at families. The other half was for investments in critical areas such as rebuilding bridges and roads, supporting teacher jobs, or providing temporary help for those who found themselves unemployed because of the impact of the Great Recession.

The economic picture today is much brighter. GDP per capita started expanding in the third quarter of 2009 and reached its pre-crisis level in about four years, considerably faster than the historical record suggests is the typical pace of recovery following a systemic financial crisis.¹ Since 2010, the U.S. economy has also consistently added over 2 million private-sector jobs a year, bringing the overall unemployment rate down to its lowest level since October 2008. Job growth has been broad-based across sectors and has withstood significant headwinds, including more recent fiscal contraction at all levels of government, and concerns stemming from the European sovereign debt crisis.

As part of the unprecedented accountability and transparency provisions included in the Recovery Act, the Council of Economic Advisers (CEA) was charged with providing to Congress quarterly reports on the effects of the Recovery Act on overall economic activity, and on employment in particular. In this chapter, CEA provides an assessment of the effects of the Act through the third quarter of 2013, and of subsequent jobs measures through 2012.

This chapter assesses the role of the Recovery Act and the subsequent jobs measures in helping to facilitate the economic turnaround since 2009. It updates previous estimates from the Council of Economic Advisers and other sources on the Act's contribution to employment and output growth, and expands the estimates to reflect the impact of the full set of fiscal measures undertaken. The chapter also considers how many investments contained in the Recovery Act have laid the groundwork for a more productive economy in the years ahead and will support growth long after the spending authorized by the Act has fully phased out.

Consistent with the preponderance of evidence from numerous private-sector, academic, and government analyses, this chapter finds that the Recovery Act substantially boosted employment and output. CEA estimates that, by itself, the Recovery Act saved or created an average of 1.6 million jobs a year for four years through the end of 2012 (cumulatively, equivalent to about 6 million job-years, where a job-year is defined as one full-time job for one year). In addition, the Recovery Act alone raised the level of GDP by between 2 and 2.5 percent from late 2009 through mid-2011. The Recovery Act also helped individuals, businesses, and State and local governments directly affected by the downturn, and put the economy on a better trajectory for long-run growth by undertaking targeted investments in education, energy, and health care, among other areas.

Combining effects of the Recovery Act and additional countercyclical fiscal legislation that followed, CEA estimates that the cumulative gain in employment was about 9 million job-years through the end of 2012. The cumulative boost to GDP from 2009 through 2012 is equivalent to 9.5 percent of fourth quarter 2008 GDP.

While these estimates are substantial, they still understate the full impact of the Administration's economic policies in tackling the Great Recession due to being based only on the effect of fiscal measures. CEA estimates do not account for the broader set of responses that included policies to stabilize the financial system, rescue the auto industry, and provide support for the housing sector -- in addition to the independent actions undertaken by the Federal Reserve.

Fallout from the housing crisis quickly spread to the broader economy through a complex web of opaque financial instruments and questionable business practices, including excessive leverage and an overreliance on short-term debt (Financial Crisis Inquiry Report 2011). Investors pulled back from risky assets and, during one fateful week in September 2008, the investment bank Lehman Brothers went out of business, a prominent money market fund "broke the buck" (meaning that depositors could no longer count on getting their money back in its entirety, an almost unprecedented event), and the large insurance firm American International Group (AIG) teetered on the edge of bankruptcy until the U.S. government provided $85 billion in financial support.

This financial turmoil led to sharp declines in real economic activity. From the third quarter of 2007 through the first quarter of 2009, the economy lost more than $13 trillion in wealth, nearly one-fifth of the total, because of rapidly declining stock and house prices. This was much larger than the initial decline in wealth at the outset of the Great Depression.² Falling asset prices reduced the value of collateral and further restricted the availability of credit and, as credit dried up, many small businesses and even some large, well-known corporations reported trouble meeting basic expenses such as payroll. Faced with extraordinary uncertainty about the economic future, businesses stopped hiring, laid off workers, and shelved investment plans. As housing and financial wealth plummeted and concerns over job security mounted, consumers cut back on spending. The effect was immediate and drastic: in the fourth quarter of 2008, personal consumption expenditures fell by nearly 5 percent and private investment shrunk 31 percent at an annual rate.

Most economic forecasters underestimated the magnitude of the toll these shocks would take on the economy, in large part because the United States had not gone through a systemic financial crisis since the Great Depression. Forecasts made at the time were also subject to considerable uncertainty about the spillovers to the rest of the world, and about how the economy would respond to other macroeconomic policy interventions after the federal funds rate had already hit zero. As shown in Table 3-1, in December 2008, for example, the Blue Chip panel of economic forecasters projected that real GDP would fall at a 1.4 percent annual rate in the first half of 2009, less than half the 2.9 percent annualized rate of decline that actually occurred. Moreover, the Blue Chip panel of forecasters estimated that the unemployment rate would rise to 7.7 percent in the second quarter of 2009, well below the actual rate of 9.3 percent. Other indicators showed similarly large deteriorations relative to forecasts.

Initial Policy Responses

As the economy slid into recession, Congress and the Bush Administration enacted the Economic Stimulus Act of 2008 in February. They designed the Act to counteract a short recession by providing temporary support to consumer spending, but it was not sufficient to reverse the emerging distress and, by design, did not have long-lasting effects. In fall 2008, as the initially mild recession turned into a full-blown financial crisis, the U.S. government mounted a coordinated emergency response to prevent a meltdown of the financial system.³ The Federal Reserve, which had progressively cut its federal funds target rate several times over the previous year, lowered the rate still further in December 2008 to near zero, where it remains to this day.

 

Table 3-1

 

Forecasted and Actual Real GDP Growth and Unemployment Rate

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 95.]

 

 

 

These early policy responses proved fundamental to rescuing the global financial system. They helped repair the balance sheets of both financial and non-financial institutions, restored investor confidence, and restored the flow of credit to struggling businesses and families. Nevertheless, the economy continued to deteriorate, and aggregate demand remained depressed. With the traditional tool of monetary policy, the federal funds rate, reaching its lower bound of zero, conventional countercyclical monetary policy could go no further, and the Federal Reserve ultimately opted for additional, non-standard measures.

The Recovery Act

In early 2008, before the Nation realized the full extent of the economic challenge, fiscal expansion policy was guided by the "3T's" advocated by Summers (2007), Sperling (2007), and Elmendorf and Furman (2008): timely, targeted, and temporary. By the end of 2008, however, it was clear that the recession had turned into a major financial crisis and that a new approach was needed, what Former Treasury Secretary Lawrence Summers called "speedy, substantial, and sustained."⁵

Several principles guided the new Administration's policymaking. First, the fiscal effort was to be implemented speedily, unlike previous incoming presidents' economic programs, which were generally not passed until they were six months or more into office. Second, it should be substantial, given the very large scope of the economic problem. Finally, it should be a sustained effort that would not only have significant spend-out over the first two years, but would continue some temporary support thereafter. The new approach would require a mix of instruments, with some being faster to spend-out, such as tax cuts and other temporary assistance that put cash in the hands of households who immediately needed it. Other components would be more lagged but have larger cumulative countercyclical impacts and greater longer-run benefits, such as investments in infrastructure and innovation. In all cases, however, the measures would end and would not have long-term impacts on the Federal Government's primary budget deficit.⁶

Goals of the Recovery Act. Overall, this approach was embodied in the stated goals of the Recovery Act, as written into the legislation:

(1) To preserve and create jobs and promote economic recovery;

(2) To assist those most impacted by the recession;

(3) To provide investments needed to increase economic efficiency by spurring technological advances in science and health;

(4) To invest in transportation, environmental protection, and other infrastructure that will provide long-term economic benefits;

(5) To stabilize State and local government budgets, in order to minimize and avoid reductions in essential services and counterproductive State and local tax increases.

Scale of the Recovery Act. At passage, CBO estimated that the Recovery Act would cost $787 billion, although this estimate would increase as the full magnitude of the recession became apparent. The most recent CBO estimates show that the fiscal support from the Recovery Act will total $832 billion through 2019.⁷ Of this total, $69 billion was allocated to a routine set of patches for the Alternative Minimum Tax (AMT). This part of the Act, a continuation of a long-standing practice, is best thought of as ongoing fiscal policy, not as a temporary fiscal impulse designed specifically to counter the effects of an economic recession. Excluding the AMT patch, the Recovery Act provided a total fiscal impulse of $763 billion.

Composition of the Recovery Act. The initial cost projections of the Recovery Act showed the law would be fairly evenly distributed across tax cuts ($212 billion), expansions to mandatory programs such as Medicaid and unemployment benefits ($296 billion), and discretionary spending ($279 billion) in areas ranging from direct assistance to individuals to investments in infrastructure, education, job training, energy, and health information technology. More specifically, Figure 3-1 shows how Recovery Act policies excluding the AMT patch can be divided into five functional categories: individual tax cuts, business tax incentives, State fiscal relief, aid to directly impacted individuals, and public investments.⁸

Timing of the Recovery Act Spend-Out. The Nation felt the early effects of the Recovery Act almost immediately, as enhanced Medicaid payments started to flow to states on March 13, 2009 and individual income tax withholdings were reduced by April 1, 2009. As of the third quarter of 2009, roughly one-quarter of all spending and tax cuts had occurred, with another half spread across the four quarters after that, roughly consistent with CBO projections as of 2009. By September 30, 2013, the Federal Government had disbursed $805 billion on Recovery Act programs, as shown in Table 3-2.

As shown in Table 3-3, individual tax cuts, aid to States, and aid to individuals directly affected by the recession were among the first Recovery Act programs to take effect, providing the largest initial boost to spending in fiscal year 2009. Each of these categories tapered after 2010, with only a small amount of outlays in 2012 and 2013, while public investment outlays now constitute the bulk of continuing Recovery Act expenditures.

 

Figure 3-1

 

Recovery Act Programs by Functional Categories

 

 

 

Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury, Office of Tax Analysis, based on the FY2013 Mid-Session Review.

Accountability, Transparency, and Oversight. In keeping with the Administration's commitment to the highest standards of accountability, transparency, and oversight, the Recovery Act took unprecedented steps to track and report the use of Federal funds and to prevent waste, fraud, and abuse. The Act established a Recovery Accountability Transparency Board comprised of an independent director and 12 agency inspectors general, as well as a Recovery Implementation Office that reported directly to the Vice President. Recipients (including vendors, nonprofit organizations, and State and local governments) were required to report regularly to the Board on their use of funds and the number of jobs created or saved.⁹

All of the information received from agencies and recipients has been posted on a website (www.recovery.gov). Users can sort and display data on funding in different ways (by category of funding, by agency, by state), making it easy to obtain and analyze information. The website also offers the opportunity for the public to report fraud or waste. Reported instances of waste, fraud, and abuse remain low -- at less than 1 percent of all grant awards.

 

Table 3-2

 

An Overview of Recovery Act Fiscal Impact

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 99.]

 

 

 

 

Table 3-3

 

Recovery Act Programs by Functional Categories

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 99.]

 

 

 

While the Recovery Act was the first and largest fiscal action undertaken after the financial crisis to create jobs and strengthen the economy, many subsequent actions extended, expanded, and built on the Recovery Act. Parts of the Recovery Act were extended to address the continuing needs of the economy, including Emergency Unemployment Compensation, accelerated depreciation of business investment for tax purposes (that is, "bonus depreciation"), measures for teacher jobs, and aid to states for Medicaid. In other cases, new measures expanded on elements of the Recovery Act, such as the temporary payroll tax cut in 2011 and 2012, which was nearly 50 percent larger than the Making Work Pay credit it replaced, and an even greater allowance for businesses to write off the cost of investments when computing their tax liability (that is, "expensing"). The following measures built on the goals of the Recovery Act and are counted as part of the fiscal impulse in the analysis that follows: the cash-for-clunkers program enacted in summer 2009, an expanded homebuyer tax credit and business tax incentives in fall 2009, the HIRE Act tax credit and additional infrastructure investment incentives in March 2010, a small business tax cut and credit bill in fall 2010, veterans hiring incentives in fall 2011, plus the additional payroll tax cut extensions and unemployment insurance extensions passed in 2011 and 2012. All told, these subsequent jobs measures, listed in Table 3-4, provided an additional $674 billion in countercyclical fiscal support through the end of 2012. This total excludes routine or expected policies such as continuing the 2001 and 2003 tax cuts, passing so-called "tax extenders" to address regularly expiring tax provisions, and fixing Medicare's Sustainable Growth Rate formula.¹⁰

Of the $674 billion in fiscal support following the Recovery Act, 31 percent was accounted for by the payroll tax cut from 2011 to 2012, 24 percent was accounted for by extended unemployment insurance, and the remainder included a variety of actions such as relief for States and tax incentives for businesses. Figure 3-2 shows a breakout of the policies of the Recovery Act and the subsequent jobs legislation.

In addition, the President proposed further measures for the economy that were not passed by Congress, most notably the American Jobs Act, which was proposed in September 2011 and would have provided additional investments -- totaling $447 billion -- in everything from infrastructure to teacher jobs to a robust tax credit for small business hiring.¹¹

Automatic Countercylical Measures

In addition to Obama Administration policies, previously enacted laws have built-in provisions that allow for automatic support when economic conditions worsen. For example, personal income tax payments decline when income declines, and spending on unemployment insurance picks up as more individuals struggle to find work. These automatic responses -- known as "automatic stabilizers" -- can help moderate business cycles (as shown for instance by Auerbach and Feenberg (2000) and Follette and Lutz (2010)) in addition to alleviating the human costs of economic downturns.

 

Table 3-4

 

Fiscal Support for the Economy Enacted After the Recovery Act

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 . page 101.]

 

 

 

 

Figure 3-2

 

Recovery Act and Subsequent Fiscal Measures

 

by Functional Category

 

 

 

Source: Office of Management and Budget, Agency Financial and Activity Reports; Department of the Treasury, Office of Tax Analysis, based on the FY2013 Mid-Session Review; Congressional Budget Office.

 

Figure 3-3

 

Automatic Stabilizers and the Budget Balance, 2009-2013

 

Federal Budget Deficit as a Percent of GDP

 

 

 

 

Figure 3-4

 

Fiscal Expansion as a Percentage of GDP

 

 

 

Source: Congressional Budget Office, The Budget and Economic Outlook: 2014 to 2024; Office of Management and Budget; Bureau of Economic Analysis, National Income and Product Accounts.

Total Fiscal Response

All told, the Great Recession triggered a substantial fiscal response. Figure 3-4 illustrates the scale of the Recovery Act and of the other major fiscal measures implemented by the Administration. As noted earlier, fiscal policy represented only one part of the Administration's broader economic strategy to foster recovery and protect households, as described more fully in Box 3-1.

Overall, CEA estimates that the Recovery Act saved or created about 6 million job-years (where a job-year is the equivalent of one full time job for one year) through 2012 and raised GDP by between 2 and 2.5 percent in FY 2010 and part of FY 2011. These estimates are consistent with estimates made by CBO (2013a) and by independent academic studies, which use a variety of methodologies and data sources. Adding the estimated effect of subsequent fiscal policy measures, CEA model finds that the combined effect of these actions increased GDP above what it otherwise would have been by more than 2 percent a year for three years, and created or saved about 9 million job-years through 2012. Moreover, research on economic growth generally finds that these types of benefits have a long-lasting impact on the economy even after the initial policy impetus has expired. This is even more true when the policy itself included significant measures for long-term growth, as described later in this chapter.

Evaluating effects of fiscal policy in general, and the Recovery Act in particular, is challenging for several reasons. Appendix 2 describes these challenges, and how economists have addressed them, in greater detail. A key issue is that estimating effects entails comparing what actually happened with what might have happened (what economists call the "counterfactual"). However, because counterfactual outcomes are not actually observed, other methods are needed.

Estimating the Short-run Macroeconomic Effect of Fiscal Policy. A key concept for estimating the macroeconomic effects of fiscal policy is what economists call the fiscal multiplier. The fiscal expenditure multiplier is the change in GDP resulting from a $1 increase in government expenditures, and the tax multiplier is the change in GDP resulting from a $1 decrease in taxes. Because a $1 increase in spending or decrease in taxes has ripple effects in subsequent transactions as it passes through the broader economy, theory suggests that the fiscal multiplier may be greater than one -- a $1 increase in spending or reduction in taxes may support an increase in output of more than $1.

The standard theory of fiscal policy in a recession holds that when government demand goes up, firms hire workers and raise production, which boosts employment, income, and GDP. The initial effect is amplified as workers spend additional income, and businesses purchase more raw materials and make investments to meet increased demand. In its most basic form, the government spending multiplier is the sum of the first-round direct effect of spending on GDP, the second-round effect with consumption by those paid for providing goods and services, and the subsequent-round effects. In this model, the multiplier effect depends on the marginal propensity to consume (MPC) -- the fraction of an additional dollar of income that is spent rather than saved.¹² Because the MPC is thought to be large, especially in a recession when individuals face problems borrowing, models can generate multipliers much higher than one. Tax cuts also increase individual income, but the multiplier effect on overall output is generally thought to be slightly less than it is for government expenditures. Because the individual receiving the tax cut saves part of it, the first-round effect on overall spending is smaller, making the subsequent ripple effects smaller as well. Thus, the basic multiplier for a tax cut is less than the government spending multiplier; specifically, the tax multiplier is the spending multiplier times the MPC.

The model is a useful conceptual starting point, but it makes many simplifications. Appendix 2 to this chapter reviews recent theoretical research on the effects of fiscal policy, especially in a deep recession. This research suggests that, in normal times, fiscal multipliers can be small both because consumers save a substantial fraction of a temporary fiscal measure and because monetary policy tends to counteract the fiscal measure in an attempt to maintain stable inflation. In a severe recession, however, especially when monetary policy is constrained by the fact that interest rates cannot drop below zero (the zero lower bound), fiscal multipliers can be much larger. Taking further into account the fact that long-term unemployment can lead to transitions out of the labor force, with a resulting long-term effect of depressing output and employment (an effect referred to as "hysteresis"), multipliers can be larger yet for fiscal policies that support aggregate demand and reduce the average duration of unemployment.

Numerical estimates of fiscal multipliers are typically computed using historical data on fiscal interventions and macroeconomic outcomes, and Appendix 2 also discusses the recent empirical research on this topic. This empirical work provides estimates of multipliers for different types of fiscal interventions (government spending and individual income tax cuts). Once estimated, the resulting multipliers can be used to estimate the macroeconomic effect of the Recovery Act; that is, to compare what happened under the Recovery Act with what likely would have happened in its absence.

CEA's and CBO's Estimates of the Recovery Act. To estimate the effect of the Recovery Act on GDP, CEA applied a different fiscal multiplier to each component, and then aggregated the effects of each component to arrive at the overall GDP effect. For government spending (corresponding to public investment outlays and income and support payments) and for tax cuts, CEA used multipliers derived from the empirical estimates of the spending and tax multipliers discussed in Appendix 2. For other components of the Act, such as State and local fiscal relief, CEA used a multiplier equal to a weighted average of one or both of the tax and spending multipliers.¹³

The CBO used a similar approach in its quarterly reports on the effects of the Recovery Act (although their estimates include the impact of AMT relief and so are not completely comparable to CEA estimates).¹⁴ Because of the range of estimates of multipliers in the economic literature, CBO provided an upper and a lower bound for the fiscal multipliers on the various components of the Act. As shown in Table 3-5, CEA multipliers are within the range suggested by CBO (2013a).

 

Table 3-5

 

Estimated Output Multipliers for Different Types of Fiscal Support

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 108.]

 

 

 

Using the historical relation between increases in GDP and employment, CEA and CBO also estimated the number of jobs generated by the Recovery Act. According to CEA model, the Recovery Act increased employment by more than 2.3 million in 2010 alone, and continued to have substantial effects into 2012, as demonstrated in Figure 3-6. Cumulating these gains through the end of FY 2013, the Recovery Act is estimated to have generated about 6.4 million job-years. These estimates are also within the range of CBO's upper-and lower-range estimates of 1.6 to 8.3 million job-years.

CEA Estimates of the Recovery Act and Subsequent Fiscal Measures Combined. The combined effect of the Recovery Act and the subsequent countercyclical fiscal legislation is substantially larger and longer lasting than the effect of the Recovery Act alone.¹⁵ The Recovery Act represents only about half of total fiscal support for the economy from the beginning of 2009 through the fourth quarter of 2012. Moreover, as shown in Figures 3-7 and 3-8, the bulk of the effects of the other fiscal measures occurred as the Recovery Act was phasing down. These other measures thus served to sustain the recovery as effects of the Recovery Act waned. CEA multiplier model indicates that by themselves these additional measures increased the level of GDP by between 1.0 and 1.5 percent each quarter from mid-2011 through the end of calendar year 2012. Altogether, summing up the effects for all quarters through the end of calendar year 2012, the Recovery Act and subsequent fiscal measures raised GDP by an average of more than 2.4 percent of GDP annually -- totaling a cumulative amount equal to about 9.5 percent of fourth quarter 2008 GDP.

 

Figure 3-5

 

Estimates of the Effects of the Recovery Act

 

on the Level of GDP, 2009-2013

 

 

 

Figure 3-6

 

Estimates of the Effects of the Recovery

 

Act on Employment, 2009-2013

 

 

 

Figure 3-7

 

Quarterly Effect of the Recovery Act and Subsequent

 

Fiscal Measures on GDP, 2009-2012

 

 

 

 

Figure 3-8

 

Quarterly Effect of the Recovery Act and Subsequent

 

Fiscal Measures on Employment, 2009-2012

 

 

 

The contribution of all fiscal measures to employment is equally substantial. Other fiscal measures beyond the Recovery Act are estimated to have raised employment by 2.8 million job-years, cumulatively, through the end of calendar year 2012. Adding these jobs to those created or saved by the Recovery Act, the combined countercyclical fiscal measures created or saved more than 2.3 million jobs a year through the end of 2012 -- or 8.8 million job-years in total over the entire period.

Estimates from Private Forecasters. Private forecasters and domestic and international institutions have used large-scale macroeconomic models, mostly to estimate the effects of either the Recovery Act by itself or other policies in isolation. The models used by these individuals and organizations generally employ a similar multiplier-type analysis as is found in CEA and CBO work, although they vary considerably in their structure and underlying assumptions. Although no outside estimates of the total impact of all the fiscal measures are available, Table 3-6 displays the estimates of the impact of the Recovery Act offered by several leading private-sector forecasters before the Act was fully implemented. Despite the differences in the models, these private-sector forecasters all estimated that the Recovery Act would raise GDP substantially from 2009 to 2011, including a boost to GDP of between 2.0 and 3.4 percent in 2010.

Taking a broader view that incorporates fiscal measures in addition to the Recovery Act, Blinder and Zandi (2010) estimate the effect of the fiscal policies enacted through 2009 (the Economic Stimulus Act, the Recovery Act, cash for clunkers, the unemployment insurance benefits extensions of 2009). They find that these policies raised the level of GDP in 2009 by 3.4 percent in the third quarter and by 4.3 percent in the fourth quarter.

Cross-State Evidence

A different approach to estimating the effect of fiscal policy is to use variation in spending across states. As noted earlier, estimates of the effects of macroeconomic policy are inherently difficult because the counterfactual outcome is not observed. One way economists have attempted to address this difficulty is to isolate a component of the Act that was implemented in a random or quasi-random manner across different states, mimicking a randomized controlled trial used for research in other disciplines like medicine. If some states received more Recovery Act funds than others for reasons unrelated to their economic needs, then this portion of the funds can allow for an independent, unbiased evaluation of the effects, much like two groups of individuals participating in a drug trial that receive different dosages of the same medicine.

 

Table 3-6

 

Estimates of the Effects of the Recovery Act on the Level of GDP

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 112.]

 

 

 

One portion of the Act that was distributed independently of states' immediate economic needs was the additional grants to states for Medicaid. Under the Act, states received a 6.2 percentage point increase in their expected Federal reimbursement rate (the Federal Medical Assistance Percentage).¹⁷ This increase was worth more to states that spent more per capita on Medicaid before the recession (in FY 2007). To the extent that the relative severity of the recession at the state level was unrelated to its previous level of per-capita Medicaid spending, this portion of funds might be thought of "as if" randomly assigned. In other words, the spending was effectively independent of the strength or weakness of the state-level economy once the recession hit. As Figure 3-9 shows, states that received more funds stemming from this non-cyclical part of the formula tended to exhibit greater employment gains through the first half of 2009 compared with states receiving less funds.

 

Figure 3-9

 

Change in Nonfarm Employment

 

 

 

Source: Centers for Medicare and Medicaid Services, Data Compendium; Bureau of Labor Statistics, Current Employment Statistics; CEA Calculations.

Refining this approach, Chodorow-Reich et al. (2012) found that each additional $100,000 of formula-based Medicaid grants generated an additional 3.8 job-years, which translates into a $26,000 cost per job. Other academic papers following a similar approach, but assessing broader measures of Recovery Act spending, reached similar conclusions. For example, Wilson (2012) estimates a cost per job of about $125,000 for all Recovery Act spending programs other than those implemented by the Department of Labor (mostly unemployment insurance). Feyrer and Sacerdote (2011) and Conley and Dupor (2013) also find positive effects of the Act on employment, although the ranges of effects estimated in both papers include magnitudes similar to those discussed above as well as somewhat smaller effects.

International Comparison

The 2008 crisis reverberated worldwide. In addition to seeing sharp reductions in output and employment, many countries also experienced large government budget deficits because of countercyclical fiscal policies and a fall in tax revenues caused by the recession. These changes in budget deficits across countries can be used to derive an international estimate of the impact of fiscal policy. The International Monetary Fund's (IMF) early estimates using pre-crisis cross-country data suggested expenditure multipliers averaging 0.5, although with substantial variation across countries.¹⁸ However, subsequent research by IMF (2012) and Blanchard and Leigh (2013) reassessed this earlier work and estimated multipliers substantially above 1.0 during the crisis, consistent with the discussion earlier in this chapter about recent fiscal multipliers in the United States.

This international evidence also suggests that the structural reductions in government budget deficits (or "fiscal consolidation") implemented by many countries has had a large negative impact on economic activity in the short run, at least when interest rates are low or at the zero lower bound and when there is already substantial economic slack. Previous research, summarized in Alesina and Ardagna (2010), hypothesized that fiscal consolidation can sometimes boost GDP because it increases investors' confidence and lowers interest rates. But Blanchard and Leigh's (2013) results, as well as findings by Perotti (2011) and Guajardo, Leigh, and Pescatori (forthcoming), point instead to significant short-run costs of deficit reductions and suggest a more gradual strategy of fiscal consolidation, as explained for instance in Blanchard, Dell'Ariccia, and Mauro (2010).

It is notable that the United States is one of only two of the 12 countries that experienced systemic financial crises in 2007 and 2008 but have seen real GDP per working-age person return to pre-crisis levels (see Box 3-2). Although this does not provide any specific evidence on the effect of U.S. fiscal measures, it is consistent with the proposition that the full set of U.S. policy interventions made a sizable difference in reversing the downward spiral of falling employment and output.

Benchmarking the Economy's Performance Since 2009

While the bulk of the available evidence indicates that the Recovery Act and subsequent fiscal legislation helped avert what might have become a second Great Depression and paved the way for stronger economic growth, many households continue to struggle with the after-effects of the recession. In addition, from a macroeconomic perspective, the average rate of real GDP growth in this recovery (2.4 percent a year) has been slower than many would have liked. Some critics have argued that this slower growth is evidence that economic policymaking has gone awry, and that the interventions undertaken since 2008 have had unintended detrimental consequences on growth. Taylor (forthcoming) argues that fiscal policy not only failed to help but actually hurt the economy.

As discussed earlier, it is impossible to infer the causal impact of a set of policies from the observed outcomes because these observed outcomes do not reveal what would have happened absent the policy interventions. The research that attempts to answer that counterfactual conclusion using a variety of different methods has generally come to the conclusion that the Recovery Act and subsequent measures had a large positive impact on growth and employment.

In particular, claims based on the recovery are often based in part on a misleading apples-to-oranges comparison to past growth and also fail to take into account the key features of the recession and recovery. First, the economy's potential growth rate is slower now than it has been in previous post-World War II recoveries for long-standing reasons unrelated to the Great Recession or the policies that followed in its wake. This lower rate of potential growth reflects several key factors: slowing growth in the working-age population as baby boomers move into retirement, a plateauing of female labor force participation following several decades of transformative increases, and a slowdown in productivity growth. CBO (2012a) estimates that slower growth of potential GDP accounts for about two-thirds of the difference between observed real GDP growth in the current recovery and growth on average in the preceding postwar recoveries, an estimate that is in line with other recent studies (see the 2013 Economic Report of the President for further discussion).

Second, the economy has encountered a long list of additional headwinds in recent years. This list includes international events like the European sovereign debt crisis, the tsunami and nuclear accident in Japan, and the disruption of Libya's oil supply. It includes extreme weather like Hurricane Sandy and the 2012 drought that was described by the U.S. Department of Agriculture as the "most severe and extensive drought in at least 25 years."¹⁹ It includes fiscal austerity at the state and local level that intensified as the Recovery Act began to phase out and has cost hundreds of thousands of additional job losses even during the expansion period. And it includes measures like the sequester which CBO estimated took 0.6 percentage point off growth in 2013, the 16-day government shutdown that the Bureau of Economic Analysis (BEA) estimated directly subtracted 0.3 percentage point from growth in the fourth quarter, and dangerous brinksmanship around the debt limit in 2011 and 2013.

In addition, the unique after-effects of financial crises discussed in Box 3-2 have also created substantial challenges for faster growth. The Great Recession was the first downturn brought about by a systemic financial crisis in nearly 80 years. Macroeconomic data from the Great Depression is limited, and models based on more readily available post-World War II data do not contain any comparable benchmark for the shock that hit the economy in 2008. Many of these models are still being refined to include more extensive and detailed linkages between the macroeconomy and the financial sector, based on lessons learned as a result of the crisis.

All these factors must be taken into account in assessing the economy's performance in recent years -- and understanding what would have happened without the significant policy actions that are described in this chapter.

Real GDP Per Working Age Population

 

in 2007-2008 Banking Crisis Countries, 2007-2013

 

 

 

 

Table 3-7

 

Tax Relief and Income Support in the Recovery Act

 

and Subsequent Measures, 2009 2012

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 118.]

 

 

 

The Recovery Act's income support and individual tax cut provisions allowed households to maintain their purchasing power through one of the worst recessions of the century. The Making Work Pay tax credit in effect in 2009 and 2010 provided 95 percent of workers with a tax cut worth $400 for a typical single worker and $800 for a typical married couple. Without these provisions, aggregate real disposable personal income would have been $354 billion lower than what it actually was in 2009. As shown in Figure 3-10, despite $300 billion in lost private wages and salaries, real disposable incomes actually grew throughout calendar year 2009 (CEA 2010b), primarily due to tax cuts for families, the largest of which was the Making Work Pay tax credit in the Recovery Act. The Making Work Pay credit was replaced by the even-larger payroll tax cut in 2011 and 2012 that provided a tax cut for all 160 million workers, with $1,000 for the typical worker making $50,000 per year.

 

Figure 3-10

 

Disposable Personal Income With and Without ARRA

 

 

 

Sources: Bureau of Economic Analysis, Effect of the ARRA on Selected Federal Government Sector Transactions, National Income and Product Accounts.

Unemployment Insurance

Regular state-based unemployment insurance (UI) programs typically provide benefits for 26 weeks, but as the average duration of unemployment rose to record highs in the 2007-09 recession and its wake, additional steps were needed. The Recovery Act expanded unemployment insurance in several ways. First, the Act provided for a 100 percent Federal contribution to the Extended Benefits program, which has been in place since 1970 to assist states that experience especially sharp increases in unemployment but has traditionally been jointly financed by Federal and State governments. The Recovery Act also extended the Emergency Unemployment Compensation (EUC) program enacted in 2008, which extended the duration of benefits available under Extended Benefits. It also provided an additional $25 a week in benefits through the end of 2009 and offered incentive funds for States that chose to modernize their unemployment insurance systems. Subsequent to the Recovery Act, Congress passed several more extensions and expansions of unemployment insurance, and another round of reforms aimed at assisting people searching for work.

Effects of Unemployment Insurance on Workers. In total, 24 million U.S. workers have received extended unemployment insurance benefits. Counting workers' families, over 70 million people have been supported by extended UI benefits, including more than 17 million children. Benefits have helped a broad swath of individuals, including 4.8 million with a bachelor's degree or higher. The impact was profound: the Census Bureau estimates that from 2008 to 2012, unemployment insurance kept over 11 million people out of poverty.

Beyond providing income support and keeping families out of poverty, unemployment benefits also affect labor markets. As discussed in the Executive Office of the President report on unemployment insurance (Council of Economic Advisers and Department of Labor, 2014), elevated unemployment rates in recent years were driven by declines in the demand for labor, with only slight reductions in labor supply stemming from unemployment insurance extensions. Moreover, as shown by Chetty (2008), unemployment benefits can also have a positive effect on labor productivity, because they give people time to search for a job better suited to their skills.

In addition to supporting incomes, unemployment benefits deter the long-term unemployed from dropping out of the labor force. After the extension of the unemployment benefits program in 2008, the long-term unemployed dropped out of the labor force at a considerably reduced rate, and Rothstein (2011) suggests that most of the small increase in unemployment rates due to extended benefits can be attributed to this phenomenon. While job-finding rates for the long-term unemployed remain low, keeping people in the labor market increases the chance that they will eventually resume working, which supports the economy's long-run potential.

Unemployment Insurance Reforms. The Recovery Act also included the most significant reforms to unemployment insurance in decades through a $7 billion fund to incentivize states to modernize their UI systems and to update eligibility rules to reflect the changing labor market. States received an incentive payment if they implemented some suggested improvements to their eligibility rules. These suggested improvements included allowing a worker to become eligible based on his or her most recent earnings (rather than earnings in the previous calendar year) or when quitting a job because of certain circumstances (compelling family responsibilities, a relocating spouse, domestic violence, or sexual assault). Proposed reforms also included offering benefits to individuals seeking only part-time work and providing for a dependent allowance.

Overall, states invested $3.5 billion of Recovery Act funds toward these modernization efforts. The law prompted 41 states to make nearly 100 reforms to their unemployment insurance programs. Numerous states expanded eligibility to workers whose job loss was due to compelling family circumstances, with 13 states adding coverage for domestic violence, 14 states adding coverage to care for a sick family member, and 16 states extending coverage to a relocating spouse.

In February 2012, the President signed into law more reforms to unemployment insurance -- many of which were originally proposed in the American Jobs Act -- including measures to help the long-term unemployed get back to work. Specifically, the new law created opportunities for states to test new strategies to help the long-term unemployed find new work. The Administration also expanded "work-sharing" programs across the country, which will help prevent layoffs by encouraging struggling employers to reduce hours for workers rather than cut headcount. Additionally, for the first time, the reforms allowed the long-term unemployed who were receiving federal benefits to start their own businesses, while also providing support to states to expand entrepreneurship programs.

Protecting the Most Vulnerable. The Recovery Act and subsequent legislation also included a range of proposals focused on protecting the most vulnerable. These measures included expanding the Earned Income Tax Credit (EITC) and the refundable portion of the child tax credit, both of which provide an additional reward to work for low-income families. The Administration also sought to ensure that the Making Work Pay credit was refundable, so that it benefited not just middle-class families but moderate-income working families as well. The Recovery Act expanded the Supplemental Nutrition Assistance Program (SNAP) to help families through tough times while also providing emergency benefits through Temporary Assistance to Needy Families (TANF), including subsidies to encourage hiring of low-income parents. The Recovery Act also ended or prevented homelessness for over 1.3 million families through the Homelessness Prevention and Rapid Rehousing Program.

All told, the pre-existing social insurance system combined with the expansions in the Recovery Act and subsequent extensions were very effective in preventing a large rise in the poverty rate, despite a substantial downturn in the economy. Even though the economy was dealt its most severe blow since the Great Depression, Wimer et al. (2013) find that from 2007 to 2010, the poverty rate measured to include the effects of antipoverty policy measures rose just half a percentage point. Excluding these measures, the poverty rate would have risen 4.5 percentage points -- nine times greater than the actual increase. Chapter 6 further discusses the effects of the Administration's policies on reducing poverty.

For this reason, the Recovery Act was designed not just to provide an immediate, short-term boost to the economy, but also to make investments that would enhance the economy's productivity and overall capacity even after the direct spending authorized by the Act had phased out. The Act's investments in expanding broadband infrastructure and laying the ground-work for high-speed rail, to take two examples, are a far cry from the broken window in Bastiat's parable because they do so much more than simply restore things to how they once were. Rather, these types of investments will raise the economy's potential output for years to come, from a rural school that can now offer its students and teachers high-speed Internet access, to a business that has a new option to transport its goods more quickly.

As shown in Table 3-8, the Recovery Act included $300 billion of these types of investments in areas such as clean energy, health information technology, roads, and worker skills and training. Figure 3-11 suggests that the timing of these investments was relatively more spread out than some of the Act's other measures, consistent with the longer-term focus of these projects.

 

Table 3-8

 

Recovery Act Long Term Growth Investment by Category

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 123.]

 

 

 

The Recovery Act and subsequent jobs measures were designed to expand both private capital and public capital.

Business Tax Incentives for Private Capital. The theory behind incentivizing private capital holds that, at a time of systemic financial crisis, firms might not have access to sufficient capital through financial markets to invest or might be overly deterred from investing due to uncertainty, as explained in a report by the Treasury's Office of Tax Policy (2010a). To overcome these impediments to private investment, the Recovery Act and subsequent measures included business tax cuts designed to increase cash flows -- like extended periods for net operating loss carrybacks and bonus depreciation -- that, in effect, constituted an interest-free loan to businesses. Some economic research (House and Shapiro 2008) has shown that bonus depreciation policies can noticeably raise investment. At low interest rates, these measures had a small net present value cost to the Federal Government, but provided resources to credit-constrained firms to support investment. Building on this approach, in fall 2010 the President proposed 100 percent expensing for business investment, which, as passed by Congress in December 2010, became the largest temporary business investment tax incentive in history.

 

Figure 3-11

 

Recovery Act Cumulative Public Investment Outlays, 2009-2013

 

 

 

Transportation and Other Investments in Public Capital. A modern and effective transportation infrastructure network is both necessary for the economy to function and a prerequisite for future growth. Numerous studies have found evidence of large private-sector productivity gains from public infrastructure investments, as highlighted in a report by the Department of Treasury and CEA (2012).²⁰ The early stage of the recovery has been a particularly opportune time to undertake such investments because of the high level of underutilized resources in the economy and low construction costs. The Treasury report also points out that transportation investments can create middle-class jobs and lower transportation costs, which would otherwise weigh on household budgets.

The Recovery Act allocated $48 billion to programs administered by the Department of Transportation, with almost 60 percent for highways and 37 percent for public transportation and intercity passenger rail. The magnitude of this aid was substantial. While it is difficult to estimate what transportation expenditures would have been without the Recovery Act, total highway spending in 2010 was about $27 billion (or 24 percent) higher than in 2007. This increase occurred in a period when user revenues (such as fuel taxes and other fees), the usual source of funding for states for transportation projects, were declining. Moreover, an equal dollar amount of expenditures was more effective during the recession, because construction costs for highways (as measured by the National Highway Construction Cost Index) declined about 20 percent between mid-2008 and mid-2009, and remained relatively flat through 2011.²¹

In addition to these direct programs, the Recovery Act also provided indirect support for transportation projects through Build America Bonds. The Federal Highway Administration estimated that 26 percent of the total funds raised by Build America Bonds (or $48 billion) were used by states for transportation projects. Further, a Recovery Act provision that temporarily exempted Private Activity Bonds (PABs) from the Alternative Minimum Tax (AMT) enabled airports across the country to access credit at affordable rates. The Federal Aviation Administration estimated that 24 U.S. airports issued $12.7 billion in bonds under the Recovery Act AMT exemption, realizing $1.06 billion of present value savings ($1.8 billion in gross savings) through early November 2010.

With these funds, shovels went in on more than 15,000 transportation projects across the Nation. The Department of Transportation estimates that these projects will improve nearly 42,000 miles of road, mend or replace over 2,700 bridges, and provide funds for over 12,220 transit vehicles. The Recovery Act also made the largest-ever investments in American high-speed rail, constructing or improving approximately 6,000 miles of high-performance passenger rail corridors and procurement of 120 next-generation rail cars or locomotives.

Finally, the Recovery Act initiated the Transportation Investment Generating Economic Recovery (TIGER) grant program, which allowed the Department of Transportation to invest in critical projects that were difficult to fund through traditional means. The TIGER program included a competitive process that encouraged innovation and regional collaboration. The program made extensive use of benefit-cost analysis to evaluate project applications, and required grant recipients to track the performance of their projects once launched to ensure that they achieve the promised benefits. The program also allowed many cities, counties, and other government entities to access direct Federal funds for the first time. The initial $1.5 billion TIGER program was deemed so successful that it was extended five additional times and is currently in effect through September 2014.

The Recovery Act also invested in restoring or otherwise improving infrastructure to allow Americans to safely and easily access public lands and waters. Investments included about $1 billion to the National Park Service, Fish and Wildlife Service, and Forest Service for deferred maintenance of facilities and trails and for other critical repair and rehabilitation projects. These projects help support the infrastructure needed to sustain the outdoor recreation economy and contribute to the enjoyment of public lands.

The Recovery Act included funding for programs administered by the Environmental Protection Agency (EPA) to protect and promote both a healthier environment and jobs. These investments have generated substantial environmental benefits, such as cleaning up contaminated land and putting that land back to economic use, reducing air pollution from diesel engines, and reducing contaminants in both surface water and drinking water. EPA's Brownfields program used $100 million in Recovery Act funds to leverage additional funds and cleaned up 1,566 acres of properties that are now ready for reuse, far exceeding the original target of 500 acres. The Act's funding led to 30,900 old diesel engines being retrofitted, replaced, or retired, which has reduced lifetime emissions of carbon dioxide by 840,300 tons and particulate matter by 3,900 tons.²² More than 3,000 water quality infrastructure projects and Clean Water projects are improving or maintaining sewage treatment infrastructure for over 78 million people nationwide, as another Act investment. The Recovery Act funds have also enabled 693 drinking water systems, serving over 48 million Americans, to return to compliance with Safe Drinking Water Act standards.²³

Protecting and Expanding Investments in Human Capital

The Recovery Act was also aimed at protecting and expanding human capital. Saving and creating jobs helps protect human capital, in part, by preventing the loss of skills -- including job search skills -- that can come from prolonged periods of unemployment. The evidence shows that protracted unemployment in Europe in the 1980s and 1990s resulted in sustained loss of human capital (Blanchard and Summers 1986, Ljungqvist and Sargent 1998). Helping workers better connect with jobs, whether through unemployment insurance reforms or job subsidies in the TANF emergency fund, has helped protect human capital.

Significant investments and reforms in education were critical to actually expanding human capital. State and local governments typically provide more than 90 percent of the funding for elementary and secondary education and about 40 percent of the funding for public institutions of higher education in the United States. As the economy slowed in 2008, State revenues declined, putting pressure on education budgets.

The Recovery Act dramatically increased funding for education through Title I grants to local education agencies (LEAs), School Improvement Grants, and grants for special education. In addition, the Act increased student aid and support for post-secondary institutions to invest in new buildings and research in innovative health and energy technologies. In response to these grants, recipients reported that more than 800,000 education job-years were saved or created, keeping teachers, principals, librarians, and counselors as well as university faculty and staff on the job.

States also reported that they used State Fiscal Stabilization Funds from the Recovery Act to restore sizable shares of K-12 education funding. For example, the Recovery Act restored 9 percent of K-12 education funding in California, Indiana, Alabama and Oregon; 12 percent of such funding in Florida, Wisconsin and South Carolina; and 23 percent of K-12 education funding in Illinois in fiscal year 2009. In at least 31 states, Recovery Act funds prevented or lessened tuition increases at public universities, including universities in Massachusetts, Minnesota, and Virginia. Without this influx of State Fiscal Stabilization Funds, these states would have endured drastic cuts in education funding.

The Recovery Act launched the innovative Race to the Top Program with $4.35 billion. Race to the Top is a competitive grant program designed to encourage and reward States to implement critical reforms designed to help close the achievement gap and improve student outcomes, including better student assessments; better data systems to provide teachers and parents with information about student progress; new steps to develop and support effective teachers; and efforts to turn around low-achieving schools. Encouraged by the incentives included in Race to the Top, states across the country chose to adopt more rigorous academic standards aligned to higher expectations for college and career readiness. To date, 19 states, representing 45 percent of all K-12 students, have received Race to the Top funds; and to compete for funds, 34 states modified state education laws and policies in ways known to improve education.

The Recovery Act also expanded the Pell Grant program, raising the maximum grant from $4,731 to $5,550, and it created the American Opportunity Tax Credit to modify and replace the Hope higher education credit (this policy was later extended by the Tax Relief, Unemployment Insurance Reauthorization and Job Creation Act of 2010 and the American Taxpayer Relief Act of 2012). The passage of the Health Care and Education Reconciliation Act of 2010 enabled further expansion of the Pell Grant award. Together, these efforts to expand higher education opportunity helped individuals who chose to return to school or remain in school to bolster their skills in a demanding job market. As a result, the Pell Grant program offered $36.5 billion in aid to more than 8.8 million undergraduate students in FY 2010, compared to roughly half as much aid, $18.3 billion, for 6.2 million students in FY 2008 (U.S. Department of Education, 2011). The largest growth in Pell Grant applications came from students in the lowest-income categories. For tax year 2009, 8.3 million tax returns claimed $14.4 billion in American Opportunity Tax Credits. This level of education credits (including the lifetime learning credit) was a nearly $10 billion increase from the prior year (U.S. Department of the Treasury, 2010b).

Investments in Technology and Innovation

Some of the highest returns to investment are in the area of innovation. Often, innovation produces large returns for the economy as a whole, but since businesses do not capture all these society-wide returns, they tend to underinvest in innovation. For example, firms may not undertake research and development even though it would benefit the rest of their industry, other industries, and their regional economies. In general, looking across a range of industries, economists have estimated that the divergence between private and social returns to investment may be as great as two-to-one (for instance, Hall et al. 2009). This is especially true when investments can result in externalities that are not captured by the entity making the investment. For instance, in the energy sector, there are substantial climate and national security benefits to cleaner energy that are not fully internalized in the form of financial rewards for individual firms. The Recovery Act made a significant impact on innovation -- complementing the other measures this Administration has taken to encourage innovation.

Scientific Research. The Recovery Act provided a one-time supplemental appropriation of over $3 billion for the National Science Foundation and $1 billion for the National Aeronautics and Space Administration. It also increased support for the National Institute of Standards and Technology and provided the Advanced Research Projects Agency-Energy (ARPA-E) with funding of $400 million. The ARPA-E is charged with researching transformative energy technologies. So far it has pioneered research into so-called second-generation biofuels, which utilize agricultural and municipal waste, as well as more efficient batteries, superconducting wires, and vehicles powered by natural gas.

Clean energy. Clean energy was the focus of more than $90 billion in government investment and tax incentives in the Recovery Act. The purpose of these investments was to help create new jobs, reduce dependence on foreign oil, enhance national security, and improve the environment by countering climate change. Key targets included energy efficiency (with programs such as the weatherization assistance program), renewable generation (with investments in wind turbines, solar panels, and other renewable energy sources), and grid modernization. Many of these clean energy programs were administered though the $38 billion Recovery Act portfolio of the Department of Energy.

 

Figure 3-12

 

Advanced Renewable Electric Power Net Generation, 2000-2012

 

 

 

Using the multiplier model described earlier in the chapter, CEA estimated that clean energy investment created or saved about 650,000 job-years, directly or indirectly, through 2012.²⁴ These investments have started to drive changes in energy production, as highlighted, for instance, by Aldy (2013). Owing in large part to these clean energy incentives and investments, renewable wind, solar, and geothermal energy have increased their contributions to U.S. energy supply each year since 2008. For instance, as shown in Figure 3-12, wind electricity net generation nationwide grew by 145 percent from 2008 to 2012. Solar thermal and photovoltaic electricity net generation more than quadrupled during the same period. Meanwhile, carbon dioxide emissions from the electric power sector fell approximately 14 percent over the period, even though total power generation declined by only about 2 percent.

Many of the clean energy provisions of the Recovery Act were designed to bring in private funds through coinvestment. For example, through Energy Cash Assistance, individuals and businesses that installed certain types of renewable energy generation received a grant equal to 30 percent of the project's cost.

Of course, not every investment in clean energy will ultimately result in a transformative technology. Because funding is often directed to projects based on ideas that are at the frontier of scientific research, there is a certain degree of risk involved. But given the grave economic, environmental, and national security consequences of climate change, these types of investments must continue. An independent review released in 2012 found that, on the whole, the Department of Energy loan guarantee programs are expected to perform well and hold even less risk than initially envisioned by Congress.

Health Care Information Systems. The Health Information Technology for Economic and Clinical Health (HITECH) Act, enacted as a part of the Recovery Act, encouraged adoption and use of health information technology. The core of the HITECH Act is a set of financial incentives to health care providers to adopt and make "meaningful use" of electronic health records. The HITECH Act also provided $2 billion to the Department of Health and Human Services to fund activities to encourage the diffusion of health information technology, such as investing in infrastructure and disseminating best practices. The Act also made a variety of other changes, including provisions to facilitate data sharing across health care providers to support coordinated care and protect patient privacy.

Fully integrated electronic health record systems allow immediate and complete access to all relevant patient information. These innovations have the potential to greatly improve coordination of care -- for example, by limiting the unnecessary duplication of tests and procedures -- and also to reduce medical errors, thereby lowering health care costs. Chapter 5 further explains the benefits of fully integrated electronic health record (EHR)systems and discusses the dramatic increase in the share of medical providers using electronic health records in recent years.

Broadband. The Recovery Act helped increase access to broadband and drive its adoption across the country, both directly through grants, and indirectly through tax incentives such as increased expensing of investment costs.²⁵ It provided $4.4 billion through the Department of Commerce's National Telecommunications and Information Administration to deploy broadband infrastructure (for instance, laying new fiber-optic cables or upgrading wireless towers, and connecting key institutions such as schools, libraries, hospitals, and public safety facilities) and support public computer centers (establishing new public computer facilities to provide broadband access to the general public or specific vulnerable and underserved populations). The funding also encouraged sustainable adoption of broadband (for instance, through digital literacy training and outreach campaigns), led to the publication of the National Broadband Map (www.broadbandmap.gov), and supported state broadband leadership and capacity building activities (through, for example, local broadband planning teams and information technology assistance provided to small businesses, schools, libraries, and local governments). The Recovery Act also provided $2.5 billion through the Department of Agriculture Rural Utilities Service to expand broadband access in rural areas.

Because of these grants, over 110,000 miles of broadband infrastructure have been added or improved, and high-speed connection has been made available to about 20,000 community institutions. These projects have also delivered about 16 million hours of technology training to more than 4 million users.

In part as a result of the Recovery Act and related policies, broadband access has risen substantially in recent years. Chapter 5 discusses broadband development in more depth.

Fiscal Sustainability and the Recovery Act

The Recovery Act and subsequent fiscal measures were part of an overall fiscally responsible economic strategy that cut the deficit in the medium and long run. Moreover, given the economic context in which the jobs measures were passed, these measures alone had little if any impact on long-run fiscal sustainability.

The Recovery Act is entirely temporary -- it cost $763 billion over the first decade (not counting the extension of the AMT patch) and it has no long-term impact on non-interest outlays or revenues. Overall, assuming the CBO score, the Act at most added less than 0.1 percentage point to the 75-year fiscal gap.

These estimates are small but may nevertheless overstate the true cost of the Recovery Act. To the degree that the Act successfully expanded output and boosted employment, those gains would result in additional revenue and less spending on countercyclical programs than would otherwise have occurred. Taking the estimates presented in this chapter for the increase in GDP over the 2009-12 period -- and assuming these increases led to additional tax revenue at 18 percent of GDP, roughly the recent historical average -- then the resulting increase in revenue would alone be enough to offset roughly one-quarter of the Act's cost.

Moreover, to the degree that effects on output are persistent -- a factor that is not captured in the estimates in this chapter but is assumed by the IMF (2009) and Reifschneider, Wascher, and Wilcox (2013), then the positive fiscal feedback effects could be even larger. DeLong and Summers (2012) have shown, for example, that with plausible multipliers and persistence in output effects, it is possible that the additional output associated with the Recovery Act, and associated additions to revenue and reductions to debt, could result in a reduced debt-to-GDP ratio by the end of the decade.

These estimates do not reflect the potential benefits for long-term growth of the productivity-enhancing investments in the Recovery Act. For example, if an infrastructure project has a total rate of return of 10 percent, and if overall revenues are about 18 percent of GDP, then it would have a rate of return to the Federal taxpayer of about 2 percent. Given the Federal borrowing costs at the time of the Recovery Act, the investment would conceivably pay for itself over time and reduce Federal debt as a share of GDP as the investment produces returns.

None of these estimates should be taken as conclusive or as a suggestion that official budget scoring should take these feedback effects into account. When the economy is operating at full employment, and monetary policy is not constrained by the zero lower bound, many of these macroeconomic feedback effects would be less relevant or not even operative at all. Moreover, if fiscal policy actions raised the specter of substantially larger and less sustainable future deficits and debt, that could reduce confidence and raise interest rates, undermining any beneficial economic feedback. But in this case, these measures were passed at the same time that the Administration was also laying out steps for longer-term deficit reduction and reducing the fiscal gap by passing major deficit-reduction measures, including the Affordable Care Act and the Budget Control Act.

As a result, given the overall context of highly insufficient aggregate demand, monetary policy operating at the zero lower bound, and other measures for medium- and long-term deficit reduction, fiscal measures to support jobs have the potential for even larger impacts on output and thus greater associated revenue feedbacks and a much lower long-run fiscal cost, if they have any long-run fiscal cost at all.

The Administration's actions have been guided by the notion that fiscal support measures would only be needed for a temporary period, and this view is being borne out. Most temporary measures to support the economy expired in 2013, most notably the payroll tax cut. Businesses and households are now in far better shape as a result of several years of deleveraging, and private-sector growth has led the way since 2010. Although many challenges linger, and supportive measures like emergency unemployment insurance remain necessary given the unacceptably high rate of long-term unemployment, the economy has the potential for even stronger growth in 2014.

Public policy, in particular public investment in areas like research, infrastructure, and innovation, will continue to play an important role in the economy. The President is proposing additional investments and reforms in all of these areas. But, in these cases, investments are part of a longer-term, sustained commitment to expanding the productive capacity of the economy without the same need for immediate countercyclical support.

Overall, the Recovery Act and subsequent measures are one of the main reasons why the U.S. economy was able to return to record levels of per working-age population GDP within just over four years of the onset of the recession and to bring the unemployment rate down by 0.8 percentage point per year -- when many other countries with systemic financial crises have not seen their GDP per working-age population fully recover or their unemployment rates start a sustained fall. In the longer run, the benefits of all of these efforts will be more difficult to isolate from other simultaneous changes, but they will be no less profound in terms of their cumulative impact on the economic well-being of the Nation.

Table 3-10 reports the budgetary impact classified into the six broad functional categories shown also in Figure 3-1: individual tax cuts, Alternative Minimum Tax relief, business tax incentives, State fiscal relief, aid to directly impacted individuals, and public investments. The following sections of this appendix will discuss each of these categories in more detail.

Tax Relief

Within the first three categories of tax cuts, major programs included the Making Work Pay tax credit, which provided a 6.2 percent credit on earnings up to a maximum value of $400 for individuals and $800 for couples, phasing out starting at income above $75,000 and $150,000, respectively (estimated to cost about $116 billion between 2009 and 2011). The credit was administered through reducing tax withholdings and the Internal Revenue Service required that companies reduce withholding by April 1, 2009. In addition, the legislation made $250 one-time payments to seniors, veterans, and people with disabilities. The Recovery Act included the Making Work Pay tax credit for 2009 and 2010. In December of 2011 Congress enacted a 2-percentage point reduction of the Social Security payroll tax for 2011 that was extended through 2012 and expired at the start of 2013.

Additionally, the Recovery Act provided tax credits for families, such as an expansion of the child tax credit, including making it refundable for more low-income families (at a total estimated cost of $15 billion), expansions of the earned income tax credit for married couples and families with more than three children ($5 billion), and the American Opportunity Tax Credit to help make college more affordable. All of these measures have since been extended through 2017, and the President's Budget for 2014 proposes to make them permanent -- rendering them among the only items from the Recovery Act intended to be permanent.

The Recovery Act also raised the exemption amount for the AMT to $46,700 for individual taxpayers and $70,950 for joint filers, at an estimated cost of $70 billion. Because this was a widely expected continuation of previous AMT patches, this component of the Recovery Act did not represent a net new fiscal impetus for the economy and is not included in CEA's macroeconomic estimates.

For businesses, the legislation provided cost-effective incentives to expand investment by allowing businesses to immediately deduct half of the cost of their investments (bonus depreciation) and also to extend the period over which small firms (except those receiving TARP funds) could claim losses and expense capital purchases. Businesses buying back or exchanging their own debt at a discount were also allowed to defer any resulting income. All of these measures were designed to improve the cash flow for firms that might be facing credit constraints and to increase incentives to invest. Long-run costs to the Federal Government were limited because the measures largely advanced tax benefits that companies would receive anyway. The 50 percent bonus depreciation was subsequently extended and expanded to 100 percent expensing, and the net operating loss carryback was extended to larger firms. In addition, the Recovery Act included incentives for investments in renewables and advanced energy manufacturing, and in areas undergoing significant distress through State and local government-issued Recovery Zone Bonds. The Recovery Act also increased funding for the New Markets Tax Credit and provided incentives to hire unemployed veterans and disconnected youth.

 

Table 3-9

 

Recovery Act Outlays, Obligations, and Tax Reductions

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 135.]

 

 

 

 

Table 3-10

 

Recovery Act Fiscal Stimulus by Functional Category

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 136.]

 

 

 

An expansion in unemployment benefits offered significant aid to individuals.²⁶ Typically, American workers who have lost their jobs are entitled to 26 weeks of benefits under the unemployment insurance (UI) program, which tends to replace about half of lost earnings and is paid for entirely by the states through payroll taxes levied on employers. In June 2008, Congress created the Emergency Unemployment Compensation (EUC) program, which provided an additional 13 weeks of federally financed compensation in all states to eligible individuals who had exhausted their UI benefits. The Recovery Act extended and expanded the EUC program to reflect that fact that with jobs increasingly scarce the optimal balance of unemployment insurance shifted towards covering people for a longer period of time. It also provided 100 percent Federal funding of the pre-existing Extended Benefit (EB) program, which provides an additional 13 or 26 weeks of benefits in states where unemployment is exceptionally high and rising. (EB costs are usually borne half by the Federal Government and half by the States.)

The Recovery Act also added $25 a week to benefits and exempted the first $2,400 in yearly unemployment benefits from taxes. The CBO (2009a) estimated the total costs of these changes to the unemployment compensation system at $39 billion. In addition, the Federal government offered states incentives to modernize their UI programs, picking up the cost of new provisions allowing workers to become eligible based on recent earnings (rather than those from the previous calendar year) and extending benefits to part-time job seekers.

The Recovery Act provided a 13 percent increase in SNAP payments and lifted several restrictions governing the length of time that individuals could collect food stamps, at an estimated cost of $20 billion. For the first time, the Federal Government agreed to temporarily pay 65 percent of health insurance premiums for laid off workers who wanted to continue with their employer-sponsored health insurance. Other aid to individuals included funds for job training and improving skills of the hard to employ and young workers.

The Recovery Act devoted substantially more resources compared with previous antirecessionary policies to investments in education and research and development. The legislation increased the Pell Grant maximum by $500 to $5,550, at an estimated cost of $17 billion over ten years. The Recovery Act also boosted Title I aid and other programs for disadvantaged children ($13 billion) and funds for special education ($12 billion).

State Fiscal Relief

The Recovery Act provided unprecedented support for State and local governments, which often face budget challenges in a recession because their revenues rise and fall with the economy, while pressures on spending, especially on programs targeted to the disadvantaged, tend to move in the opposite direction. The result can be budget shortfalls, or gaps between expected revenues and expenditures. These gaps pose problems for state residents already affected by the downturn, and for the larger economy because most State and local governments are generally bound by constitutional or statutory requirements to balance their operating budgets each year. As shown by Poterba (1994), states and localities have to raise taxes or cut spending, precisely when doing so can most harm recovery.

To dampen such counterproductive tax increases or budget cuts, the Recovery Act boosted Federal Medicaid payments by $87 billion, including a 6.2 percent across-the-board increase in the Federal matching rate, plus delays of a planned reimbursement cut for some states (based on income growth before the recession) and an increment of aid linked to local unemployment conditions. It also established a $53.6 billion State Fiscal Stabilization Fund to be administered by the Department of Education, but with some funds available for other "high priority needs" such as public safety. Unlike most previous increases in Federal grants to States and localities during a recession, these transfers were available for general fiscal relief or left to local discretion to use, as long as recipients met basic maintenance of effort or minimal spending requirements.

Beyond direct spending, the Recovery Act made new types of borrowing available for State and local governments. Build America Bonds (BABs) allowed State and local governments to access non-traditional markets, including pension funds and international investors who would not normally purchase U.S. municipal bonds because they do not owe U.S. income taxes and therefore do not benefit from these bonds' tax-exempt status. Under BABs, State and local issuers could offer higher taxable interest rates on bonds and choose to make a Federal income tax credit available to buyers or to take a direct subsidy offsetting 35 percent of their borrowing costs. State and local governments issued $181 billion of Build America Bonds before the program expired at the end of 2010. The Treasury Department has estimated that this action saved issuers $20 billion in present value of borrowing costs as well as alleviating supply pressures in the tax-exempt market (Department of the Treasury 2011).

Investments

The Recovery Act made numerous investments in human capital, clean energy, health information technology, roads, and the skills of U.S. workers.²⁷ For example, the Recovery Act provided an additional $27.5 billion for highway construction, $18 billion for public transit and inter-city passenger rail, $10 billion for water infrastructure, and $18 billion for government facilities. It also made available $57 billion for investment in smarter grid technology, renewable energy, and energy efficiency improvements through a combination of grants, loans, and pilot programs, including $5 billion to help low-income households weatherize their homes. Scientific projects from the National Science Foundation, National Institutes of Health, NASA, the Department of Energy, and others received over $15 billion for scientific facilities, research, and instrumentation. Additionally, the Recovery Act provided $7 billion to expand broadband Internet access in underserved areas of the country. The Recovery Act also provided several investments in health care and health information technology, including an $18 billion measure to encourage hospitals and physicians to computerize medical records, $2 billion for Community Health Centers, $1 billion for fighting preventable chronic diseases, and $1 billion for researching the effectiveness of various medical treatments. In total, more than $100 billion of the investments -- including some tax incentives -- were explicitly targeted at innovation.²⁸

Subsequent Fiscal Measures

Table 3-11 shows the total fiscal support provided by the Administration, by fiscal year, with a brief description of the main programs for each measure. (These data were summarized in Table 3-4.) All measures use prospective CBO cost estimates. These totals only include measures explicitly designed to address job creation and provide relief and do not include routine extensions, like so-called "tax extenders" or the fix to Medicare's Sustainable Growth Rate formula.

 

Table 3-11

 

Fiscal Support for the Economy Enacted After the Recovery Act

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 140.]

 

 

 

Many modern macroeconomic models combine forward-looking behavior with some form of slow-moving prices or wages, sometimes called "New Keynesian" models. In normal times, when monetary policy is unconstrained and interest rates can vary, these models tend to imply fiscal expenditure multipliers that are positive but smaller than one, as shown for instance by Cogan et al. (2010) and Coenen et al. (2012), in part because of increases in the interest rate from monetary policy which partially offsets the fiscal expansion.

The onset of low interest rates has spurred considerable interest in how these models perform when monetary policy is constrained by the zero lower bound, that is, when the nominal federal funds rate falls to zero, as in the recent recession. For instance, Eggertson (2001), Christiano, Eichenbaum, and Rebelo (2011) and Woodford (2011) have shown that when nominal interest rates are near zero, government spending can be particularly effective and generate spending multipliers that are greater than one; at the zero lower bound, expansionary fiscal policies can increase inflation expectations and thereby reduce real interest rates, which spurs investment and consumption, and monetary policy does not counteract fiscal policy. Coenen et al. (2012) simulate the effect of the Recovery Act spending in some forward-looking models with rigidities, both conventional models (such as Smets and Wouters (2007)) and models augmented by the zero lower-bound effects. Their results show that the standard models imply a notable increase in output for several years, but with multipliers smaller than one, while the models augmented by zero-lower-bound effects imply multipliers that are much larger than one over the first few years.

The 2007-09 recession was unusual both because the Federal Reserve was at the zero lower bound and because of its severity. This severity raises the specter of high unemployment and -- because the path to recovery from a deep shock is long -- unusually long spells of unemployment. Long-term unemployment can lead to deterioration of skills and to stigmatization, which makes finding employment even more difficult. For these and other reasons, the longer the spell of unemployment, the less likely is an individual to find a job in any given month, and the more likely he or she is to remain unemployed or stop looking for a job altogether. This can lead to a vicious circle: persistent slack demand means many people out of work and long spells of unemployment, which in turn reduces the chances of the unemployed finding a job, which perpetuates slack and further lengthens spells. Because the resulting unemployment dynamics depend on the path of unemployment, not just on its current level, this phenomenon is often referred to as "hysteresis" in the rate of unemployment.

The potential for hysteresis in unemployment -- the economy getting stuck at high rates of unemployment for an extended period -- provides a further argument for activist fiscal policy, and models that build in hysteresis effects can have large and sustained multipliers (see for example Phelps 1972, Blanchard and Summers 1986, Ball 2009, and DeLong and Summers 2012). Reifschneider, Wascher, and Wilcox (2013) stress the relevance of these channels to the current recovery. Their research shows that the financial crisis damaged the productive capacity of the economy, by causing a steep decline in capital accumulation, lower productivity growth, and structural damages to the labor market, and a large portion of this damage to the productive capacity stemmed from weak demand. These results suggest that under such conditions fiscal policy can continue to have a meaningful effect on output with a substantial lag.

This recent work has moved far beyond the basic multiplier. It shows that fiscal and monetary policy can influence each other in substantial ways. While fiscal multipliers might be less than the basic model suggests in mild recessions and when monetary policy is unconstrained, they can be large when monetary policy is at the zero lower bound. In addition, fiscal expansion in a deep recession can have additional long-term benefits, and therefore high multipliers, by shortening spells of unemployment, minimizing the erosion of human capital, and increasing future productivity.

Time Series Evidence

Evaluations of fiscal effects using the structural models described above reflect the economic theory used to construct the models. The reliability of the resulting estimates therefore depends on the reliability of the underlying macroeconomic theory. A complementary approach to evaluating the effects of fiscal policy is instead to use models that rely less on economic theory and more on historical empirical evidence.

The main challenge to credibly implementing this data-driven approach is using just enough theory, or finding enough independent variation in the data, to estimate the causal effect of fiscal policy on the economy: simply noting that two variables move together does not establish causality. For example, if Congress passed countercyclical fiscal policy whenever a recession loomed, a figure plotting the countercyclical policy variable and GDP growth would show that countercyclical policy occurred at the beginning of recessions. An analyst might conclude, incorrectly, that this policy caused the recession, when in fact the policy was itself caused by the recognized onset of the recession. Analysis based on this hypothetical figure suffers from two central problems in the estimation of causal effects from observational (as opposed to experimental) data: simultaneous causality (the looming recession spurred Congressional action and the fiscal policy potentially affected the course of the economy) and the presence of other omitted, confounding factors (perhaps the Federal Reserve moved countercyclically and it was those actions, not Congress's, that muted the recession). The latter problem of omitted variables can be partially addressed using multiple regression methods, but the problem of simultaneous causality requires other approaches, and relying on simple plots or multiple regression can lead to misleading results.²⁹ Because such plots or regressions are uninformative, a vast literature developed over the past four decades uses more sophisticated methods to estimate causal effects in general and the effect of fiscal policy in particular.³⁰

Evaluation of the effects of fiscal policy in general, and the Recovery Act in particular, faces several additional challenges. First, the effect of activist fiscal policy must be disentangled from the automatic stabilizers built into the tax and safety net system. Second, the effect of fiscal policy unfolds over time, so there is not a single causal effect but rather a sequence of dynamic causal effects, including long-lasting effects of investment on productivity that could last for many years. Third, different fiscal policy instruments (expenditures, taxes, transfers) will in general have different effects. Fourth, as discussed above, theory suggests that the effect of fiscal policy could depend on the economic environment, and in particular could depend both on the severity of the recession and on the reaction of monetary policy.

A vast body of empirical literature now employs time-series data to estimate the macroeconomic effect of fiscal policy. Broadly speaking, this literature uses two different approaches to isolate (to "identify") the effect of fiscal policy. The first is to impose a minimal amount of structure on an otherwise unrestricted time series model, typically a so-called structural vector autoregression. In an influential contribution, Blanchard and Perotti (2002) assume that, because of implementation lags and limitations on the information considered by policymakers, fiscal policy does not respond immediately to other economic shocks. Under this assumption, any unpredicted movements in the fiscal variable (that is, movements that differ from what standard fiscal policy would have suggested) are unrelated to contemporaneous economic shocks, so the effect of fiscal policy can be estimated by tracing out the effect of those unpredicted movements on output and employment. Using this approach, Blanchard and Perotti (2002) estimate the government spending multiplier on GDP to be in the range 0.9 to 1.2. Ramey (2011b) reviews the large body of research that uses structural vector autoregressions to build on this approach to identifying the effects of fiscal shocks. The common theme of this work is using a component of fiscal policy -- in Blanchard-Perotti (2002), the unpredictable component -- which is "as-if random" in the sense that it is unrelated to other economic shocks.

A second approach to identifying the effect of fiscal policy is to exploit external information, such as institutional or historical knowledge, to find changes in fiscal policy that are in effect random (that is, independent of macroeconomic conditions), which can in turn be used to trace out the fiscal effect. Because this information falls outside the time series model being estimated, this approach is called the method of external instruments. In this vein, Ramey and Shapiro (1998) and Ramey (2011a) use expenditures on wars and military buildups, arguing that they are determined by international and political, not economic, considerations. These authors estimate GDP multipliers in the range of 0.6 and 1.2. Romer and Romer (2010), instead, use narrative evidence from Presidential and Congressional records and similar documents to identify tax changes that were not implemented in response to current or forecasted economic conditions. They find that the identified tax cuts have a sustained and large effect on output, with multipliers as high as 3. Mertens and Ravn (2012) use Romer and Romer's (2010) narrative to distinguish between the effects of anticipated and unanticipated tax changes and, surprisingly and in contrast to Ramey (2011a), find little difference in the two effects. Additional recent contributions include Favero and Giavazzi (2012) and Mertens and Ravn (2013). Estimates of fiscal policy effects obtained using this so-called method of external instruments are reviewed in Ramey (2011b) and Stock and Watson (2012).

The foregoing time series estimates are predicated upon fiscal multipliers having the same size in booms and in recessions. Recent work by Auerbach and Gorodnichenko (2012) suggests that while the spending multiplier can be relatively small during expansions, it can be much greater than one during recessions. These results are consistent with conventional models in recessions, but with neoclassical ones in booms, and suggest that multipliers obtained also using fiscal policy changes that happen in booms (such as the military buildup used by Ramey and Shapiro (1998) and Ramey (2011a) to identify fiscal shocks) could underestimate the effect of the policies in recessions.

Finally, a different approach is to use consumer-level microeconomic data on specific policy events, as highlighted by Parker (2011). For instance, looking at the 2001 and 2008 tax rebates, which reached recipients in different months, Johnson, Parker, and Souleles (2006) and Parker, Souleles, Johnson, and McClelland (2011) show that a sizable fraction of the rebate was spent, especially by lower income or liquidity constrained households. Their results indicate that income transfers can be an effective way to raise consumption in the short run. This approach has the advantage of directly estimating consumption effects, although it does not capture the full dynamic, indirect response of the economy to the fiscal shock.

Cross-Sectional Multipliers

In addition to the works on the Recovery Act cited in the cross-state evidence section, recent work has also exploited other sources of cross-sectional variation in government spending to estimate the size of the fiscal multiplier. For instance, looking at the effects of windfall returns on pension funds, Shoag (2013) estimated a local output multiplier of 2.1. Suarez, Serrato, and Wingender (2011) reached a similar conclusion using on changes in Federal transfers due to the decennial census. Nakamura and Steinsson (2011) detected a 1.5 local multiplier based on regional differences in Federal defense spending.

Table 3-12 summarizes the fiscal multipliers implied by the economic literature on state-level effects of fiscal policies.

 

Table 3-12

 

Summary of Cross-Sectional Fiscal Multiplier Estimates

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 146.]

 

 

 

² See Romer (2011).

³ A comprehensive timeline of the policy actions taken by the U.S. government can be found on the Federal Reserve Bank of St. Louis website http://timeline.stlouisfed.org/

⁴ The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) later reduced that amount to $475 billion. A detailed description of the TARP can be found on the Treasury website http://www.treasury.gov/initiatives/financial-stability/Pages/default.aspx

⁵ Speech at the Wall Street Journal CEO Council conference in Washington, DC, Nov 19, 2008.

⁶ The primary deficit excludes interest payments on the national debt.

⁷ CBO's original estimate of the cost of the Recovery Act, $787 billion (CBO 2009b), was revised to $862 billion (CBO 2010a), then to $814 billion (CBO, 2010b), $821 billion (CBO 2011a), $831 billion (CBO 2012a), $830 billion (CBO 2013a), and most recently to $832 billion (CBO 2014). The estimates evolved because economic conditions deteriorated more than had been assumed in earlier projections, resulting in higher-than-expected use of certain assistance programs.

⁸ Additional detail on these components of the Recovery Act can be found in Appendix 1.

⁹ Title XV, Section 1512

¹⁰ This category includes items like the Research and Experimentation tax credit, the tax deduction for State and local sales taxes for States without income taxes, and numerous other tax provisions that have been routinely extended as a group in the past. Going forward, the President's budget is proposing that all tax extenders are either made permanent and paid for or allowed to expire.

¹¹ See http://www.whitehouse.gov/the-press-office/2011/09/08/fact-sheet-american-jobs-act

¹² This basic multiplier thus equals 1 + MPC + MPC² + . . . = 1/(1 - MPC).

¹³ For State and local fiscal relief, CEA assumed that 60 percent of the transfer is used to avoid spending reductions, and 30 percent is used to avoid tax increases. One-time tax rebates and one-time payments to seniors, veterans, and disabled are assumed to have half of the effects of conventional tax cuts. The effect of business tax incentives is very uncertain. Conservatively, the multiplier to this component is set equal to 1/12 of the spending multiplier. See CEA (2009a).

¹⁴ CBO's methodology is described in Reichling and Whalen (2012).

¹⁵ CEA's estimates of the effects of the subsequent fiscal measures are based on CBO's initial cost estimates, not actual spending. CEA assigned each of the subsequent fiscal measures to the same functional categories that were used to analyze the Recovery Act, and then applied the corresponding multipliers as discussed above. Quarterly costs were interpolated when only annual cost estimates were available.

¹⁶ See Nakamura and Steinsson (forthcoming) and Farhi and Werning (2012) for formal discussion of the relationship between local multipliers and the economy-wide multiplier.

¹⁷ In addition, states were "held harmless" from planned reductions in FMAP rates due to personal income growth prior to the recession and they received an additional increment in the FMAP linked to local unemployment. The analysis presented here relies only on the 6.2 percent (non-cyclical) increase.

¹⁸ See for example Ilzetzki, Mendoza, and Vegh (2011).

¹⁹ http://www.ers.usda.gov/topics/in-the-news/us-drought-2012-farm-and-food-impacts.aspx#.Uu1MXfldV5A

²⁰ Many of these studies are summarized in Munnell (1992) and Fernald (1999).

²¹ See Transportation Investments in Response to Economic Downturns, Special Report 312, Transportation Research Board of the National Academies.

²² See Environmental Protection Agency (2013).

²³ See http://www.epa.gov/recovery/accomplishments.html

²⁴ See CEA's (2010a) second report on the Recovery Act of 2009 for a detailed discussion on the macroeconomic effects of clean energy investment. The latest estimates are presented in CEA's (2010c) fourth quarterly report.

²⁵ See the Office of Science and Technology Policy and the National Economic Council report Four years of broadband growth (2013).

²⁶ For a comprehensive discussion of the various employment benefits programs implemented in recent years, see the Council of Economic Advisers and the Department of Labor report The economic benefits of extending unemployment insurance (2014).

²⁷ CEA counts as public investment any Recovery Act expenditure or tax program that directly results in activity that increases the capital stock of the Federal government, State and local governments, or private firms. We also count provisions that affect the Nation's human capital and knowledge capital, areas not measured in the national income accounts but which economists have identified as crucial to generating long-run economic growth. Note that tax programs are included if they function similarly to direct spending. In other words, entities can claim tax benefits only when associated spending occurs (e.g., the Advanced Energy Manufacturing Tax Credit).

²⁸ Executive Office of the President and Office of the Vice President, The Recovery Act: Transforming the American Economy through Innovation, August 2010.

²⁹ The multiple regression analysis in Taylor (2011), which estimates the effect of fiscal policy in the 2000s, addresses in part the problem of omitted variables but not the problem of simultaneous causality. Taylor measures the direct impact of fiscal policy on income by the component of disposable income due to countercyclical fiscal policy from 2009-based on the 2001, 2008, and Recovery Act fiscal programs. The level of quarterly consumption is then regressed on contemporaneous values of on personal income, the fiscal policy measure, wealth, and oil prices. Thus this regression controls for the separate effects of oil price movements, in case they co-move with fiscal policy. As observed in the text, however, the sign and magnitude of the coefficient on fiscal policy is ambiguous a-priori because of simultaneous causality: it could be positive, zero, or negative. As it turns out, the coefficient is positive but small, a finding that is consistent with fiscal policy having a large positive effect which, in the regression, is offset by the fact that Congress passed it in a recession, or with fiscal policy having little effect. Because of simultaneous causality, this regression analysis, like its graphical equivalent, sheds little light on the question of the effect of fiscal policy.

³⁰ The field of the econometric estimation of causal effects has seen tremendous advances in both methods and applications; for a review see Angrist and Pischke (2010), Sims (2010), and Stock (2010). For additional methodological discussion of simultaneous causality, see Stock and Watson (2010, Chapters 9 and 12).

But the U.S. health care system also faces another enduring challenge: decades of rapid growth in health care spending. While much of this historical increase reflects the development of new treatments that have greatly improved health and well-being (Cutler 2004), most agree that the system suffers from serious inefficiencies that hike costs and reduce the quality of care that patients receive. Another key goal of the ACA was to begin wringing these inefficiencies out of the health care system, simultaneously reducing the growth of health care spending -- and its burden on families, employers, and State and Federal budgets -- and increasing the quality of the care delivered.

This chapter analyzes recent trends in U.S. health care costs and documents a dramatic slowdown in recent years. According to the final data on national health expenditures, real per-capita health spending grew at an average annual rate of just 1.1 percent from 2010 to 2012. Preliminary data as well as projections by the Office of the Actuary at the Centers for Medicare and Medicaid Services (CMS) imply this slow growth continued in 2013, and the CMS projections show real per-capita health spending growth averaging just 1.2 percent over the three years since 2010. These spending growth rates are the lowest rates on record for any two- and three-year periods, and less than one-third the long-term historical average of 4.6 percent that stretches back to 1960. Moreover, they have occurred at a time when the aging of the population would have been expected to modestly increase the growth rate of health care spending.

The historically slow growth in health costs has appeared not only in health care spending, but also in the prices paid for health care goods and services. Measured using personal consumption expenditure price indices, health care inflation is currently running at around 1 percent on a year-over-year basis, a level not seen since 1963. Health care inflation measured using the consumer price index (CPI) for medical care is at levels not seen since 1972. Health care inflation measured relative to general price inflation is also unusually low in historical terms.

An important question is what has caused these trends and whether they are likely to persist in the years ahead. Although the slowdown is not yet fully understood, the evidence available to date supports several conclusions about its causes and the role of the Affordable Care Act.

The 2007-09 recession and its aftermath have likely played some role in the recent slowdown in health costs, and this portion of the slowdown is likely to fade as the economic recovery continues. However, several pieces of evidence imply that the slowdown in health care cost growth is more than just an artifact of the 2007-09 recession: something has changed. The fact that the health cost slowdown has persisted even as the economy is recovering; the fact that it is reflected in health care prices, not just utilization; and, the fact that it has also shown up in Medicare, which is more insulated from economic trends, all imply that the current slowdown is the result of more than just the recession and its aftermath. Rather, much of the slow-down appears to reflect "structural" changes in the U.S. health care system, suggesting that at least part of this trend -- although it is uncertain how much -- is likely to persist. This conclusion is consistent with a substantial body of recent research that seeks to quantify the recession's contribution to the slowdown and has found that the recession alone cannot explain recent trends.

While various non-recession factors unrelated to the ACA appear to be contributing to the recent slow growth in spending -- including a long-term decline in the development of new prescription drugs and a long-term increase in cost-sharing in employer sponsored plans -- the ACA is also playing a meaningful role. For example, by curtailing excessive Medicare payments to private insurers and medical providers, the law has contributed to the recent slow growth in health care prices and spending, reducing health care price inflation by an estimated 0.2 percentage points each year since 2010.

The ACA's measures to reduce costs and improve quality by improving the payment incentives faced by medical providers also appear to be beginning to bear fruit. For example, hospital readmission rates have turned sharply lower since the ACA began penalizing hospitals that readmit a larger number of patients soon after discharge. Similarly, the ACA has substantially increased health care provider participation in payment models designed to promote high-quality, integrated care. These are hopeful signs and provide reason to believe that, as the ACA's payment reforms continue to take effect over the coming years, they will make an important contribution to extending the recent slowdown.

An emerging literature also suggests that the ACA's payment reforms, which operate primarily through Medicare (and, to a lesser extent, through Medicaid), may generate "spillover" benefits throughout the health system. This literature finds that when Medicare reduces payments to medical providers, private payers tend to follow suit, and also finds that the same is true for changes to the structure of how Medicare pays providers. Some recent evidence also suggests that changes in payment structures by one insurer may benefit patients covered by other insurers, even if those other insurers do not adopt the new payment structures. One possibility is that changes by one insurer induce changes in providers' "practice styles" that affect all patients that providers see. This evidence suggests that the ACA's reforms to the Medicare payment system may be, in economic terms, "public goods."

The presence of spillover benefits would imply that the contribution of the ACA to the recent slowdown in health costs growth is considerably larger than previously understood. As noted above, ACA provisions that curb excessive Medicare payments to private insurers and medical providers have directly reduced health care price inflation by an estimated 0.2 percent a year since 2010. A calculation accounting for spillovers raises this estimate to 0.5 percent a year -- a substantial share of the recent slowdown in health care price inflation.

This chapter concludes with a consideration of the economic benefits of a sustained slowdown in health care costs. Over the long run, slower growth in health care spending that is achieved without compromising the quality of care will raise living standards. These gains may be substantial. If even just one-third of the recent slowdown in spending can be sustained, health care spending a decade from now will be about $1,200 per person lower than if growth returned to its 2000-07 trend, the lion's share of which will accrue to workers as higher wages and to Federal and State governments as lower costs.

Recent Congressional Budget Office estimates offer a concrete illustration of the potential for improvements in the Federal fiscal outlook. Since August 2010, CBO has reduced its projections of combined Medicare and Medicaid spending in 2020 by $168 billion and 0.5 percent of gross domestic product (GDP). The $168 billion reduction represents a 13 percent reduction in previously projected spending on these programs and primarily reflects the recent slow growth in health care spending. These revisions are, however, distinct from the deficit reduction directly attributable to the ACA, which CBO estimates will be substantial. Due in large part to the ACA's role in slowing the growth of health care spending, CBO estimates that the provisions of the ACA will directly reduce deficits by about $100 billion over the coming decade and by an average of 0.5 percent of GDP a year over the following decade.

Slower growth in long-term health spending also reduces employers' compensation costs in the short run, increasing firms' incentives to hire additional workers. This chapter surveys the available evidence on the likely effects on employment to conclude that short-run employment gains could be substantial, although the magnitude of these gains is quite uncertain.

This chapter proceeds as follows. The first section quantifies the recent slowdown in health care costs. The second section discusses possible factors behind the slowdown in costs, and also discusses the effects of the ACA on quality of care so far and in the future. The final section discusses the slowdown's potential economic benefits.

The analysis is extended through 2013 using the most recent NHE projections, which were published by CMS in September 2013 (Cuckler et al. 2013) and reflect Medicare and Medicaid spending data and macroeconomic data available through June 2013 (CMS Office of the Actuary 2013).¹ NHE "tracking" estimates constructed by the Altarum Institute using data on health spending from the Bureau of Economic Analysis imply that final estimates of NHE for 2013 will come in very close to the CMS projection (Altarum 2014).

 

Table 4-1

 

Real Per-Capita NHE Annual Growth Rates

 

by Payer and Spending Category

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 151.]

 

 

 

The slow growth is reflected in all three payer categories depicted in Figure 4-3, which appears on page 157. Real per enrollee spending growth in private insurance over the 2010-13 period is less than one-quarter its level from 2000-07 and less than one-third its level from 2007-10. The change in Medicare spending growth has been similarly dramatic, with real growth in per beneficiary Medicare costs essentially ceasing over this period. In Medicaid, the already slow growth in real per beneficiary costs seen in recent years has continued and turned slightly negative from 2010 to 2013.

Panel A of Table 4-2 documents a similar slowdown in the growth of prices paid for health care goods and services, which is also depicted in Figure 4-2. Health care inflation, whether measured using the personal consumption expenditure (PCE) price indices or the CPI for medical care, is running at half or less the rate seen historically, and below the rates seen over the last decade. Indeed, year-over-year inflation as measured using PCE data is currently running at around 1 percent, a level last seen in 1963. The recent behavior of the CPI for medical care is similar, with recent months' year-over-year inflation rates reaching low levels not seen since 1972.

It is important to note that this slow growth in prices for health care goods and services is not simply a reflection of the fact that the prices of all goods and services have grown slowly in recent years. Panel B demonstrates that health care inflation relative to general price inflation has also been unusually low over the last few years.

 

Figure 4-1

 

Growth in Real Per Capita National Health Expenditures,

 

1961-2013

 

 

 

Source: Centers for Medicare and Medicaid Services, National Health Expenditure Accounts; Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

 

Figure 4-2

 

General and Health Care Price Inflation, 1960-2013

 

 

 

 

Table 4-2

 

Recent Trends in Several Indicators of

 

Health Care Spending and Price Growth

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 155.]

 

 

 

While final conclusions about the causes of the recent slow growth and its persistence await additional data and analysis, some conclusions are possible with the data currently available. Most importantly, the recent slow growth does not appear to be the result of idiosyncratic factors affecting a single category of spending or a particular payer. As documented in Table 4-1, the slowdown has affected all major payers and each of the major categories of spending. The search for explanations must, therefore, look for factors affecting behavior system-wide. The first part of this section examines the role of the 2007-09 recession, the second part discusses potential non-ACA, non-recession explanations for the recent slow growth, and the third part considers the role of the Affordable Care Act, both to date and in the future.

The Role of the 2007-09 Recession

Some have identified the 2007-09 recession and its aftermath as a potential driver of system-wide changes. For example, job losses may have caused reductions in insurance coverage that curtailed access to health care, or the accompanying falls in families' disposable incomes could have forced households to prioritize other needs over medical care. Alternatively, disruptions in financial markets could have depleted providers' cash reserves or reduced their ability to borrow in order to invest in new equipment or facilities, leading to lower utilization in subsequent years.³ If the recession were the primary driver of the current slow growth in health spending, then growth would likely return to its earlier rapid rate as the economic recovery continues.

 

Figure 4-3

 

Growth in Real Per Enrollee Health Spending by Payer

 

 

 

Source: Centers for Medicare and Medicaid Services; Bureau of Economic Analysis, National Income and Product Accounts; CEA calculations.

Three features of the recent slow growth in health care costs are inconsistent with the theory that this slow growth results only from the recession, suggesting that a substantial portion of the recent slowdown is "structural" and likely to persist. First, and most simply, the slowdown has now persisted well beyond the end of the recession. The Great Recession began in December 2007 and concluded by June 2009. Since that time, the economy has recorded four years of steady growth. Yet, as shown in Table 4-1 and Figure 4-3, health spending growth has remained subdued relative to the years during and immediately following the recession. While the economy may affect health spending with a lag, if the recession were the primary force driving the slowdown, more substantial acceleration would likely be visible by now.

Second, as documented in Table 4-1 and Figure 4-3, the slowdown has affected Medicare in addition to the private sector, a fact highlighted in a recent analysis by CBO economists (Levine and Buntin 2013). Because seniors are generally more insulated from a weak labor market, this fact undermines the notion that the slowdown results primarily from economic disruptions attributable to the recession. In addition, Levine and Buntin find that even those seniors who did experience relatively larger economic disruptions during the recession did not spend less on health care. Levine and Buntin also document, using State-level data, that Medicare spending growth has historically risen when unemployment rises -- the opposite of the pattern required for the economic downturn to explain the slowdown in cost growth.⁴

Third, the recent behavior of health care inflation is difficult to square with the theory that the slowdown is primarily a result of the recession. As documented in Table 4-2 and Figure 4-2, health care inflation has decelerated sharply of late, even when measured relative to inflation in the broader economy. While there are a variety of plausible mechanisms by which the recession could reduce the quantity of health care services people demand, and thus reduce total spending, it is difficult to explain why a recession should cause a reduction in the growth rate of health care prices relative to price growth in other sectors of the economy.

Many recent studies have also attempted to directly quantify the role of the recession in driving recent slow growth in health care spending. These analyses, which use a variety of methods, have generally concluded that, while the recession likely has depressed health care spending growth in recent years, health spending is low in historical terms even after accounting for the recession, and a substantial fraction of the slowdown likely reflects structural changes that are likely to persist. The remainder of this section provides a review of this growing literature.

Chandra, Holmes, and Skinner (2013) provide one approach to evaluating the role of the recession. They survey the available micro-econometric estimates of the effect of income on the demand for health care. Virtually all such estimates in the existing literature are small, with the largest credible estimates of the income elasticity being the 0.7 estimate provided by Acemoglu et al. (2013). Applying this upper-bound estimate to the observed slowdown in GDP growth, they show that the slow economic growth in recent years explains less than half of the recent slow growth in health spending. Although they express some uncertainty about the future outlook for health spending, they nevertheless project that a substantial fraction of the slowdown will persist, due in part to the potential of payment reforms included in the Affordable Care Act.

Ryu et al. (2013) take another approach. They examine the role of two specific mechanisms by which the recession could have affected health care cost growth: by reducing insurance coverage via job loss and by causing firms to offer their employees leaner health plans that require greater cost-sharing. Focusing on the period 2009-11, they find that recent reductions in spending growth are, if anything, larger among employed individuals, and that increases in cost-sharing can account for only one-fifth of the slow-down. On the basis of their results, they advise a "cautious optimism that the slowdown in health spending may persist."

Another set of studies evaluates the effect of the recession by estimating the historical relationship between economic growth and health spending growth and using this estimated relationship to simulate how health spending would have evolved had the recession not occurred. Econometric time series analyses like these have the important advantage that, by virtue of their nationwide, aggregate approach, they can capture the effects of a wide variety of potential mechanisms connecting economic growth to health spending growth. But the nationwide, aggregate nature of these analyses is also a weakness; it can be difficult to plausibly control for important confounding factors, and the paucity of data (only about 50 years of data, or about 50 total data points, are available) can make these analyses sensitive to seemingly innocuous changes in methodology, as demonstrated by Chandra, Holmes, and Skinner. The current literature does not, unfortunately, provide persuasive evidence on which econometric specifications are likely to provide the most reliable results.

Cutler and Sahni (2013) estimate a model relating current health spending growth to a five-year average of economic growth. Based on their results, they estimate that spending growth in 2011 and 2012 would have been on the low end of the historical range even accounting for the recession, and that more than half of the slowdown over the longer period 2003-12 is due to factors other than the recession. They conclude that "fundamental changes" are underway in the health sector, changes that are not attributable to the recession alone.

A contrary perspective comes from an analysis from the Kaiser Family Foundation and the Altarum Institute (KFF and Altarum 2013). They estimate a model relating current health spending growth to economic growth the current year and each of the prior five years and general price inflation in the current year and each of the prior two years. On the basis of their estimated model, the authors conclude that most of the slowdown in health care spending from the 2001-03 period to the 2008-12 period is attributable to the macroeconomic factors, although they still attribute 23 percent of the slowdown to non-macroeconomic factors.

It is important to note, however, that the authors' calculation applies to the slowdown in nominal health spending growth over this period, while the slowdown in real (that is, inflation-adjusted) health spending growth is of greater economic interest. Because inflation was, on average, lower during the 2008-12 period than during the 2001-03 period, the authors' approach overstates the role of macroeconomic factors in explaining the slowdown in real spending growth. In addition, the authors' model, by virtue of its relative complexity, is particularly subject to the shortcomings of the time series approach described above. Indeed, the model estimated by KFF and Altarum has one particularly unusual feature: the effect of reduced economic growth on health spending actually peaks four years later. While not impossible, such lags seem implausibly long.

Non-ACA Factors Affecting Health Spending Growth

As discussed above, the recession does not provide a full, or even necessarily a major, explanation for the recent slow growth in health spending. While additional factors may be identified in the future, two non-ACA factors have received substantial attention to date -- although it is important to note that at least one non-ACA factor is modestly increasing health spending growth.

The long-term trend toward increased patient cost-sharing is one factor that can plausibly explain why slow growth has affected many different categories of spending at the same time (Cutler and Sahni 2013; Ryu et al. 2013; Chandra, Holmes, and Skinner 2013). The Kaiser Family Foundation/ Health Research and Educational Trust Employer Health Benefits Survey indicates that recent increases in cost-sharing in employer plans have been substantial; the typical deductible in an employer plan has increased from $584 in 2006 to $1135 in 2013, a 70 percent increase after adjusting for inflation (Kaiser Family Foundation 2013a).

Some research suggests that the observed increase in cost-sharing is having an effect. As noted above, Ryu et al. (2013) examine the importance of increased cost-sharing in the employer context and conclude that it can account for 20 percent of the reduction in growth over the 2009-11 period. Chandra, Holmes, and Skinner (2013) evaluate the role of increased cost-sharing using estimates from the literature of how utilization responds to cost-sharing. They conclude that cost-sharing may have played a larger role, although the precision of their estimates is limited by the poor quality of the available data on recent changes in cost-sharing and the current incomplete understanding of how cost-sharing affects utilization.

While it seems possible and perhaps likely that increased cost-sharing is playing a role, it cannot be the whole story. As discussed in detail above, the slowdown in Medicare fee-for-service spending has been even more dramatic than the slowdown in the private sector, and there have been no substantial changes to the core Medicare benefit design in recent years that parallel the changes seen in the private sector.

The striking slowdown in prescription drug spending, documented in Table 4-1, also factors into the slow growth trend. Various sources attribute this sharp drop in prescription drug spending to the expiration of patent protection for many important drugs. Due to a slowdown in the invention of new drugs that stretches back more than a decade, the drugs that have come off patent in recent years are not being replaced by more-recently patented drugs. As a result, the share of prescriptions accounted for by generic drugs -- which typically cost much less -- has increased sharply, substantially reducing costs (Aitken, Berdnt, and Cutler 2009; Cutler and Sahni 2013; IMS 2013). However, these changes in the prescription drug market are probably only making a modest contribution to aggregate trends in health spending since prescription drugs account for less than 10 percent of total health spending.

There is, however, at least one easily identified factor working against the recent slowdown: the aging of the U.S. population. In recent decades, population aging has made a small positive contribution to the growth of U.S. health spending; White (2007) estimates that over the period 1970-2002, population aging added about 0.3 percentage points to annual growth. The contribution of population aging to health care spending growth appears to have increased by a small amount in recent years. Using data on the age distribution from the U.S. Census Bureau, data on spending by age reported by Yamamato (2013), and a methodology similar to that used by Yamamato, the CEA estimates that population aging added about 0.5 percent to annual growth in health care spending over the 2000-07 and 2007-10 periods and added about 0.8 percent to growth over the 2010-13 period.⁵ These demographic headwinds mean that the slowdown in the growth of health care costs for an individual of any particular age is actually slightly larger than shown in Table 4-1.⁶

The Role of the Affordable Care Act

The evidence discussed above shows that the recession is not the sole cause of the recent slow growth in health spending, and that the other factors identified to date cannot explain the magnitude or broad scope of the slowdown. What, then, is the Affordable Care Act's role in driving changes in the Nation's health care system? To be sure, the ACA is not the sole cause of the slowdown. Health care spending growth had slowed somewhat even before the ACA was passed (as shown in Table 4-1), the recession and other changes in the health system have certainly made contributions (as discussed above), and many of the ACA's reforms have yet to take full effect.

Nevertheless, the ACA's reforms aimed at driving out waste and improving quality are contributing in a meaningful way to recent slow growth in health costs -- including by building on pre-existing trends in delivery system reform and initiating new ones -- and are likely to make larger contributions in the future. Recent economic research also provides support for the premise that implementing reforms in Medicare can reduce the cost and improve the quality of care system-wide. This research supports the idea that the ACA will play an important role in slowing health care cost growth over the long term, but also suggests that its provider payment reforms may be having a larger-than-anticipated impact today.

Deployment of new payment models. The ACA also includes many reforms intended to identify and promote payment models that encourage efficient care delivery, reduce care fragmentation, and reward physicians, hospitals, and others that invest in providing high-quality care rather than just a high quantity of care.

The ACA made direct changes in Medicare payment systems aimed at achieving these goals, including creating the readmissions reduction and shared savings programs discussed in detail below and various "value-based" purchasing initiatives that tie provider reimbursement to measures of the quality of the care received by patients. The ACA also provided additional financial assistance to states through Medicaid to establish health homes to improve care management for patients with chronic conditions.

In addition, the ACA created the Center for Medicare and Medicaid Innovation (the "Innovation Center") to experiment with diverse new payment approaches, including bundled payments, various accountable care models, and multi-payer initiatives, each of which will be touched on later in this section. To date, more than 50,000 health care providers from across every state are participating in an Innovation Center initiative. The Secretary of Health and Human Services has the authority to take successful pilots to scale.

Finally, through the Patient-Centered Outcomes Research Institute, the ACA is funding efforts to identify which treatments work -- and for which patients -- and to identify strategies for translating that evidence into practice. By giving providers the information they need to provide efficient, high-quality care, this research initiative directly complements the ACA's efforts to change provider incentives.

The full benefits of the initiatives described above will only be realized in the years to come. However, the next two subsections discuss a pair of payment reforms -- the ACA's incentives to reduce hospital readmissions and its deployment of accountable care payment models -- that are already beginning to show results.

Incentives to reduce hospital readmissions. The ACA made important changes in how Medicare's hospital payment system treats hospital readmissions -- cases in which a patient returns to the hospital soon after being discharged. Historically, nearly one-in-five Medicare patients were readmitted within 30 days of discharge, and it is commonly believed that many of these readmissions result from low-quality care during the initial admission or poor planning for how the patient will obtain care after discharge. Prior to the ACA reform, hospitals faced no financial incentive to invest in activities aimed at reducing readmissions, and could actually be made financially worse off by doing so since they lose payment for the avoided readmissions. This misalignment of incentives likely both increased costs and reduced quality.

The ACA aims to correct these incentives by penalizing hospitals with high readmission rates for patients with a specified set of diagnoses. Many of the rules governing these penalties were finalized in August 2011. The penalties took effect at the start of FY 2013 (October 2012), but because penalties for a given fiscal year are based on hospitals' readmission rates in prior years, hospitals' incentives to begin reducing readmissions began as soon as the rules were finalized (or earlier, to the extent that hospitals anticipated the structure of the payment rules).⁸ The number of conditions included in the program and the maximum penalty amount will grow over time.

Figure 4-4 provides evidence that this readmission policy has begun changing patterns of care. After having been flat for several years, overall 30-day hospital readmission rates for Medicare patients turned sharply lower soon after the program rules were finalized, and, as of July 2013, were more than one percentage points below their average level from 2007-11. From January 2012 through August 2013, this reduction corresponded to 130,000 avoided readmissions (CMS, 2013a). The sharp change in trend -- and its timing -- implies that the readmissions program played an important role in causing these changes, although other efforts to reduce readmissions were underway during this period as well. Among those other activities were efforts by the Department of Health and Human Services efforts to actively engage hospitals and community-based organizations in improving discharge processes through the Partnership for Patients and the Community-Based Care Transitions Program (Gerhardt et al. 2013).

 

Figure 4-4

 

Medicare 30-Day, All-Condition Hospital

 

Readmission Rate, 2007-2013

 

 

 

Source: Centers for Medicare and Medicaid Services, Office of Information Products and Data Analytics.

Accountable care payment models. Another important ongoing ACA reform is the creation of "accountable care" payment models through the Medicare Shared Savings Program and the Innovation Center. These programs seek to realign provider incentives to encourage provision of efficient, high-quality care. Under fee-for-service payment systems, providers delivering more efficient care often end up financially worse off because lower service volume translates into lower payments from Medicare. In addition, since provider payments were based on service volume, the pre-ACA payment system gave providers no direct financial incentive to deliver high-quality care. Prevailing fee-for-service payment systems also pay each provider separately without regard to how services furnished by that provider fit into the patient's broader plan of care, and thus create no incentive for efficient coordination of care across providers.

Under these accountable care programs, a provider or group of providers can seek designation as an Accountable Care Organization (ACO). ACOs are eligible to share in the savings created when they reduce the cost of caring for patients assigned to them, which encourages providers to be efficient in the use of additional services. In addition, because the ACOs earn shared savings based on the total costs of a patient's care across all providers and not merely the costs for any particular visit or procedure, ACOs have incentives to invest in care coordination and avoid duplication. Perhaps most important, ACOs must achieve designated benchmarks for the quality of care received by their patients in order to be eligible for shared savings, which provides strong incentives to ensure that patients receive high-quality care.

Today, more than 360 organizations serving 5.3 million Medicare beneficiaries have adopted the ACO model, and the number of beneficiaries covered will likely grow in the years ahead. A preliminary evaluation of the Pioneer ACO program (the Innovation Center ACO program for large and advanced systems) found that costs for beneficiaries aligned with Pioneer ACOs grew more slowly from 2011 to 2012 than costs for similar beneficiaries not aligned with ACOs (L&M Policy Research, 2013). The annual cost savings for each enrollee aligned with a Pioneer ACO in 2012, the first year of the program, were estimated to be at least $150, more than 1 percent of average Medicare spending per beneficiary in that year. In addition, overall, the ACOs performed better than fee-for-service benchmarks on all quality measures for which comparable data are available (CMS 2013b). Academic research on similar private models also suggests that these payment models can achieve their intended purpose of reducing costs while improving quality (Song et al. 2012).

The Innovation Center is experimenting with related payment models through its Bundled Payment for Care Improvement Initiative, which got underway in 2013.⁹ Under these models, Medicare will make a single "bundled" payment for all services provided during an "episode" of care connected to a hospital stay, rather than paying separately for each service provided during that episode. In the model using the most comprehensive bundle definition, this payment will cover the hospital stay, physician services provided during the stay, and post-hospital care. The Innovation Center is also testing models with narrower bundles covering only services provided during the hospital stay or only services provided after the hospital stay. Although the details vary across payment models, the bundled payment will then be allocated across the participating providers according to agreements among the providers themselves.¹⁰ The models are being tested for a set of common types of hospital care episodes that account for a significant fraction of all hospital stays.

Much like the accountable care payment models, these bundled payment models encourage providers to be more efficient because providers receive no additional payment for providing additional services (if the service is included in the bundle). Similarly, because all providers involved in an episode of hospital care are jointly accountable for the total cost of the care episode, the bundled payment structure gives providers strong incentives to coordinate their activities, with attendant benefits for efficiency and quality of care. Because of this scope for increased efficiency, Medicare can (and does under the models being tested) set the bundled payment amount below the total amount it would pay under the existing fee-for-service payment systems. The efficiency gains from these sources could be substantial. CBO recently estimated that if a bundled payment model that covered services provided during and after the hospital stay and used a 5 percent savings target were phased in nationwide starting in 2017, the savings to Medicare would total $47 billion over 10 years (CBO 2013c).

Recent research on cross-payer "spillovers." In evaluating the direct effects of the ACA's Medicare and Medicaid reforms so far and considering their likely effects going forward, an important question is how these reforms will influence the rest of the health care system. Recent empirical work in economics and health policy strengthens the premise that reforms to public-sector health programs that reduce waste and improve quality will have spillover benefits for the private sector.¹¹

In particular, various recent studies suggest that efforts by Medicare to reduce excessive payments for particular services are likely to generate corresponding savings for private insurers and their enrollees. Clemens and Gottlieb (2013) study how the prices that private insurers pay to physicians change when Medicare changes its prices, exploiting a natural experiment created by regional differences in the effect of earlier reforms to the way Medicare pays physicians. They find that when Medicare reduces the price it pays for services, private insurers are able to reduce the amount they pay for care by similar amounts.

White (2013) and White and Wu (2013) undertake a similar analysis focused on Medicare payment to hospitals that exploits natural experiments created by cross-hospital differences in the effect of earlier Medicare payment changes. White (2013) finds that when Medicare reduces its payment rates, private payers reduce their payment rates by approximately 77 percent of that amount. White and Wu (2013) find that for each dollar of Medicare savings, private insurers realize additional savings of 55 cents.

The implications of these estimates are striking. For example, the $17 billion in Medicare savings estimated to have been achieved in FY 2013 as a result of reducing excessive Medicare payments. Using the same logic applied previously, these estimated savings correspond to a 0.2 percentage point reduction in the average growth of health care prices over the period 2010-13. If just half of these price reductions spilled over to the rest of the health care system to the extent estimated by White (2013), then the implied reduction in health care inflation economy-wide due to these Medicare changes would be about 0.5 percent.¹² In this scenario, the ACA would be playing a significant role in driving the observed slow growth in health care prices -- representing about half of the recent slowdown in health care inflation relative to general price inflation.¹³

Potentially even more important, the work by Clemens and Gottlieb provides evidence that the benefits of the ACA's improvements to the structure of public-sector payment systems may be realized system-wide,not just among enrollees of those programs. Again focusing on Medicare payment for physician services, they show that Medicare payment changes that increase payment for some services and reduce payment for others tend to be matched by private insurers. Clemens and Gottlieb's results provide empirical support for the widely believed notion that Medicare's payment structure serves as the "starting point" in negotiations between providers and private insurers in many circumstances, in which case changes in Medicare will reasonably quickly get picked up in the private sector as well. This evidence is consistent with historical experience. Medicare introduced "prospective" payment for inpatient services in the 1980s, under which all care during an inpatient admission was covered via a single payment determined based on the patient's diagnosis; virtually all private insurers pay hospitals using this type of system today.

Some recent evidence suggests that spillover benefits from the ACA's public-sector payment reforms may occur even if private payers do not directly adopt these payment models. McWilliams et al. (2013) study the Alternative Quality Contract (AQC) -- a contract similar to accountable care payment structures currently being deployed by CMS-that Blue Cross Blue Shield of Massachusetts has been experimenting with since 2009. Research cited above (Song et al. 2012) finds that the AQC reduces costs and improves quality for patients whose care is directly subject to the contract. The research by McWilliams et al. finds, however, that patients associated with AQC-participating providers whose care was not subject to the contract (in this case, Medicare patients) also experienced improvements. In this case, the cost savings amounted to 3.4 percent, on average, and was accompanied by improvement on some quality measures. The results may arise because providers adopt a single "practice style" for all their patients, so that when incentives from one induce a provider to change its approach in ways that improve efficiency or quality, all patients seen by that provider benefit.

Taken together, the evidence of cross-payer spillovers reviewed above suggests that not only are reforms to the structure of the public-sector payment systems helpful in reducing costs and improving quality system-wide, but that the public sector may be essential to fully realizing the potential for improvement. In economic terms, the presence of spillovers means that payment system reforms are "public goods" -- investments that generate benefits for many people other than the purchaser and for which the purchaser cannot capture all the resulting benefits (Clemens and Gottlieb 2013). Because no individual investor captures the full benefits of investment in public goods, the private market generates too few of them. As with other public goods, one solution to the underinvestment is for the government to invest directly, in this case by implementing reforms through Medicare and Medicaid.

Recognizing the importance of other payers' decisions in determining providers' response to new payment arrangements, CMS has launched demonstration projects that actively engage multiple payers. Incorporating multiple payers into reform efforts at the outset may increase the possibility that the payment models that emerge can easily cross payer boundaries, once proven. These initiatives also recognize that engaging private payers in reform efforts is important for Medicare and Medicaid beneficiaries themselves, in light of the evidence described above that spillovers can run in both directions: from Medicare and Medicaid to the private sector, and vice versa.

Two multi-payer initiatives merit special mention. Through the Comprehensive Primary Care Initiative, CMS has enlisted public and private payers in eight states to join with Medicare to invest in primary care practices, with the potential for shared savings after two years. Another promising effort is the State Innovation Models Initiative, which provides grants to states that wish to make statewide, multi-payer changes to provider payment systems. With support from this program, Oregon has embarked upon an effort to move its Medicaid beneficiaries, State employees, and individuals who have purchased coverage through the state's ACA-created health insurance marketplace into ACO-like payment models. Arkansas has undertaken an initiative involving public and private payers aimed at ensuring that half of Arkansans have access to a patient-centered medical home by 2016, and expanding its existing system of episode-based payment.

Higher Living Standards

All else equal, when the health sector consumes less of the Nation's output, more resources are available for meeting other needs. As a result, reductions in health care spending that stem from improving efficiency or eliminating low-value care have the potential to improve living standards. Because of the large share of the Nation's resources devoted to health care, even relatively modest reductions can have very large effects on economic well-being.

 

Figure 4-5

 

Inflation-Adjusted Premiums for Medicare

 

Parts B and D, 2000-2014

 

 

 

One straightforward way of illustrating the magnitude of the potential impacts is to consider the effect of continuing the slow growth of the last few years. To that end, recall from Table 4-1 that national health expenditures have grown at a 1.2 percent real per capita annual rate from 2010-13, whereas health spending grew at a 4.0 percent rate from 2000-07. Suppose that even just one-third of that slowdown continued, so that instead of returning to the recent historical rate of 4.0 percent, real per capita health care costs instead grew at a 3.1 percent rate, similar to the rate projected in the recent work by Chandra, Holmes and Skinner (2013). Under this illustrative scenario, the savings after a decade would amount to about $1,200 per person. As discussed above, these savings would materialize primarily in the form of higher wages and lower State and Federal costs.

Lower Deficits

In 2013, the Federal Government devoted 22 percent of the U.S. budget, or 4.6 percent of GDP, to Medicare and Medicaid. For this reason, the future path of health care costs has major implications for the long-term budget outlook.

Over the last three years, CBO has made a series of downward revisions to its forecast of future spending on Medicare and Medicaid (CBO 2010a; 2011; 2012c; 2013a; 2014), which are depicted in Figure 4-6. From the projections CBO published in August 2010 to its most recent set of projections in February 2014, CBO has reduced its estimate of Medicare and Medicaid spending in 2020 (the latest year covered by all of the projections examined here) by $168 billion and 0.5 percent of GDP.¹⁴ This $168 billion represents a 13 percent reduction in spending relative to CBO's earlier projection of spending on these programs.

These reductions primarily reflect lower projections of future growth in health care costs.¹⁵ To that point, in a recent presentation, CBO Director Douglas Elmendorf commented: "The slowdown in health care cost growth has been sufficiently broad and persistent to persuade us to make significant downward revisions to our projections of Federal health care spending" (Elmendorf 2013).

It is important to note that the reductions in projected Medicare and Medicaid spending described above are separate from the deficit reduction that CBO estimates will occur as a direct result of the ACA. The most recent CBO estimates indicate that the ACA will reduce the deficit by about $100 billion over the decade 2013-22, and that it will reduce the deficit, on average, by about 0.5 percent of GDP in the subsequent decade (CBO 2012b). CBO notes that these deficit-reducing effects are likely to continue to grow in following decades.

 

Figure 4-6

 

Recent CBO Projections of Medicare and Medicaid Outlays

 

 

 

Sources: Congressional Budget Office, Budget and Economic Outlook; CEA calculations.

Higher Employment and Economic Growth

Slower growth in health care costs reduces the growth of the health insurance premiums paid by employers. As discussed above, in the long run, because employers must compete for workers, reductions in the cost of health care are likely to be passed through to workers in the form of higher wages. Thus, over the long run, changes in the growth rate of health care costs are unlikely to substantially affect employer's hiring costs and decisions.¹⁶

In the short run, however, the picture may differ. Wage setting is subject to various "rigidities" that mean that lower health insurance costs may not be fully passed through in the short and medium run, potentially reducing employer costs and spurring hiring (Sommers 2005). Rigidities of this kind may be particularly important in the aftermath of the 2007-09 recession, as abnormally low inflation has increased the importance of constraints on the adjustment of nominal wages (Daly et al. 2012).

There is relatively little empirical literature on the effect of slower growth in employer health insurance premiums on employment, and there is no consensus among economists about the likely size of these effects. There are, however, at least two empirical studies suggesting that these effects could be substantial.

Baicker and Chandra (2006) use variation in employer health insurance costs resulting from within-state changes in medical malpractice costs over time to estimate the effect of higher health insurance premiums on employment. They find that a 10 percent reduction in insurance premiums increases the share of working-age individuals who are employed by 1.2 percentage points. This estimate suggests that the recent slowdown in the growth of health insurance premiums could have had a substantial positive effect on employment.

Sood, Ghosh, and Escarce (2009) take an alternative approach to quantifying the effect of faster premium growth on employment. Specifically, they examine whether industries that provide insurance to a large share of their employees experience relatively lower employment growth during periods when health costs are growing particularly rapidly. They find that, for an industry that provides health insurance to all of its workers, increasing health insurance premiums by 1 percent reduces the industry's employment by 1.6 percent relative to an industry that insures none of its workers.

Translating the Sood, Ghosh, and Escarce estimates into effects on aggregate employment is difficult because their results could arise either because higher health insurance costs reduce employment overall or because they cause a reallocation of employment from high-coverage industries to low-coverage industries. Cutler and Sood (2010) make one set of plausible assumptions about the importance of these two types of employment changes, and given their estimates of the effect of the ACA on the path of health care costs, find that the ACA will increase job growth by 250,000 to 400,000 a year by the second half of this decade.

In the longer run, lower deficits due to the ACA and the slowdown in health costs also have the potential to improve economic growth. Reductions in long-term deficits increase national saving, which increases capital accumulation and reduces foreign borrowing, and thereby increase national income and living standards over time. As discussed in detail in a 2009 CEA report on the potential benefits of health care reform for the economy, this means that even modest sustained reductions in health care cost growth can generate substantial economic benefits (CEA 2009).

² The periods 2000-07 and 2007-10 were chosen as comparison periods in order to facilitate the discussion in the next section of the role of the 2007-09 recession in driving recent trends.

³ The NHE data do show a very sharp reduction in investment in equipment and structures in the health care sector over 2009 and 2010 of about 12 percent in real per capita terms. It is worth noting, however, that this contraction followed two years of very strong investment growth. Moreover, even as financial conditions have normalized, investment has remained subdued, suggesting that providers do not view themselves as having incurred a substantial investment deficit, nor suggesting an imminent investment-driven rebound in health care cost growth.

⁴ The 2013 Economic Report of the President undertakes a related analysis (CEA, 2013). The report analyzes changes in state-level unemployment from 2007-09 to state-level health spending growth over that period. While that analysis finds that unemployment is associated with lower health spending growth, the effect is small and cannot explain a substantial fraction of the recent downturn in health spending.

⁵ As Yamamato notes, this methodology assumes that spending does not change discontinuously at age 65 when individuals transition to Medicare. It also does not account for differences in coverage mix by age in the under-65 population. It does not appear that accounting for these factors would meaningfully alter the results, but further research in this area would be worthwhile.

⁶ The effect of changing demographics on per beneficiary spending by particular payers may differ from the effect on the overall population. The average age of Medicare beneficiaries is currently falling as the youngest baby boomers reach age 65. Consistent with that, Levine and Buntin (2013) estimate that changes in the age mix of Medicare beneficiaries had no effect on per beneficiary growth in Medicare spending over the 2000-07 period, but subtracted 0.2 percentage points over the 2007-10 period. Calculations like those in the main text suggest that these changes in age mix subtracted somewhat more from growth, on the order of 0.4 percentage points, over the 2010-13 period. This represents a modest, but not trivial, share of the overall slowdown in Medicare spending growth.

In addition, changes in beneficiary mix (that are not primarily attributable to the aging of the population) appear to have had a larger effect on recent trends in per beneficiary Medicaid spending. Over the period 2000-10, Medicaid enrollment among children, parents, and pregnant women increased substantially more rapidly than did enrollment among elderly and disabled individuals (Kaiser Family Foundation, 2013b). The resulting change in enrollment mix lowered per beneficiary costs since non-elderly, non-disabled beneficiaries generally use less health care. Holahan and McMorrow (2012) estimate that this change in enrollment mix subtracted 1.5 percentage points from the annual growth of Medicaid spending per beneficiary spending over the 2000-10 period. However, enrollment data reported by the Kaiser Family Foundation suggest that, if anything, changes in enrollment mix have actually increased per beneficiary costs since 2010. Thus, adjusting for enrollment mix would make the slowdown in per beneficiary Medicaid costs over the 2010-13 period more dramatic than shown in Table 4-1.

⁷ This chapter also cites a CBO estimate of the budgetary effect of repealing the ACA from July 2012, which suggests that repeal would increase Medicare spending in FY 2013 by $4 billion, a much smaller sum than the $17 billion cited here. However, as discussed in the CBO letter, because it would have been too late to unwind some ACA provisions for FY 2013 and due to other effects, this estimate does not reflect the full effect of the ACA in that year.

⁸ Under current program rules, a hospital's penalties in a given fiscal year are based on its readmission rate during the three-year period that ended five quarters earlier.

⁹ These models build on several earlier Medicare demonstration projects, with the most similar being the Acute Care Episode (ACE) demonstration, a much smaller demonstration that concluded in 2013.

¹⁰ The bundled payment is administered in different ways under the different models. In the model covering only services during the hospital stay, the bundled payment is paid "prospectively" to a single entity (e.g. the hospital), which is then responsible for paying the other providers involved in episode. In the other models, Medicare continues to pay providers according to the existing fee-for-service rules. If total fee-for-service payments are below the bundled payment amount, Medicare pays the excess to a designated provider, which distributes that excess among the other involved providers. If total fee-for-service payments are above the bundled payment amount, the reverse occurs. In principle, the two structures change provider incentives in similar ways.

¹¹ This growing literature is contrary to the traditional view in some health policy circles, which held that efforts to achieve savings in Medicare (or Medicaid) cause medical care providers to increase the prices they charge to private insurers in order to recover the lost revenue, and, thus, reforms in Medicare simply "shift" costs to the private sector rather than reducing them. The empirical support for this view was always inconsistent, and, as argued by Dranove (1988) and Morrissey (1994), this view has important conceptual shortcomings. In particular, for hospitals to be able to increase the prices charged to private payers after a reduction in Medicare payment rates, they must have been willingly charging a price below what the market would bear prior to the reduction in Medicare rates. For a comprehensive overview of this literature, particularly the older literature, see Frakt (2011; 2013).

¹² The reductions in excessive payments to Medicare Advantage plans are less likely to "spill over" to general private-sector payment rates (although to the extent they lead MA-participating insurers to negotiate lower provider payment rates, such spillovers could occur under certain models of spillovers). Since the Medicare Advantage reductions account for about half of the estimated $17 billion in payment reductions in 2013, the calculation in the text assumes that only half of this reduction would spill over.

¹³ Of course, effect on total spending may be smaller or larger to the extent that these price changes induce changes in volume. Indeed, the estimates of White and Wu, referenced above, as well as estimates reported by He and Mellor (2012) suggest that volume changes will generally work to offset these price spillovers. However, even under the estimates of White and Wu (2013), the savings to private insurers as a result of Medicare changes would be substantial.

¹⁴ In July 2013, the Bureau of Economic Analysis (BEA) released comprehensive revisions to the National Income and Product Accounts that increased BEA's estimate of GDP in recent years by more than 3 percent. CBO projections of GDP released before and after these revisions are, therefore, not directly comparable.

The figures reported in the text and displayed in Figure 4-6 account for this issue in the following manner. For May 2013, CBO released two sets of GDP projections, one before and one after the BEA revisions; the figures shown use the GDP projections released after the BEA revisions. For earlier CBO baselines, CEA adjusted CBO's projections of GDP upward by the ratio of CBO's post- and pre-revision May 2013 GDP projections. Without these adjustments, the reduction in projected Medicare and Medicaid spending as a share of GDP in 2020 from CBO's August 2010 baseline to its February 2014 baseline would be 0.6 percent, rather than the 0.5 percent reported in the text, and the decline in Medicare and Medicaid spending shown in Figure 4-6 would be larger.

¹⁵ Several factors other than recent slow growth in health care costs have affected CBO's projections of Medicare and Medicaid spending over this period. These factors work in different directions. First, CBO has revised its general economic projections in ways that, on net, increase projected future Medicare and Medicaid spending by around $25 billion. Second, CBO estimates issued after June 2012 incorporate the Supreme Court decision in NFIB v. Sebelius. CBO materials indicate that this ruling reduced projected Medicaid spending in 2020 by roughly $30 billion as of July 2012, although this figure has likely fluctuated as CBO has changed its assumptions about how many states will adopt the Medicaid expansion. For more detailed information, see CBO's analysis of the budgetary effects of the Supreme Court decision (CBO 2012c) and CBO's March 2012 baseline (CBO 2012a). Third, projections issued in August 2011 and later incorporate the effects of sequestration under the Budget Control Act, which CBO estimated in May 2013 would reduce Medicare spending by $11 billion in 2020 (CBO 2013a).

CBO itself has cited somewhat larger figures when discussing the extent to which it has revised down its projections in response to slower health care cost growth. For example, CBO recently reported that slower growth in health costs has led it to revise down its estimate of Medicare spending in 2020 by $109 billion since March 2010 (CBO 2014), whereas the comparable figure based on the approach in the text is $87 billion. CBO's figure is larger because it excludes the changes due to updated economic projections discussed above, because it considers a slightly different time period, and because its figure appears to apply to gross, rather than net, Medicare spending. On the other hand, CBO's figure excludes the effect of sequestration, which partially offsets these differences. The estimates presented in the text were chosen over the estimates presented by CBO to simplify exposition and presentation.

¹⁶ Faster growth in health insurance costs could reduce employment through another mechanism. In particular, if workers do not value the additional health spending, then the combination of more expensive health insurance and lower wages could make employment less attractive over time, inducing them to reduce their labor supply. Because evidence suggests that workers' labor supply is only modestly responsive to the returns to work, these effects are likely to be modest in size.

This extraordinary increase in corn output, fall in the real price of corn, and the resulting improvement in physical well-being, did not come about because we are stronger, harder-working, or tougher today than the early settlers who first plowed the prairies. Rather, through a combination of invention, more advanced equipment, and better education, the Iowa farmer today uses more productive strains of corn and sophisticated farming methods to get more output an acre. Today's farmer harnesses more capital equipment, such as advanced planters and combines, to plant more acres, and has the know-how to operate this sophisticated equipment.

Technological advances such as corn hybridization, fertilizer technology, disease resistance, and mechanical planting and harvesting have resulted from decades of research and development. While the government has supported some of this research and its dissemination -- for example, through basic biological research and land-grant universities -- much of this research occurred in the private sector. However, the government has facilitated this private-sector technological innovation by providing the infrastructure to transport and sell increasing quantities of the products and a regulatory and legal environment, such as the U.S. patent system, which clarifies and enforces rights to inventions (more generally, to intellectual property) so that the private sector can reap the rewards of research. These property rights create incentives for innovators, while also allowing others to build on their inventions. The improvements in productivity made possible by technological progress have appeared not just in agriculture, but also throughout the U.S. economy.

The framework of government support for technological innovation is facing new challenges that stem from an ever-changing scientific and legal landscape. Many of these challenges center on the best way to support and encourage development of intellectual property which now encompasses improvements, not just to tractor design, but also technological changes to the software that optimizes its performance. Farmers can now use the Internet to do market research, purchase inputs, make direct sales, and participate in online crop and livestock auctions. Other challenges involve issues surrounding the allocation of the electromagnetic spectrum in a way that supports the efficient development of new wireless and communications technologies that will improve productivity and connectivity -- for the farmer in the combine's cab as well as for millions of other consumers and businesses -- while weighing national security and other concerns. These challenges also include striking the appropriate balance between the need for the government to support fundamental research, which can have large positive externalities that will not be realized by any individual private actor, and the importance of private-sector innovation in driving technology forward.

Another set of challenges relates to how the gains from innovation are shared. In the decades following World War II, productivity improvements translated relatively automatically into compensation increases for families across the income spectrum. But starting in the 1970s, inequality began its relentless rise and productivity growth became increasingly disconnected from compensation growth for typical families. The trends in inequality are related to the trends in productivity, as well as to other broad economic trends. Some of the technological changes over the past three decades, especially those related to information technology, have raised the relative reward to skills obtained through advanced academic study. Thus, the slowing growth of educational attainment both potentially slows innovation and increases inequality by raising the returns to the most highly educated workers. Although expanding the size of markets through globalization can help increase the productivity of the economy, it can also create challenges for inequality.

This chapter begins with a review of the history of productivity growth since World War II, emerging inequality trends, and the government's role in fostering productivity growth. It then focuses on two important current issues in more detail: wired and wireless broadband infrastructure and the efficient allocation of the electromagnetic spectrum; and, new challenges to the U.S. patent system posed by standard-essential patents and patent-assertion entities.

Because of the importance of technological progress in enhancing long-run growth, economists also use another measure of productivity called total factor productivity, or TFP, which proxies for the effect of technological progress. From 1948 to 2012, labor productivity growth in the private nonfarm business sector has averaged 2.2 percent per year, and total factor productivity growth, as measured by the series on multifactor productivity produced by the Bureau of Labor Statistics (BLS), has averaged 1.1 percent per year. This growth of productivity has not been constant, however, and can usefully be thought of as occurring in three episodes: a period of fast productivity growth through the early 1970s, a period of slow productivity growth through the mid-1990s, and a period of somewhat faster productivity growth since then, but still not as fast as in the 1950s and 1960s.

Labor Productivity, Total Factor Productivity, and Multifactor Productivity

The growth rate of labor productivity equals the growth rate of output, minus the growth rate of labor input (worker hours), thus yielding the growth rate of output per worker hour. In contrast, the growth rate of TFP is the growth rate of output, minus the growth rate of output that would be expected solely from the growth rate of the inputs to production. The resulting gap between the actual growth rate of output and the growth rate arising solely because of the growth of inputs is also known as the Solow residual, and is a measure of how well those inputs are combined. Thus, the growth rate of TFP tracks a broadly defined concept of technological change that encompasses scientific innovation and invention, managerial innovations, effects of reorganization of the production process, and other efficiency improvements that do not accrue uniquely to a single measured input.

The concept of total factor productivity is appealing because it estimates the contribution of technological developments to economic growth, and because it can be applied at the level of an industry as well as to the overall economy. In practice, measuring TFP poses several challenges. First, TFP is not observed directly and instead must be estimated using measured inputs and estimates of how the inputs contribute to output. Second, the inputs discussed so far have been capital and labor, but other inputs to production also include, in particular, energy, materials, and business services. Third, for a given level of other inputs, output can increase by hiring better-trained or higher-skilled workers; so for the purpose of measuring TFP, the desired concept of labor input captures changes in both the quantity and quality of labor input. Because labor quality is not observed, proxies such as age and education must be used. Both academics and the U.S. Government have tackled these and other measurement challenges, and have developed estimates of the growth of TFP. This chapter uses an estimate of TFP produced by the Bureau of Labor Statistics called multifactor productivity, or MFP, which is described in Box 5-1.²

Postwar U.S. Productivity Growth

According to the BLS measure of labor productivity shown in Table 5-1, an American worker could produce more than four times as much output per hour in 2012 as in 1948.³ Because MFP takes into account the growth of capital and other factors, labor productivity growth generally exceeds MFP growth. For example, even absent technological change, labor can be more productive simply by using more capital; that is, by increasing the capital-labor ratio or so-called capital deepening. Mathematically, the growth rate of labor productivity is the sum of the MFP growth rate, the contribution of changes in labor quality (as measured by changes in the composition of the workforce), and the contribution of the growth in the amount of capital per worker.⁴ The final column of Table 5-1 gives this decomposition, showing that 10 percent of the growth in labor productivity is due to improvements in the composition of labor (primarily greater educational attainment), 38 percent is due to increases in the amount of capital the worker has at his or her disposal, and 52 percent is due to increases in broad technological progress as measured by MFP.

_____________________________________________________________________

 

 

Table 5-1

 

Sources of Productivity Improvement, Nonfarm Private Business,

 

1948-2012

 

 

[ Editor's Notes: To view the table,

 

see Doc 2014-5547 , page 184.]

 

 

 

As can be seen in Figure 5-1, both labor productivity and MFP are quite volatile from year to year. One reason for this volatility is measurement error in the estimation of both series; indeed, proper measurement of the inputs and outputs is a daunting task and for this reason alone not too much should be read into the growth of productivity in any one year. Another reason is that these series, and the gap between them, varies cyclically. For example, MFP growth fell -- in fact, took on negative values -- during the recessions that started in 1969, 1980-81, 1990, and 2007. These negative values do not mean that, during recessions, firms make negative technological progress or collectively forget about the innovations they have produced over the preceding years. Rather, such declines in MFP could come about from changes in relative prices, so that existing methods of production are no longer the optimal way to combine inputs to produce output. Negative MFP growth can also arise from variation in the utilization rates of capital and labor over the business cycle.

 

Figure 5-1

 

Nonfarm Private Business Productivity Growth, 1949-2012

 

 

 

From the perspective of policies to foster long-term economic growth, these annual and cyclical fluctuations are less relevant than long-term trends in the growth rates of productivity. Figure 5-2 shows a centered 15-year moving average of the growth rates of labor productivity and MFP; and, Table 5-2 summarizes the compound annual growth rates of these series over 10- and 20-year periods ending in 2012, as well as the 60-year period from 1953-2012.

Table 5-2 and Figure 5-2 tell a similar story, which has two parts. First, over the long run the gap between labor productivity growth and MFP growth has fluctuated in a small range, with a difference of between 1.0 and 1.3 percentage points in decadal averages. Moreover, there is no noticeable trend in this gap: the mean difference in the growth rates of these two productivity measures over 2003-12 is within 0.2 percentage point of the mean difference over 1953-62. The stability of the difference between these measures underscores the role of broad technological change -- as measured by MFP -- as a key driver of long-term growth of output per worker.

 

Table 5-2

 

Nonfarm Private Business Productivity Growth

 

 

[ Editor's Notes: To view the table,

 

see Doc 2014-5547 , page 186.]

 

 

 

Because productivity is the key to raising output per person, a great deal of academic research has focused on understanding why productivity growth varies over time. Research points to several factors that contributed to the productivity slowdown of the 1970s. A major culprit seems to be the sharp rise in energy prices during the 1970s that made less energy-intensive technologies more attractive, thus changing the optimal way to combine inputs and reducing MFP growth (Jorgenson 1988, Nordhaus 2004). One lesson learned from this period is how important energy cost fluctuations are in determining the growth of potential output.

 

Figure 5-2

 

15-Year Centered Moving Average of Annual Growth Rates for

 

Labor and Multifactor Productivity, 1956-2005

 

 

 

Another explanation is due to rapid changes in the labor force in the 1970s, primarily shifting the workforce to newer, less-experienced workers. The Baby Boom generation (the cohort born between 1946 and 1964) came of age in the 1970s and 1980s, lowering the overall work experience of the economy. This was a period of rapid entry of women into the workforce for the first time, a shift that also temporarily reduced the overall level of work-force experience in the economy (Feyrer 2007, 2011). Moreover, the rapid entry of these new workers into the workforce outpaced investment, slowing the growth of the capital-labor ratio.

Another possible part of the story is that productivity growth in the 1950s and 1960s was temporarily spurred by large public investments such as the interstate highway system and the commercialization of military innovations from World War II like the jet engine and synthetic rubber.

The productivity rebound of the 1990s and 2000s is widely attributable to the information technology (IT) revolution. For the nine years from 1996 to 2005, MFP grew at 1.6 percent per year, a rate not seen in a nine-year period since the mid-1960s. Although many of the basic technologies that facilitated this growth, like the personal computer and the software to run it, were invented in the 1970s and 1980s; improvements in speed, breadth of applications, and the ability of firms to exploit this technology stretched through the ensuing decades. The BLS MFP measure suggests that much of the productivity improvement resulted from technological and process improvements, a position supported, for example, by Basu, Fernald, Oulton, and Srinivasan (2004). Alternatively, Jorgenson (2001) and Jorgenson and Ho (2012) emphasize the importance of the accumulation of physical IT capital. Oliner, Sichel, and Stiroh (2007) provide a detailed review of the literature on the 1990s productivity boom.

A key current question is what the rate of productivity growth will be going forward -- will the U.S. economy maintain the pace of recent decades, will new innovations accelerate the pace of productivity growth, or will productivity growth revert to the slower rates before the recent boom? MFP growth fell sharply in the recession, grew sharply in the early stages of the recovery, and has averaged 1 percent for 2011 and 2012. These large cyclical swings make it difficult to assess whether there has been a recent change in the rate of technological progress, relative to the late 1990s and early 2000s. The academic literature reaches mixed findings concerning whether the IT productivity boom was temporary.⁵ This literature also requires qualification because it predates the substantial data revisions to historical GDP and productivity that were released in the summer of 2013, which substantially revised upwards estimated productivity growth in some years in the 2000s.

Some contributions to this debate look further into the future. While some economists predict labor productivity growth could decline in coming decades because the scope for future transformative general-purpose inventions is limited (Gordon 2012), others argue that IT is in fact a general-purpose invention and, at least in the medium run, presents an ongoing stream of opportunities for workplace reorganization and efficiency gains, as well as spin-off technologies and improvements.⁶ Bernanke (2012) argued that making these improvements often requires more than just purchasing hardware and software, and realizing potential productivity gains can require changes within and between organizations and thus take a considerable time to be fully realized.⁷

Ultimately, it is very hard to predict future growth rates in innovation, and there is no economic reason that these growth rates should be constant over time. Moreover, the past four decades have seen substantial changes in the extent to which productivity gains translate into higher incomes across the board, the topic of the next section.

Trends in Inequality, Productivity Growth, and Compensation

Real output per hour was 99 percent higher by the end of 1972 than in 1947, while real average hourly earnings (GDP deflator) grew by 73 percent. Figure 5-3 shows that since the early 1970s, the paths of labor productivity and average hourly earnings diverged more widely. As a result, by the end of September 2013 real output per hour was 107 percent higher than at the end of 1972, but average hourly earnings had only grown 31 percent.⁸

Table 5-3 examines the real output per hour and average hourly earnings for private production and nonsupervisory workers by decade. From 1953 to 1962, productivity growth exceeded the average annual rate of change in hourly earnings by only 0.4 percentage point. In the next decade, the difference in growth had ticked up to 0.6 percentage point. However, from 1973 through 2012 labor productivity grew 1.4 percentage points faster than earnings.

Since the 1970s, these trends generally have been worse for lower-income households than for higher-income households (DiNardo, Fortin, Lemieux 1996; Piketty and Saez 2003; Lemieux 2008; CEA 2012; Haskel, Lawrence, Leamer, and Slaughter 2012).⁹ In particular, the income growth in the top percentile of the income distribution has been much stronger than other percentiles. For example, the Congressional Budget Office (CBO 2011) reports that from 1979 to 2007, real before-tax income at the median of the household income distribution increased by about 19 percent, while incomes for the top 1 percent of households have increased by around 200 percent.¹⁰

 

Figure 5-3

 

Growth in Productivity and Average Wage, 1947-2013

 

 

 

Source: Bureau of Labor Statistics, Productivity and Costs, Current Employment Statistics; CEA calculations.

Technological Change and Inequality

The lesson from Figure 5-3 is that productivity growth is important for wage growth, but that does not mean that it automatically leads to wage growth. One possibility is that the sources of labor productivity and MFP growth since the early 1970s are qualitatively different than earlier, and that these different sources of growth drove the trends in inequality over the last 40 years. In the early 1990s, a broad consensus emerged among economists that an increase in the demand for skill relative to the supply of educated labor was the primary driver of the sharp rise in inequality in the 1980s (Bound and Johnson 1995; Katz and Murphy 1992; and Juhn, Murphy, and Pierce 1993). It soon became accepted that "skill-biased technological change" (SBTC) was the most important cause of increased inequality (Berman, Bound, and Griliches 1994; Krueger 1993). The crux of the argument is that, as computer technology became increasingly less expensive, relative demand increased for workers with complementary skills. This explanation has remained popular among economists with few modifications to the basic argument until recently (for example, Acemoglu 2002).

 

Table 5-3

 

Average Annual Rates of Change in the Nonfarm Business Sector

 

 

[ Editor's Notes: To view the table,

 

see Doc 2014-5547 , page 190.]

 

 

 

Goldin and Katz (2008) focus on changes in the growth of the supply of skills rather than on episodic increases in technological change. Using the ratio of college to non-college workers as a measure of the relative supply of skills, they show this relative skill supply grew by 3.9 percent from 1960 to 1980. But in 1980, as confirmed by Heckman and LaFontaine (2010) and others, this increase slowed as high school graduation rates stopped improving and college completion rates slowed. Goldin and Katz (2008) show that a constant increase in the demand for relative skill, combined with the post-1980 slowdown in the supply of relative skill, explains the time path of the logarithm of the college wage premium, which is one measured aspect of wage inequality.¹¹ The nature of rising wage inequality started to change around the early 1990s, becoming increasingly concentrated in the top end of the wage distribution. The ratio of the 90th to the 50th percentile of the wage distribution continued to grow at roughly the same rate it had since the early 1980s, whereas inequality at the bottom (the 50-10 ratio) declined somewhat after the late 1980s. Piketty and Saez (2006) find that income gains have increasingly concentrated in the top 10 percent and top 1 percent since the 1980s. The result has been a "polarization" or a "hollowing out" of the wage distribution, with relative wage growth in the bottom and especially the top of the wage distribution relative to the middle (Goos and Manning 2007; Autor 2010; Acemoglu and Autor 2011; Lemieux 2006).

Autor and coauthors refine the earlier skill-biased technological change literature and argue that the changes in inequality are driven by technological change that substitutes for some tasks but not others (Autor, Levy, and Murnane 2003; Autor, Katz, and Kearney 2006; Acemoglu and Autor 2011). In particular, this new research argues that computer technologies complement non-routine cognitive tasks, which tend to be highly paid; substitute for routine tasks, which tend to be in occupations with wages in the middle of the distribution; and have little effect on manual tasks that tend to be associated with lower wages. This technological explanation for polarization has been controversial, however, and Mishel, Shierholz, and Schmitt (2013) suggest that the theory does not explain the timing of changes in polarization, and more generally that occupational employment and wage trends do not explain a large part of the trends in wages or inequality over time. Moreover, one of the most striking changes in inequality over the past three decades -- the sharp growth of incomes at the very top of the distribution -- is unlikely to be related to technological changes or to a relative demand for skill (Alvaredo, Atkinson, Piketty, and Saez 2013).

This discussion has focused on whether increases in productivity translate into increases in earnings or lead to increasing inequality. A related, less-understood question is whether increasing inequality might directly dampen productivity growth, and this question is addressed further in Box 5-2.

Investments in R&D often have "spillover" effects; that is, a part of the returns to the investment accrue to parties other than the investor. As a result, investments that are worth making for society at large might not be profitable for any one firm, leaving aggregate R&D investment below the socially optimal level (for example, Nelson 1959). This tendency toward underinvestment creates a role for research that is performed or funded by the government as well as by nonprofit organizations such as universities.

These positive spillovers can be particularly large for basic scientific research. Discoveries made through basic research are often of great social value because of their broad applicability, but are of little value to any individual private firm, which would likely have few, if any, profitable applications for them. The empirical analyses of Jones and Williams (1998) and Bloom et al. (2012) suggest that the optimal level of R&D investment is two to four times the actual level. Akcigit et al. (2013) also find underinvestment in basic research (although, contrary to the bulk of the literature, they find overinvestment in applied research), and suggest policies that are specifically targeted at basic research.

Consistent with the presence of large spillover benefits, most basic research in the United States is funded by the government and other nonprofit entities. As Figure 5-4 shows, over half comes from government sources, and less than one-quarter comes from private industry. However, expenditures on basic research are only a fraction of total R&D expenditures, as seen in Figure 5-5, and the private-sector share of funding for applied research and development is much higher than it is for basic research.

 

END OF FOOTNOTE TO BOX 5-2

 

 

_____________________________________________________________________

 

 

Figure 5-4

 

Basic Research Expenditures in the U.S. by Source Funding, 2010

 

 

 

In addition to direct funding of R&D, the government also provides financial incentives for private R&D investment through tax policy. Government can also facilitate private R&D investment and technological progress by providing an institutional, legal, and regulatory framework that clarifies and enforces intellectual property rights and thereby ensures that innovators reap enough financial rewards from their innovations to provide sufficient incentive to engage in a closer-to-optimal level of R&D.¹²

One important type of intellectual property right is patents. A patent grants the inventor a temporary exclusive right over the invention. Exercising that right results in high prices and profits for investments that are successfully commercialized, and those profits provide an incentive to invent. However, the exercise of the exclusive right will also raise prices on inventions that would have been created even with weaker patent protection or with none at all, and these higher prices harm consumers. Moreover, because patented inventions are sometimes used as inputs in creating additional innovations, the higher prices created by patents (as well as the associated legal and administrative burdens, such as negotiating licenses) could slow down subsequent innovation. As discussed further below, a central economic challenge of patent policy is to strike the right balance between providing an economic incentive to invent and the potential harm from the exercise of patent rights. At a minimum, it is important to ensure that patents are not wrongly issued, but rather are only issued for inventions that are non-obvious, useful, and inventive.

 

Figure 5-5

 

Composition of Total R&D Spending

 

 

 

The government can also lay the groundwork for greater creativity and invention by supporting the development of human capital. Investments and improvements in education and training, particularly in the Science, Technology, Engineering, and Mathematics (STEM) fields, foster the innovation workforce of the future.¹³ The productivity of these workers can be enhanced by investment in "innovation clusters," which are dense concentrations of firms and of highly skilled personnel, usually close to a major research university, whose mutual proximity can further promote innovation (see Greenstone, Hornbeck, and Moretti 2008).

Immigration reform is another human capital policy that has the potential to increase the pace of innovation. Studies have found that foreign-nationals living in the United States authored or co-authored over 25 percent of U.S. patent applications in 2006, and that over 75 percent of patents awarded to the top 10 patent-producing American universities in 2011 had at least one foreign-born inventor. Moreover, the innovation benefits of immigration are not confined to the immigration of innovators. Immigration of low-skilled workers, as well as immigration of high-skilled workers who are not innovators, can spur innovation indirectly by increasing specialization. When more non-innovators are present to specialize more completely in their occupations, they enable innovators to specialize more completely in theirs. The Congressional Budget Office (2013) projects that the additional immigration resulting from the Border Security, Economic Opportunity, and Immigration Modernization Act, as passed by the Senate, would raise total factor productivity by roughly 0.7 percent in 2023 and by roughly 1.0 percent in 2033 as a result of increased innovation and task specialization.

Finally, the government has an important role in ensuring that access to the technologies that catalyze productivity growth, and to the technologies and products that are the fruits of that productivity growth, are broadly available throughout American society. Sharing these benefits increases welfare directly, and also ensures that the broad population maintains the technological skills needed in the workplace and for the education of current and future generations.

This chapter now turns from a general discussion of the role of government policy in achieving technological progress to a focus on two key current areas that are important for productivity growth and that are also a focus of the Administration's policies: telecommunications and patent reform.

Innovation and Investment

The telecommunications sector is a major success story in the U.S. economy. A recent White House (2013) report, Four Years of Broadband Growth, documents many of the striking facts, including:

• Just two of the largest U.S. telecommunications companies account for greater combined stateside investment than the top five oil/gas companies, and nearly four times more than the big three auto companies combined, as seen in Figure 5-6.

• Between 2009 and 2012, annual investment in U.S. wireless networks grew more than 40 percent, from $21 billion to $30 billion. During that period, investment in European wireless networks remained flat, and wireless investment in Asia (including China) rose only 4 percent. The report projected that U.S. wireless network investment would increase further in 2013, to $35 billion.

• The United States leads the world in the availability of advanced 4G wireless broadband Internet services such as LTE; nearly half of the global subscriber base for 4G LTE is in the United States.

• The United States ranks among the top countries in the world in the amount of currently licensed spectrum available for mobile broadband.

Four Key Areas for Telecommunications Policy

The U.S. Government can support innovation and investment in tele-communications through the same general policies discussed above: direct government investment in research and development; catalyzing private innovation through policies such as reforming and extending the Research and Experimentation Tax Credit; catalyzing technological infrastructure investment in areas like broadband; and ensuring that everyone benefits from broadband technologies.

Government Investments in Research and Development. As discussed above, spillover benefits to research and development, especially for basic science and technology, creates a role for direct government investment. Perhaps the most famous government investment in telecommunications technology was the Defense Advanced Research Projects Agency (DARPA) development of the Internet. But DARPA has provided other important defense-based public research contributions as well. These contributions include the radio and, more recently, Global Positioning Systems, which today are central to a huge number of consumer applications.

 

Figure 5-6

 

Relative Investment of the Telecommunications Sector, 2011

 

 

 

Today, the Department of Defense (DOD) continues to play an important role in telecommunications research, particularly in helping to develop ideas and technologies for sharing of electromagnetic spectrum frequency bands between different users, including between government and private users. This, as discussed further below, has been identified as important for efficient spectrum management in the future (PCAST 2012). For example, DOD has solicited innovative research proposals aimed at efficient and reliable sharing of spectrum between radar and communications systems. All told, $100 million in Federal investments are being targeted toward spectrum sharing and advanced communications through the National Science Foundation (NSF), DARPA, and the Commerce Department.

Catalyzing Private Investment. Reforming, expanding, and making permanent the Research and Experimentation Tax Credit would increase investment in telecommunications technology, accelerating innovation. Immigration reform would accelerate innovation as well. Reforming the patent system is also important in this industry, especially for technology deployed in smartphones, which are complex devices that embody thousands of patents. The increasing frequency of patent disputes in this area suggests that there may be increasing costs to navigating the appropriate licenses. If these costs are high enough to adversely affect the introduction of new products, then patent reform is particularly important for the telecommunications industry.

Public policy encouraged these investments. Many private carriers, as regulated monopolies, were permitted to charge high rates for long-distance calls, business service, and the telephones themselves. A portion of the resulting funds were required to be used to subsidize basic local phone service, particularly in rural and other areas that are costly to serve due to low population density and geographic factors. The Telecommunications Act of 1996 sought to reform and improve upon telecommunications regulation by enabling greater competition, particularly in local and long-distance telephone service, and by rationalizing, and making explicit, the subsidy system supporting service in high-cost areas. Substantial additional private investment followed.

In recent years, the U.S. Government has further facilitated private telecommunications investment through favorable tax policy. In 2010, the President proposed and signed into law the largest temporary investment incentive in history -- 100-percent expensing -- that, together with the bonus depreciation that preceded and followed it, played a critical role in increasing and accelerating investment, including the substantial increases in both wired and wireless investment in the telecommunications sector. For example, two major companies in a joint statement said that, "despite the downturn in the economy, the cable communications sector has been able to continue steady investment and to retain jobs as a result of policies like 100-percent expensing."

Catalyzing investment in mobile broadband infrastructure is especially important given the rapidly growing usage and the scarcity of electromagnetic spectrum for carrying wireless broadband traffic. In 2010, Federal Communication Commission experts predicted that the needs for broadband capacity will overwhelm available spectrum (the "spectrum crunch"). If allowed to happen, this would result in higher prices for mobile broadband services, as well as reduced growth in broadband-based innovation, services, and employment. The scarcity of broadband capacity can be alleviated through increased investment (denser transmission infrastructure means more traffic on a given spectrum frequency), fuller deployment of spectrum already licensed to wireless carriers, spectrum license consolidation, technological advancement, and improvements in spectrum policy.

One important initiative is to seek to reallocate public spectrum when it has a more valuable private use. The Federal Government is a major user of spectrum, as Figure 5-7 shows. Most of this usage involves national security and law enforcement functions, as shown in Figure 5-8. Federal use of spectrum is valuable, but it is not costless. As an economic matter, if a particular spectrum band would produce a larger net social surplus in private hands than in public hands, then it should be reallocated, and vice-versa. That is, the Federal Government can alleviate spectrum scarcity by having government agencies vacate certain spectrum bands entirely, or share them with private users, when this can be achieved without compromising the agencies' vital missions (which in many cases involve safety-of-life and national security) and when the associated costs of relocating government operations out of those bands are justified by the social value that will be unlocked as a result of the reallocation to the private sector. The vacated spectrum could then be auctioned off to commercial users or, if appropriate, made widely available on an unlicensed basis (more on this below). The Federal user would be relocated to alternative spectrum that could be used more intensively and economically, particularly if additional resources were made available for investment in newer equipment.

 

Figure 5-7

 

Exclusive and Shared Allocation of Radio Spectrum

 

 

 

In addition to economizing on its aggregate spectrum usage, the government can further alleviate spectrum scarcity by rationalizing spectrum allocation. There are some spectrum bands that, above and beyond the properties that make them valuable in general (for example, strong propagation through buildings and in rural areas), are particularly valuable for commercial applications, such as if they are complementary to other commercial spectrum bands. In those cases, value can be unlocked by having the government relocate from those bands to other bands that do not have that property -- again, under the condition that this can be done without compromising vital missions and that the relocation costs are not prohibitively high.

Box 5-4 describes several spectrum investment policies that have been undertaken or proposed by the Administration.

There is also substantial scope to reallocate some spectrum currently licensed to private entities to a more valuable use in wireless broadband. Some incumbent firms, such as over-the-air broadcast television stations, hold rights to spectrum that are much more valuable as wireless broadband spectrum. The 2010 National Broadband Plan introduced the idea of "incentive auctions" as a tool to help meet the nation's spectrum needs by giving those rights-holders a share of the auction proceeds if they relinquish their rights. In the Spectrum Act of 2012, Congress authorized the Federal Communications Commission to conduct incentive auctions and directed that the FCC use this innovative tool for an incentive auction of broadcast television spectrum. In September 2012, the FCC adopted a Notice of Proposed Rulemaking in order to develop a rulemaking record that will enable the Commission to meet the challenges presented by the Spectrum Act's unique grant of authority. The magnitude of potential gains to social surplus are enormous when broadcasters with access to new, more-efficient transmission technologies that use less spectrum, or with a small and shrinking base of over-the-air viewers and annual revenue in the low millions of dollars, will have an incentive to relinquish spectrum that, when reconfigured for commercial broadband use, will be sold for hundreds of millions of dollars to companies that will use it to improve services for a vastly greater number of broadband customers.

 

Figure 5-8

 

Federal Agencies with Most Spectrum Assignments

 

 

 

Source: National Telecommunications and Information Adminstration, Government Master file (2010); Government Accountability Office Analysis.

Some spectrum can be used effectively without being licensed at all, but rather made available for anyone to use on an unlicensed basis. Just as some roads seldom experience traffic jams, in some instances certain spectrum bands do not become highly congested even when access is free. Unlicensed spectrum plays an important role in the broadband ecosystem, enabling Wi-Fi, Bluetooth, "smart homes," and more, which operate on unlicensed spectrum using devices whose power is low enough that interference among numerous devices sharing the spectrum is not a major concern. It also helps to alleviate scarcity in licensed spectrum bands. This is because a great deal of mobile usage is not the "on-the-go/in transit" mobile usage that must be transmitted on a carrier's licensed mobile network, but rather is so-called "nomadic" usage (for example, at home, office, or other fixed location), that is amenable to carriage mostly by a wired broadband connection and then wirelessly completed using a nearby unlicensed Wi-Fi router. For this reason, the licensed carriers are investing heavily in the deployment and use of Wi-Fi networks. The value of this unlicensed spectrum has been estimated at $16 billion to $37 billion per year.

Clearing Federal Government spectrum for exclusive licensed use, and making it available for shared unlicensed use, remain viable solutions in the near term. However, given the dramatic spectrum challenge and the fact that much of the lowest-hanging fruit for reallocation has already been picked, it is also important to focus on newer and more innovative ideas. These ideas include new advances in the sharing of spectrum between different users, particularly between government and private users. Innovation in spectrum sharing is both promising and necessary, as there are some spectrum bands that the government cannot vacate entirely, but that nevertheless have unused capacity, and that with appropriate processes and procedures in place could be shared, accommodating some valuable private usage without compromising mission-critical functions.

The President's Council of Advisors on Science and Technology (PCAST) released a report estimating that "in the best circumstances, the amount of effective capacity that can be obtained from a given band of spectrum can be increased thousands of times over current usage through dynamic sharing techniques that make optimal use of frequency, geography, time and certain other physical properties of the specific new radio systems (PCAST 2012)."

The 2010 Presidential Memorandum that set the Administration's spectrum goal contemplated the sharing of Federal Government spectrum as one means of achieving that goal. More recently, in June 2013, another Presidential Memorandum established a Spectrum Policy Team in the Executive Office of the President, which was charged with the mandate to "monitor and support advances in spectrum sharing policies and technologies." That Memorandum also contains measures to facilitate research, development, testing, and evaluation of technologies to enhance spectrum sharing and other spectrum-related efficiencies.

To stimulate investment in more advanced forms of spectrum sharing, the Defense Advanced Research Projects Agency (DARPA) is soliciting innovative research proposals aimed at efficient and reliable sharing of spectrum between radar and communications systems. Consistent with its history of promoting groundbreaking technological breakthroughs for both military and commercial use, DARPA is seeking "innovative approaches that enable revolutionary advances" in spectrum sharing, specifically in the spectrum bands that are most amenable to broadband and communications services. The program may fund multi-year projects designed either to significantly modify existing radar and communications systems or to unveil new system architectures redesigned from the ground up.

By itself, making additional Federal spectrum available for commercial use, whether on an exclusive or a shared basis, is unlikely to be sufficient to keep up with the exploding demand for bandwidth. The ambitious goal of freeing up 500 MHz of spectrum would nearly double the amount of wireless spectrum available for mobile broadband over the course of a decade, but even that may not be enough to keep up with spectrum usage growth. Therefore, it is important to do everything from increasing investments in wired broadband networks that can offload some of the demand (often making the last connection wireless, but through Wi-Fi rather than cellular), to increasing the density of wireless cells, to encouraging technological innovations for using spectrum more efficiently.

The Administration is trying to help with these efforts in a variety of other ways, including the June 2012 Executive Order issued by the President specifying a number of steps that will ease and facilitate carriers' access to Federal land and buildings for purposes of deploying broadband infrastructure, including cell towers.

Ensuring Everyone Benefits. It is important to ensure broad participation in the benefits of broadband telecommunications technologies, because broad participation allows more people to use those benefits to develop their talents, which lead to higher economic growth and higher living standards in the future. One element of broad participation is ensuring that technology and its products are affordable. To that end, vigorous antitrust enforcement is critical to ensure that that prices are not inflated and choices not limited by lack of competition. This has been a focus of the law enforcement agencies, and is also important as a policy consideration going forward.

The Obama Administration has made critical investments in expanding broadband to underserved communities. The American Recovery and Reinvestment Act of 2009 included $6.9 billion in funding to upgrade the nation's broadband infrastructure, with $4.4 billion administered by the Department of Commerce's National Telecommunications and Information Administration, and $2.5 billion by the Department of Agriculture's Rural Utilities Service. Of these funds, a total of $4.4 billion (as of the end of May 2013) went to fund more than more than 325 broadband projects through the Broadband Technology Opportunities Program and the Broadband Initiatives Program. The Federal Communications Commission has also played an important role in expanding broadband deployment in unserved and underserved areas through Universal Service Reform and the establishment of a $4.5 billion annual Connect America Fund, which reallocates funds previously used to support voice service.

Education researchers have long believed that technology holds the potential to profoundly impact the classroom experience, from allowing students to interact with course content in new and personalized ways to helping teachers understand what lessons and techniques are most effective. By making the ever-expanding collection of educational resources available on the Internet accessible to teachers and students in classrooms, technologically equipped schools enhance learning by gaining access to those resources, rather than being limited to resources that are physically at hand.

Although more high quality research on the effectiveness of online educational tools is still needed, these tools do show promise. A meta-analysis of experimental or quasi-experimental studies of the effects of online education conducted by the Department of Education in 2010 found that students who receive instruction that combines online and face-to-face elements performed better than students who received either exclusively online, or exclusively face-to-face, instruction. Other factors such as instruction time or curriculum may contribute to this positive effect, but the meta-analysis suggests that further research on designing, implementing, and evaluating these blended approaches may be worthwhile.

Instruction methods that incorporate computers have also shown promise in mathematics education. Barrow, Markman, and Rouse (2009) found that students who were randomly assigned to participate in and complete computerized math lessons at their own pace scored 0.17 to 0.25 standard deviations higher on mathematics achievement tests than students who received traditional instruction. Computer-aided mathematics instruction has been shown to have similar effects in other contexts. In an experimental study, Banerjee et al. (2007) find that playing educational math games on computers for two hours a week improved the math scores of impoverished elementary school students in India by 0.47 standard deviations. In another experiment, Carillo, Onofa, and Ponce (2010) find poor Ecuadoran elementary school students who used adaptive math and language software for three hours a week improved their math scores by 0.30 standard deviations.

Because current uses of technology can enhance learning and the potential of future developments is untold, it is critically important that all students have access to 21st century classrooms. The ConnectED program, announced by President Obama in June 2013, takes important steps to ensure that the benefits of improvements in educational technology will be made widely available. While initiatives like the FCC's E-Rate program, established under the Telecommunications Act of 1996, have helped bring Internet access to almost every school in the nation, many schools do not have access to the fast broadband speeds enjoyed by most businesses and households. Further, the E-Rate program was designed primarily to bring Internet access to the school, with less priority on ensuring that access was available throughout the school, such as via Wi-Fi technology. As a result, 62 percent of school districts say their bandwidth needs will outstrip their connections within the next 12 months, and 99 percent say that this will happen within three years.

ConnectED will bring high-speed broadband and wireless Internet access to 99 percent of America's students in their school classrooms and libraries within five years. To make the most of this enhanced connectivity, ConnectED will refocus existing professional development funds to train teachers to take full advantage of these resources in order to improve student learning. Finally, by equipping schools with the broadband Internet access necessary to make use of high-tech educational devices, ConnectED will deepen the market for such devices, as well as the digital educational content with which they interact, spurring private-sector innovation in this area.

The President has called on the FCC to modernize the E-Rate program, and has also called on the expertise of the NTIA, in order to deliver this connectivity and meet the goal of connecting 99 percent of America's students to the digital age within five years through next-generation broadband and high-speed wireless in their schools and libraries. Answering that call, the FCC announced in February 2014 that it would invest $2 billion to connect 20 million students over the next two years, representing a crucial down-payment on reaching the President's goal. The initiative, however, is not just about infrastructure. The President announced in February over $750 million in private sector commitments to help fill out this vision of a connected classroom through the digital devices, content and learning software, home wireless access, and teacher training necessary to make the best possible use of this infrastructure. By leveraging all these resources, we are making substantial progress toward a world-class education for every student that does not depend on their family's income or on the zip code in which they were born.

 

Figure 5-9

 

Percentage of Households with Access to Download Speeds

 

of 6 Megabit per Second or Greater

 

 

 

Finally, it is crucial that the benefits of broadband technology growth be consistent with privacy and security. Also, the free expression of ideas must be protected, so technological development must proceed in a way that is consistent with an open Internet.

Challenges to Broad Adoption of Telecommunications Technology

Broad adoption of telecommunications technologies faces several challenges. For example, these technologies are unevenly adopted across different education and income levels. Home broadband adoption is more than twice as high for college graduates as for high school dropouts. Overall, 30 percent of Americans do not use broadband at home, and many of these non-users are in lower-income households. Rural areas also lag in adoption. As illustrated in Figure 5-9, nearly all urban residents have access to 6 megabits per second downloads, but only 82 percent of residents in rural communities can access those speeds, and the disparity becomes even larger at faster speeds.

One reason some households do not adopt broadband is cost: unlike the sharp price declines seen for technological hardware, such as computers, the prices consumers pay for Internet access have remained steady or risen. But while broadband prices have not fallen sharply, the speeds that are available at a given price today are often considerably faster than the available speeds at the same price several years ago, which means that value for money has improved. Further, while international comparisons are difficult (due to variations in factors like taxes, government subsidies, geography, population density, and product bundles), the United States compares favorably in a number of respects, including entry-level pricing for slower but still useful broadband speeds.

While this chapter has focused heavily on telecommunications technology, there are many other areas where technological advances promise large social and economic benefits, and where public policy can play an important role. One important example, discussed in Box 5-5, is Electronic Health Records and related technologies.

Patents

The rights that prospective innovators have to the economic returns on their innovations are known as intellectual property (IP) rights, of which one major category is patents, which apply to inventions. Patents are granted on inventions in many different areas of technology, as shown in Figure 5-10. The basic economic logic behind patent protection is simple: successful inventions are valuable to society, as they lead to new and better products. But attempting to invent is costly and risky. If successful inventions could be easily imitated by competitors, then prospective inventors may be in a position where they lose if their invention fails, but gain little or nothing if it succeeds. This diminishes the incentive to expend resources and effort on inventing, and the reduced rate of invention is harmful to society. To prevent this problem, patent protection allows inventors to enjoy a temporary exclusive right to their invention. The super-competitive pricing that results from this exclusivity provides an incentive to invest. Another benefit of patent protection is that patents are published, so they can be licensed and put to other socially valuable purposes other than those of the inventor. But patent protection can also harm consumers: for inventions that would have been created with weaker patent protection or even with no protection at all, patents simply lead to higher prices for the same inventions, not to additional inventions. The economically optimal strength of patent protection (for example, how many years a patent should run) is the one that best balances the benefit from accelerated invention with the harm from higher prices.

There are some additional effects of patent protection that also deserve mention. One effect is that some inventions are complementary to each other, meaning that the availability of one makes it easier to develop others. In those cases, the higher prices resulting from patent protection, as well as the related legal and administrative burdens (such as negotiating licenses), raise the cost of, and hence reduce the rate of, subsequent innovation. This effect is relevant for determining the economically optimal patent strength. Another effect is that, in some cases, patent rights can be used to harm rival firms or to extract license fees that do not correspond to the value of the patent. As discussed below, it is important to curb such behaviors by developing sound policies related to patent examination and enforcement. It is also important to ensure that patents are not wrongly issued, but rather are only issued for inventions that are non-obvious, useful, and inventive.

 

Figure 5-10

 

Patents Issued in the U.S. by Technological Category, 2003-2012

 

 

 

Source: United States Patent and Trademark Office.

The chapter now turns to two specific patent issues that have been the subject of recent policy scrutiny: how to support standard setting by appropriate use of standard-essential patents, and the activities of Patent Assertion Entities and the effects of those activities.

Standard-Essential Patents

We take for granted that we can drive our car up to a gas pump and have the hose fit the car's nozzle. Similarly, that smartphones created by different manufacturers will communicate with each other. These are examples of interoperability that result from the standardization of certain product features. An interesting problem arises when an industry seeks to adopt an interoperability standard and the available choices for the standard may include patented inventions.

The nature of the economic problem is to develop a mechanism that determines when standardization would make market participants better off and, in such cases, provides parties with incentives to invent and propose the invention as a standard, while ensuring that all parties will later find it in their interest to implement the standard. The central premise of the economic theory of patents is that granting limited exclusive rights through the issuing of patents provides an incentive for private investment in invention; absent such rights, the entity making the investment may not receive sufficient returns to make the investment worthwhile. These exclusive rights are not meant to preclude similar technologies from being developed and marketed. In principle, some degree of competition in consumer markets bounds the power conferred by these exclusive rights. But that bound is removed if a patented technology becomes the standard and is used in all products sold in the market. As a result, the patent holder may seek to charge higher prices than originally agreed on during the standard-setting process and to use the patent to inefficiently restrict access to the technology. Such behavior may delay implementation of the technology, as others who may adapt the technology exit the market or seek other ways around it. The sought-after standard then fails to become standard (for example, Gilbert 2011).

Because industry actors are most likely to understand the substantial complexities of new technologies and the potential products and markets for their dissemination, there is value in having the standards set voluntarily by industry-based standards-developing organizations (DOJ 2013). These organizations provide a place for industry actors to propose their patented technologies as part of a standard, and to reach consensus on the technologies incorporated into the standard (or to decide on no standard). After a decision is taken, a chosen patent becomes known as standard essential. Actual implementation of the agreed-on standard as an observed standard follows when all implementers and potential implementers pay an agreed justified price (reasonable, or both fair and reasonable) for the technology, and their access to the patented technology cannot be improperly restricted (there is not discrimination). By proposing a patent for use in the standard, the patent owner is giving up the power to charge higher per-unit prices for use of the technology, but enjoys returns from the diffusion of the technology more widely across more units.

Because the notion behind standard-developing organizations is voluntary collaboration, there is no guarantee that a standard will be produced. A standard-essential patent holder can refrain from committing to licensing on reasonable and non-discriminatory (RAND) terms.¹⁴ In such cases, the declared standard is less likely to be the implemented standard and market forces may be suggesting that a standard is not needed or may be best determined over time and in the marketplace directly (Farrell et al. 2007).

When voluntary agreement does produce a standard, there are instances when parties to the agreement do not feel that others are living up to the agreement. In such instances, when patent holders have committed to license on RAND terms, judicial and enforcement procedures should aim to reproduce the intent of the agreement; that is, to ensure that the patent holder receives a RAND royalty. Otherwise, judicial and enforcement procedures can tip the balance of power in favor of one party or the other, leading either to excessive market power in the hands of the patent holder or to non-payment of reasonable royalties by implementers, and to greater incentives against establishing a standard in the first place (Lemley and Shapiro 2005).

Patent Assertion Entities

In recent years, organizations known as Patent Assertion Entities (PAEs) have become common. PAEs brought 24 percent of all patent lawsuits in 2011, and over the 2007-11 period they brought approximately one-fifth of all patent lawsuits, covering about one-third of all defendants (GAO 2013). These PAEs purchase rights to patents belonging to other firms, and then assert them against firms or individuals who are using the patented technology. Some of this activity is valuable: incentives to invent are stronger if inventors know they can later sell their patent to, or merely engage the services of, a PAE that can assert it more effectively than they could do themselves. Also, in some cases, it may be efficient for PAEs to act as intermediaries by obtaining the rights to patents held by disparate inventors in order to decrease the transaction cost of negotiating licenses. However, many industry observers believe that PAEs often do not assert patents in good faith, but rather assert them simply in order to extract nuisance payments from firms looking to avoid costly and risky litigation. In some cases, these patents are valid but of low value, meaning that absent the high cost of litigation they would only command very low licensing fees. In other cases, the patents are invalid (or not infringed), and absent the high litigation costs they would not command any license fees at all (Scott Morton & Shapiro 2013).

This issue is particularly pronounced in smartphones and other consumer electronics devices (Chien 2012). Many of these products contain technology based on thousands of patents, and as shown in Figure 5-10 above, the number of patents issued in the "Electrical and Electronics" category has been increasing over the past decade. The large number of patents embodied in these products, and their complexity, often makes manufacturers subject to patent-infringement claims, with the associated threat of costly and risky litigation, from owners of low-value valid patents or even from owners of invalid patents. It is therefore an important public policy goal to find ways to reduce the cost of defending patent lawsuits, and also to reduce the number of invalid patents, either by reducing the number of invalid patents that are granted, or by making it easier for them to be challenged.

Several provisions of the America Invents Act may help address some of the problematic behavior of PAEs by developing at the Patent and Trademark Office new programs to create alternatives to litigation over patent validity, new methods for postgrant review of issued patents, and major steps to increase patent quality through clarifying and tightening standards. Yet challenges remain, notably the asymmetry between the cost to a PAE of bringing a patent lawsuit and the cost to a target firm of defending one, which enables PAEs to bring weak cases in the hopes of extracting a settlement.

In June 2013, the President issued a set of five executive actions and seven legislative recommendations to address these challenges. These included measures to make it more difficult for overly broad claims to receive patents in the first place, as well as to make it easier to challenge weak patents once they have been granted. The President's priorities also include measures to require greater clarity in patent applications regarding the precise nature of the claimed invention, as well as the identity of the patent holder. Other measures include ways to make it more difficult for patent holders to sue end-users (as opposed to manufacturers) of products that contain patented technology, and to provide judges with more discretion to award attorney fees and other costs to the prevailing party in patent lawsuits.

Congress has taken up these issues. In December 2013, the House of Representatives passed a bipartisan bill containing many of the Administration's priority items. A related bill is currently under consideration in the Senate Judiciary Committee.

Another important policy issue related to patents is the phenomenon of "pay-for-delay" settlements of patent lawsuits in the pharmaceutical industry. This is discussed in Box 5-7 below.

² One of the many other challenges in estimating total factor productivity is that the intensity of utilization of inputs varies over the business cycle. For example, because hiring and training workers is expensive, firms might retain some workers in a mild downturn, so that fluctuations in output are greater than fluctuations in employment (a relationship which, when recast in terms of the unemployment rate, is known as Okun's Law). The BLS MFP series does not adjust for changes in factor utilization, which can produce cyclical fluctuations in MFP. Basu, Fernald, and Kimball (2006) provide an approach to adjusting for such cyclical variation, and a quarterly TFP series produced using their method is currently maintained by the Federal Reserve Bank of San Francisco (Fernald 2012).

³ This discussion of postwar productivity performance cites statistics for nonfarm private businesses. Recall the earlier discussion about how productivity growth in farming allowed fewer resources to be devoted to it. By 1947, farming accounted for less than a nine percent share of GDP. Today that share is about one percent.

⁴ Suppose aggregate production can be represented by the Cobb-Douglas production function, where Y is real output, L is labor input measured in labor-quality units, K is capital, and A summarizes the contribution of technology to production, that is, A is TFP, and is a constant. Then output per worker-hour (H) is Thus the annual growth of output per worker, that is, the growth of labor productivity, is the sum of the growth of A, that is, the growth of TFP, plus times the growth of L/H, that is the growth of labor quality per worker-hour, plus times the growth of K/H, that is, the growth of the capital-labor ratio. By using Tornqvist aggregation, the BLS MFP measure allows shares to change over time and does not require an aggregate Cobb-Douglas production function.

⁵ The findings in the literature on recent productivity growth trends tend to depend on the statistical approaches used to discern different productivity regimes. Authors that adopt discrete breaks or regime shifts, including Kahn and Rich (2011) and Fernald (2012), tend to conclude that the productivity growth boom has passed, whereas Oliner, Sichel, and Stiroh (2007), who use methods in which productivity growth evolves more slowly, find less of a slowdown.

⁶ An example of such now-possible workplace reorganization is telecommuting; see, for example, Bloom, Liang, Roberts and Ying (2013); Noonan and Glass (2012), Bailey and Kurland (2002); and Busch, Nash, and Bell (2011).

⁷ These issues are argued in the February 2013 TED debate between Gordon and Brynjolfsson.

⁸ An alternative series from BLS measures real total hourly compensation (CPI deflator) for all nonfarm workers. This measure includes benefits as well as earnings. Since 1972, total hourly compensation has increased more than hourly earnings, but still only by 46 percent. BLS decompositions of compensation into real wage and benefit shares have been available since 1991. Since then, real wages grew 7 percent and benefits grew 22 percent, with the strongest benefit growth in the magnitude of employer contribution to health insurance.

⁹ Figure 6-2 of this report suggests that the relative slow growth in income of the lower quintiles may have subsided some recently, particularly during the Great Recession and its near-term recovery. It is too soon to tell whether this has any implication for longer-term trends.

¹⁰ The CBO notes that it chose 1979 and 2007 as points of comparison because there are cyclical fluctuations in inequality measures and both years are business cycle peaks.

¹¹ This theory is based on evidence from before 2008. The U.S. economy has long had some skills shortage, which tended to turn up in the form of wage differentials rather than unemployment. It does not account for the large shock in aggregate demand that characterized the Great Recession, or the shock-driven unemployment rates from which the economy is still recovering.

¹² Research in development economics suggests that a key factor in the economic performance of a country is its "institutions," such as rule of law and clear property rights (Hall and Jones 1999, Rodrik et al., 2004).

¹³ As discussed in Delgado et al. (2012), one determinant of a country's economic performance is its science and innovation infrastructure. The authors include in this category a number of elements that can be influenced by supportive government policy, such as the quality of scientific research institutions and the quality of math and science education.

¹⁴ Sometimes the licensing commitment is to fair, reasonable and non-discriminatory terms.

The War on Poverty ushered in a new era of Federal Government leadership in providing income and nutrition support, access to education, skills training, health insurance and a myriad of other services to low-income Americans. During President Johnson's term, Congress passed more than a dozen major pieces of legislation that provided such foundational elements of our current social welfare system as the Civil Rights Act, the Economic Opportunity Act, the Food Stamp Act, Elementary and Secondary Education Act, the Manpower Act, Medicare, Medicaid, the Higher Education Act and the Child Nutrition Act. Since then, many of these programs have been reformed and updated, ensuring that the modern safety net assists families when they need it most, while also keeping them connected to the labor force.

Optimism was high at the outset of the War on Poverty. One of its architects predicted poverty would be eradicated within a generation and saw "on the horizon a society of abundance, free of much of the misery and degradation that have been the age-old fate of man (Council of Economic Advisers 1964)." The 50th anniversary of President Johnson's bold declaration provides an opportunity to assess our achievements in reducing poverty and to evaluate the record of the poverty-fighting programs created or enhanced in the wake of his declaration. There is no question that the material conditions of those in poverty have improved: the percentage of the poor with indoor plumbing has risen from 58 percent in 1960 to 99 percent in 2011;¹ infant mortality in counties with the highest levels of deprivation has fallen from 23.2 per 1,000 in 1969 to 9.1 per 1,000 in 2000;² and today all American children in poverty have access to affordable health insurance, as do poor adults in states that have taken up the Affordable Care Act's Medicaid expansion. But to what extent have we reduced the proportion of Americans living in poverty? Indeed, the poverty rate has declined considerably since the beginning of the War on Poverty, but how much of this improvement is due to the efforts of government? And what are the lessons this contains for future policy?

This chapter answers these questions by first confronting the challenges in measuring poverty, and highlighting the limitations of the official poverty measure for tracking progress in the War on Poverty. Using new historical estimates of poverty based on modern measurement methods, this chapter presents a more accurate picture of the changes in poverty over the past five decades, and estimates the contribution of the safety net to these changes. While the Council of Economic Advisers (CEA) reviewed research on the effects of antipoverty programs on work and earnings, health, food security, educational attainment, and other valued outcomes, this chapter focuses primarily on their impacts on poverty and economic mobility.³ Finally, it discusses the role that President Obama has played in reducing material hardship among low-income Americans by expanding access to affordable health insurance and tax credits for working families, and by his proposals to help ensure that no parent who works full time will have to raise his or her children in poverty.

Measuring Poverty

Measuring poverty is not a simple task; even defining it is controversial. Just starting with a commonsense definition of the poor -- "those whose basic needs exceed their means to satisfy them" -- requires difficult conceptual choices regarding what constitutes basic needs and what resources should be counted in figuring a family's means. There are no generally accepted standards of minimum needs for most necessary consumption items such as housing, clothing, and transportation. Moreover, our ideas about minimum needs may change over time. For example, even some middle-income households did not have hot and cold running water indoors in 1963. Today, over 99 percent of all households have complete indoor plumbing.

The Official Poverty Measure

Mollie Orshansky, an economist in the Social Security Administration, developed the official poverty thresholds between 1963 and 1964 (Fisher 1992). At the time, the U.S. Department of Agriculture had a set of food plans derived using data from the 1955 Household Food Consumption Survey, the lowest cost of which was deemed adequate for "temporary or emergency use when funds are low." Because families in this survey spent about one-third of their incomes, on average, on food, Orshansky set the poverty threshold at three times the dollar cost of this "economy food plan," with adjustments for family size, composition, and whether the family lived on a farm.

These income thresholds that were first used as the poverty thresholds for the 1963 calendar year have served as the basis for the official poverty thresholds ever since. These dollar amounts have been adjusted for inflation to hold the real value of the income needed to be above poverty the same over time. There have been minor tweaks to the methodology involving which price index is used to adjust for inflation, and how adjustments are made for family structure and farm status.

The Supplemental Poverty Measure

While Orshansky's measure provided a reasonable depiction of poverty in the 1960s, it has not aged well.⁵ Today, for example, the value of the two largest non-health programs directing aid to the poor -- the Earned Income Tax Credit (EITC) and Supplemental Nutrition Assistance Program (SNAP) -- are entirely ignored by the official measure, making it impossible to assess the success of these tools in fighting poverty. Over the past five decades, researchers have pointed to many flaws in the official measure (Box 6-1), leading to the development of alternative measures of poverty with more comprehensive measures of both family needs and resources. The Census Bureau created a Supplemental Poverty Measure (SPM), which departs dramatically from the official measure in its methodology for calculating both the poverty thresholds and family resources.⁶ This measure, first published in 2011, calculates poverty thresholds using recent expenditures by families at the 33rd percentile of the expenditure distribution on an array of necessary items, including food, shelter, clothing, and utilities.⁷ The dollar amount is calculated separately for families depending on whether they own or rent their home and whether they have a mortgage, and then increased by 20 percent to allow for other necessary expenses. Further adjustments are made based on differences in family size and structure, and, unlike in the official measure, the threshold is adjusted for geographic variation in living costs (Short 2013).

_____________________________________________________________________

 

 

Trends in Consumption and SPM Poverty, 1961-2012

 

 

 

The Supplemental Poverty Measure also uses a more accurate measure of disposable income that accounts for both a greater number of income sources and a wider array of necessary expenditures. Unlike the official measure, the SPM uses a post-tax, post-transfer concept of resources that adds to family earnings all cash transfers and the cash-equivalent of in-kind transfers such as food assistance (for example, SNAP or free lunch) minus net tax liabilities, which can be negative for families receiving refundable tax credits like the EITC or CTC. Necessary expenditures on work and child-care are then subtracted from resources.

The Supplemental Poverty Measure also subtracts medical-out-of-pocket expenses from families' resources since those funds are not available to meet other needs. The SPM can thus be thought of as a measure of deprivation with respect to non-health care goods and services.⁸ However, it does not provide an accurate picture of the benefits of health care. Instead, the SPM values health insurance only insofar as it reduces households' out-of-pocket medical costs and thus frees up resources for other uses. It misses benefits that may arise because insurance improves access to health care and may therefore improve health outcomes, or reduces stress caused by exposure to financial risk. As a result, the measured trend in SPM poverty may understate progress in decreasing economic hardship since the War on Poverty began by ignoring these benefits of increased access to insurance.

One important feature of the Supplemental Poverty Measure design is that the definition of minimum needs is adjusted each year based on recent data on family expenditures on necessities rather than adjusting a fixed bundle only for inflation. By considering families' expenditures on an array of necessary items, including food, shelter, clothing, and utilities -- and then setting poverty rates based on how much families at the 33rd percentile spend -- the SPM adjusts poverty thresholds as societies' spending patterns on these necessities shifts.⁹ This type of threshold is "quasi-relative," in the sense that thresholds will tend to rise with income, but are not directly tied to income as in a purely relative definition of poverty, such as setting a threshold at half the median household income.

Alternatively, it is possible to create an "anchored" version of the Supplemental Poverty Measure, which is the focus in this chapter. The anchored version, like the official measure, fixes poverty thresholds based on expenditures on necessary items in a given year and then adjusts only for inflation in each year. This version allows for the use of the more comprehensive definition of resources in measuring poverty over time, while setting a fixed assessment of what constitutes basic needs spending on food, shelter, clothing, and utilities. The anchored measure is also more consistent with the vision of the War on Poverty architects, who believed that poverty can be eradicated. Eliminating poverty defined with a relative measure may be nearly impossible, as the threshold rises apace with incomes.¹⁰

Who is Poor?

In an attempt to "provide some understanding of the enemy" for Johnson's War, the 1964 Economic Report of the President presented tables depicting the "topography of poverty." As Table 1 shows, some of the landmarks have changed since 1960 while many remain the same. For 2012, the Table presents the poverty rates measured with both the official poverty measure and the Supplemental Poverty Measure.¹¹ The official measure is displayed for comparing the relative poverty of various groups in the two time periods, but should not be used for comparing changes in the levels of poverty between the two time periods due to the flaws in the official measure discussed above. The next section will provide trend data using a consistent measure. Since historical estimates of the SPM are available only starting in 1967, Table 1 shows only official poverty rates for 1959 using the 1960 Census.

 

Table 6-1

 

Poverty Rates by Selected Characteristics, 1959 and 2012

 

 

[ Editor's Notes: To view the table,

 

see Doc 2014-5547 , page 229.]

 

 

 

Unsurprisingly, unemployment is one of the strongest predictors of poverty. In 1959, 55.7 percent of individuals in households where the head was out of work for a full year were poor -- three times the rate of individuals in households where the head worked at least one week during the year. While this rate has declined to 29.2 percent, individuals in households where the head was out of work for a full-year were still three times as likely to be poor as those in households where the head worked.

However, even full-time employment is not enough to keep all families out of poverty today. A person working full-time, full-year in 2013 being paid the minimum wage earns $14,500 for the year. These earnings alone leave such workers below the poverty threshold if they have even one child. While the EITC, SNAP, and other benefits will help pull a family of two above the poverty line, for a larger family -- such as one with three children -- full-time, full-year minimum wage work combined with government assistance is unlikely to be enough to lift that family out of poverty.

Education Level

Education's role in poverty prevention has become more important over time: in 1959, high-school dropouts were 3.8 times more likely to be poor than college graduates; but in 2012, they were 6.1 times more likely to be poor (based on the SPM measure). The growth in the poverty gap by education is driven by growth in earnings inequality, which has led to much greater earnings for college graduates than for those with less education.

Children

As in 1959, the child poverty rate today is higher than the poverty rate of the overall population, although the SPM shows that children are disproportionately helped by our poverty-fighting programs. Once taxes and in-kind transfers are taken into account, the gap between child and non-child poverty falls. According to the official measure, the child poverty rate in 2012 was 22.3 percent -- nearly 48 percent higher than the overall rate. But the official rate ignores the contributions of the most important antipoverty programs for children: the EITC and other refundable tax credits and SNAP. Including the value of these resources, the SPM estimates that 18 percent of children are poor -- a rate that is 12.5 percent (2 percentage points) higher than the overall poverty rate.

The Elderly

One of the most heralded successes of the War on Poverty is the large reduction in elderly poverty rates. In 1959, poverty rates were highest among the elderly with 39.9 percent of people 65 and older living in poverty (based on OPM). Today, poverty rates of those 65 and older are below the national average. Using the SPM measure shows elderly poverty at 14.8 percent, which is more than 50 percent higher than when taken with the OPM measure. The reason for this difference is that the SPM subtracts expenditures on medical expenses from a family's resources, and the elderly tend to have much higher medical expenses. Indeed, if the out-of-pocket medical costs were not subtracted, the measured elderly poverty rate would be only 8.4 percent -- lower than their official poverty rate of 9.1 percent.¹² In the absence of Medicare and Medicaid, out-of-pocket medical expenses of the elderly would almost certainly cause their poverty to be much higher than the SPM poverty rate of 14.8 percent.

SPM Poverty Rates for Working Age Adults by Gender, 1967-2012

 

 

 

Women

Women are more likely than men to be in poverty, with a 2012 poverty rate of 16.7 percent compared to 15.3 percent among men. This gap largely reflects higher poverty among single women, both those age 18-64 (22.9 percent compared to 20.2 percent among single men) and those age 65 or older (21.2 percent compared to 16.1 percent among single men).

Childcare responsibilities help explain the poverty gap for single working-age women. Almost one-third (31.0 percent) of single women age 18-64 lived with their children in 2012, among whom the poverty rate was 27.5 percent. Just over one-third (35.2 percent) of single mothers age 18-64 were employed full-time, full-year in 2012, compared to 44.6 percent of single working-age males and 43.3 percent of single working-age women without children at home. Increased support for childcare for young children has been shown to positively affect mothers' employment hours and earnings (Connelly and Kimmel 2003; Misra, Budig, and Boeckmann 2011). The high poverty rate among older single women relative to men reflects a combination of lower Social Security benefits due to lower lifetime earnings; lower rates of pension coverage; and greater longevity that increases the chances of outliving their private savings (Anzick and Weaver 2001, SSA 2012).

Race and Ethnicity

Poverty rates have fallen for all racial and ethnic groups over time and gaps by race have shrunk slightly. However, troubling gaps still remain. In 1959, nearly three-fifths of African Americans were in poverty, which was nearly three times the poverty rate of Whites. The fraction of African Americans in poverty has fallen by more than half since then; yet at 25.8 percent, the SPM poverty rate for African Americans is still more than double the rate of 10.7 percent for Whites. Today, the SPM poverty rate among Hispanics is 27.8 percent, similar to that among African Americans. However, this reflects smaller declines in poverty among Hispanics over the past 50 years. Among both African Americans and Hispanics, the official and supplemental poverty rates tell similar stories. However, the SPM reveals that Asian Americans have a slightly higher poverty rate than the national average -- at 16.7 percent -- while their OPM poverty rate is lower than the national average. The higher SPM rate for Asian Americans reflects, in part, the fact that they tend to live in high-cost metropolitan areas (for example, Los Angeles and New York City) and the SPM poverty thresholds are higher in such places due to its geographic adjustments for cost of living. Finally, while measures of poverty among American Indians and Alaska Natives are not available in the earlier period, currently they have the highest rates of poverty of any race and ethnicity group at 30.3 percent in 2012.

People with Disabilities

Over one-fourth of working-age people with disabilities are estimated to live in poverty, using both the OPM (28.4 percent) and SPM (26.5 percent) measures. This largely reflects their low employment rates. The effective poverty rate of people with disabilities may be understated due to the extra costs that often accompany disability, such as for home and vehicle renovations, assistive equipment, personal assistance, and other items and services that may not be covered by insurance or government programs (Sen 2009: 258, She and Livermore 2007, Fremstad 2009, Schur, Kruse, and Blanck 2013: 32-33).

Rural and Urban Communities

The official poverty measure overestimates rural poverty since rural communities tend to have lower costs of living than urban areas and the official measure does not take geographic cost-of-living differences into account. However, the OPM has revealed that significant poverty persists in rural communities throughout the country today. The Economic Research Service of the U.S. Department of Agriculture estimates that 85 percent of persistent poverty counties -- counties that have been in high poverty (over 20 percent based on the OPM) for at least 30 years -- are in rural areas.¹³ The gap between poverty rates outside and within metropolitan areas, though, has narrowed since 1959 when poverty rates outside metropolitan areas were more than double the rates within these areas. In fact, the adjustments for different housing costs across geographic areas in the SPM show that poverty rates are higher in metropolitan areas than in rural areas today.¹⁴

Context

The architects of the War on Poverty were confident they would live to see poverty eradicated. Looking at the data at their disposal at the time, it is easy to see how an analyst might have believed the end of poverty was on the horizon. Figure 6-1 shows the trend in the official poverty measure -- the only consistently available measure tracking poverty until recently -- from 1959 through 2012. Based on the trend in poverty observed between 1959 and 1968, one would have indeed forecast -- extrapolating linearly -- that poverty would be eradicated by 1980. Poverty fell by a remarkably consistent rate of about 1.15 percentage points a year over that 10-year period, but the official poverty measure stopped declining afterwards, reaching its lowest point in 1973.

 

Figure 6-1

 

Trends in the Official Poverty Measure, 1959-2012

 

 

 

As previously noted, the OPM is a measure of cash income that does not include non-cash benefits or tax credits. While it does not accurately capture the trend in poverty over time, it is nonetheless worth considering why the improvement in this measure of cash income slowed so abruptly in the early 1970s. The first, and clearest, answer is that Social Security expansions during the 1960s brought the rate of poverty among the elderly down rapidly before leveling in the 1970s (Engelhardt and Gruber 2006). In 1959, 39.9 percent of those 65 and older were in poverty, but by 1974 that fraction had fallen to 14.6 percent (based on the OPM). Over the next 38 years, the elderly poverty rate fell further to 9.1 percent in 2012. The deceleration in poverty reduction is less pronounced for nonelderly adults and children. In fact, using the SPM measure that accounts for expansions of the EITC and non-cash transfers, children's poverty had greater declines in the 1990s than in the 1960s or 1970s.

Growth in inequality has also helped put the brakes on improvements in cash income for most households. Economic growth is an important determinant of poverty (Blank 2000) as long as the gains are shared with those in the bottom of the income distribution. When growth fails to benefit the bottom, it cannot play a role in eradicating poverty. As such, the distribution of income can have a profound impact on the level of poverty. While the real economy grew at an annual rate of about 2.1 percent during the 1970s and 1980s, since 1980 economic growth has not produced the "rising tide" heralded by President Kennedy, as rising inequality left incomes at the bottom relatively unchanged (DiNardo, Fortin, Lemieux 1996, Piketty and Saez 2003, Lemieux 2008). As shown in Figure 6-2, incomes in the top 20 percent of the income distribution rose dramatically until the 2000s and are about 50 percent higher today than in 1973. By contrast, real household incomes in the bottom 60 percent of the income distribution stagnated until the mid-1990s expansion, and today are little changed from the business cycle peak in 1973.

 

Figure 6-2

 

Average Real Household Income by Quintile, 1967-2012

 

 

 

A large group of poverty scholars have pointed to this rise in inequality as a leading explanation for the lack of progress in reducing poverty since 1980 (for example, Blank 1993; Gottschalk and Danziger 1995, 2003; Hoynes, Page, and Stevens 2006).

The failure of the minimum wage to keep up with inflation is an important reason why inequality increased in the 1980s (DiNardo, Fortin, and Lemieux 1996, Lee 1999), and progress in the fight against poverty has slowed. President Johnson extended both the level and scope of the minimum wage, with its peak reached in real terms in 1968. Since then, the minimum wage has risen and fallen, but today its level of $7.25 an hour is the same in real terms as in 1950. At this level, even factoring in the subsidy provided by the EITC, a single parent of two kids working full time would still have income near the poverty line.

• While the safety net provides crucial support for families in poverty, far more Americans benefit from the safety net than are poor in any given year. Of course, all Americans benefit from Social Security and Medicare support for the aged, shielding both them and their families from low income in retirement and the costs of adverse health shocks. And many people benefit from social insurance programs that are not means-tested. For example, nearly half of all Americans will benefit from unemployment insurance at some point over a 20-year period.

  But even programs targeting primarily low-income families serve a very high fraction of Americans at some point in their lives. A recent study using administrative tax records from 1989 to 2006 found that over 50 percent of tax-filers with children benefited from the EITC at some point over the 19-year period (Dowd and Horowitz 2011). Moreover, CEA analysis of the National Longitudinal Study of Youth 1979 finds that of all individuals aged 14 to 22 in 1979, over the 32-year period from 1978 to 2010:

• 29.6 percent benefitted from SNAP;

• 34.2 percent received support from SNAP, AFDC/TANF, or SSI; and

• 69.2 percent received income from SNAP, AFDC/TANF, SSI, or UI.

• Looking at a broader array of programs, a large fraction of the population benefits in any given year as well. According to the 2013 Annual Demographic and Economic Supplement of the Current Population Survey, nearly half (47.5 percent) of all households received support from either refundable tax credits, SNAP, Unemployment Insurance, SSI, housing assistance, school lunch, TANF, Women, Infants, and Children (WIC), Medicaid or Disability Insurance.

  An important feature of the safety net that is often overlooked: for most programs the majority of beneficiaries receive assistance for only a short period when their earnings drop for some reason, and then they bounce out again. Research has shown, for example, that 61 percent of all EITC recipients claimed the credit for two years or less (Dowd and Horowitz 2011); and, half of all new SNAP participants in the mid-2000s left the program within 10 months.¹

 

END OF FOOTNOTES TO BOX 6-4

 

 

Figure 6-3

 

Women's 50-10 Wage Gap vs Real Minimum Wage, 1973-2012

 

 

 

Several studies have documented a tight link between the value of the minimum wage and measures of wage inequality in the bottom part of the income distribution (Lee 1999, DiNardo, Fortin, and Lemieux 1996). For example, as shown in Figure 6-3, changes in the ratio of the median wage to the 10th percentile of the wage distribution (the "50-10 wage gap") -- a measure of inequality in the bottom of the wage distribution -- for women correlate very closely with changes in the real value of the minimum wage.¹⁵ The best research suggests that increases in the minimum wage do not result in job losses large enough to undermine the goal of raising incomes for the poor (Dube, Lester, Reich 2010). Moreover, a recent meta-analysis by Doucouliagos and Stanley (2009) covering over 1,000 estimates of minimum wage effects finds "no evidence of a meaningful adverse employment effect." Finally, a recent analysis and review of the literature by Dube (2013) finds consistent evidence across studies that a 10 percent increase in the minimum wage decreases the poverty rate by about 2.4 percent.

Another important factor in rising inequality and slow wage growth among low- and middle-income workers has been the decline in unionization. The percent of U.S. workers represented by unions has nearly halved from 23.3 percent in 1983 to 12.4 percent in 2013.¹⁶ This decline has contributed to inequality because unions reduce inequality by raising the wages of low- and middle-income workers and compressing the returns to skill (DiNardo, Fortin, and Lemieux 1996; DiNardo and Lemieux 1997).

Many observers have implicated various demographic changes -- most prominently, increased immigration and a decline in two-parent families -- as additional factors behind the lack of progress in market incomes in the lower part of the distribution. The recent literature on immigration rejects the claim that competition from immigrants has had a meaningfully adverse effect on the wages or poverty rates of native workers (Peri 2013). Because the country-of-origin composition of immigrants has increasingly shifted toward poorer countries, however, immigration has had a mechanical effect on poverty rates. Card and Raphael (2013) estimate that changes in the population shares and the country of origin of the foreign born increased overall poverty rates (based on OPM) by 3.7 percentage points between 1970 and 2009. While some immigrant households that may have seen their incomes rise as a result of coming to the United States, many still fall below the poverty line here. But other analysts focusing on different time periods generally find much smaller compositional effects. For example, Hoynes, Page, and Stevens (2006) find that increased immigration accounts for only a 0.1 percentage point increase in poverty between 1979 and 1999.

Another dramatic change since the 1960s is a large increase in the number of people living in single-female headed households. As shown in Table 1, individuals in such households typically have double the poverty rates of the national average, so this change also tends to increase the poverty rate. Using decomposition techniques, Hoynes, Page, and Stevens (2006) show that changes in family structure alone accounted for a 3.7 percentage point increase in the (OPM) poverty rate between 1967 and 2003.¹⁷ In fact, women's poverty rates declined over this time period due to their educational attainment, labor force participation, higher earnings, and they had fewer children (Reed and Cancian 2001). Moreover, changes in family structure can both cause and be caused by changes in economic circumstances.

The last three decades have also seen a historic rise in incarceration, which has led to greater poverty. The fraction of the population in prison rose from 221 per 100,000 in 1980 to 762 per 100,000 in 2008 (Western and Pettit 2010).¹⁸ In the short term, imprisonment removes wage earners from the family, which reduces their family's income and increases the probability of their children growing up in poverty. For example, Johnson (2008) finds that child poverty increases by 8.5 percentage points and family income falls by an average of $8,700 while a father is in prison.

There are also long-term negative impacts on earnings from incarceration that lead to higher rates of poverty among those with a criminal record. Offenders' wages are lower by between 3 and 16 percent after incarceration (Raphael 2007; Western 2002) and employment and labor force participation are also negatively affected. Research shows that each additional percentage point of imprisonment of African American men is associated with a reduction in employment or labor force participation of young African American men of approximately 1.0 to 1.5 percentage points. This relationship implies that the increases in incarceration over the last three decades have reduced employment and labor force participation among young African American males by 3 to 5 percentage points (Holzer 2007). Holzer also notes that, while the magnitude of the effect of incarceration on White and Latino offenders is less clear, most studies find that their experience with employment and labor force participation after incarceration is similar to that of African American men.

Together, the factors described above created headwinds in the fight on poverty. Evaluating the precise impact of these factors is beyond the scope of this report, but it is likely that their combined influence was to exert modest upward pressure on poverty rates. As shown below, the fact that "market poverty" stayed relatively constant over this time period suggests that improvements in education or other factors may have offset the adverse effects of these demographic and other changes. Previous studies based on the OPM support this notion. For example, Mishel et al. (2013) suggest that the impact of increases in education in reducing poverty were slightly greater than the adverse impact of changing demographics.

Correcting the Historical Account of Poverty Since the 1960s

The official poverty measure introduced by President Johnson's administration ignores, by design, the most important antipoverty programs introduced during and after the War on Poverty. In particular, resources from nutrition assistance, tax credits for working families, and access to health insurance are not considered when computing whether a family is poor by the traditional metric.

The Census Bureau has published the Supplemental Poverty Measure only as far back as 2009. But recent research by poverty scholars on alternate measures of poverty all find that the official poverty rate displayed in Figure 6-1 dramatically understates the decline in poverty since the 1960s (Fox et al. 2013, U.S. Census Bureau 2013, Meyer and Sullivan 2013, Sherman 2013). Work by Wimer, Fox, Garfinkel, Kaushal, and Waldfogel (2013) is particularly valuable since they estimate poverty rates from 1967 to 2012 following the SPM methodology for computing family resources. They also measure poverty using an "anchored" measure that uses a fixed poverty threshold based on expenditures on necessary items in 2012, adjusted only for inflation in each year (with inflation measured using a historically consistent series, the CPI-U-RS).¹⁹

Figure 6-4 shows a striking fact: poverty has declined by 38 percent since 1967, according to the anchored SPM measure. And, unlike the OPM, it continued to fall after the early 1970s. The figure shows the evolution of the poverty rate using the anchored SPM measure of poverty from 1967 to 2012, compared with the official poverty rate reproduced from Figure 6-1. Using the more accurate SPM measure of family resources changes the historical account of poverty in the United States significantly: between 1967 and 2012, poverty rates fell by 9.8 percentage points -- from 25.8 to 16.0 percent. The trend in the SPM depicted in Figure 6-4 is very similar to that of an alternative measure of poverty based on consumption data, which Meyer and Sullivan (2013) argue is a better measure of material hardship (Box 6-2).

Figure 6-4 shows that the fraction of Americans in poverty fell smoothly between 1967 and 1979 to a low of 17.4 percent, a period where the discrepancy with the trend shown by the official measure is driven primarily by more accurate accounting for inflation in the 1970s. After rising steeply during the double-dip recession of the early 1980s, poverty rates fell slightly until rising again with the early 1990s recession.

In contrast to the depiction of the OPM, the steepest declines in the fraction of people in poverty occurred during the economic expansion of the 1990s. During that period, poverty fell from 20.7 percent in 1993 to 14.6 percent in 2000, the lowest poverty rate observed since 1967. As shown in Figure 6-2, economic growth in the 1990s provided a strong boost to low-income households as earnings grew even in the bottom one-fifth of incomes, in contrast to the experience of any other decade since the 1960s. Dramatic increases in the value of the EITC leveraged this upswing in the labor market to further encourage work and channel even more resources to low-income working families.

One last, and remarkable, fact shown in Figure 6-4 is that the poverty rate ticked up only slightly during the Great Recession after remaining steady for most of the 2000s. Despite the largest rise in unemployment since the Great Depression, the poverty rate rose by only 0.5 percentage points overall between 2007 and 2010. As discussed below, this shows how effective the safety net and its expansion through the 2009 American Recovery and Reinvestment Act (the Recovery Act) have been. Since much of the credit for this is due to expansions in SNAP and tax credits, the official poverty rate fails to capture this crucial success.

The poverty rates for children and for working-age adults follow a similar pattern to the overall trend shown in Figure 6-4. The fraction of children living in poverty declined from 29.4 percent in 1967 to 18.7 percent in 2012; for working-age adults, the poverty rate fell from 19.8 to 15.1 percent over the same period. For the elderly, the trend in poverty is one of near continuous decline. Poverty fell quite rapidly up until the early 1980s, driven by large growth in per capita Social Security payments, from 46.5 percent in 1967 to 20.7 percent in 1984. Elderly poverty fell further during the 1990s expansion to a low of 16.5 percent in 1999, and then ticked up slightly in the 2000s. Driven in part by the Recovery Act stimulus payments, elderly poverty declined from 17.6 percent to 15.2 percent between 2007 and 2009, and it remained at that level in 2012.

Measuring the Direct Impact of Antipoverty Efforts

The fact that poverty rates have fallen overall, even as household incomes in the bottom of the distribution have stagnated since the 1980s, suggests a substantial direct role that policies have played in improving the well-being of the poor. Wimer et al. (2013) estimate the magnitude of this impact by constructing "counterfactual" poverty measures that simulate the fraction of the population that would have been poor in the absence of all government transfers, including the overall tax system.²⁰ In other words, they estimate the fraction of families that would have incomes below the poverty line if the value of all cash, in-kind, and tax transfers they received (or paid if the family owed taxes on net) were not counted. Comparing the difference between this measure of "market poverty" and the SPM poverty rate provides a measure of the reduction in poverty accounted for by government transfers. Figure 6-5 shows the results of this analysis. The height of the overall shaded region indicates the poverty rate counting only market income, while the height of the region shaded in black is the SPM poverty rate shown in Figure 6-4. The difference, shaded in green, represents the percentage of the population lifted from poverty by the safety net and the net effect of the tax system.²¹

 

Figure 6-4

 

Official vs Anchored Supplemental Poverty Rates, 1967-2012

 

 

 

In part because of the rising inequality in earnings described above, market poverty increased over the past 45 years by 1.7 percentage points from 27.0 percent in 1967 to 28.7 percent in 2012. In contrast, poverty rates that are measured including taxes and transfers -- these taxes and transfers are the green-shaded region in Figure 6-5 -- fell through most of this period. Government transfers reduced poverty by 1.2 percentage points in 1967. This impact grew to about 7.4 percentage points by 1975 due to the expansion of the safety net spurred by the War on Poverty, and hovered around that level until the Great Recession when it increased to 12.7 percentage points in 2012.

 

Figure 6-5

 

Trends in Market and Post-Tax, Post-Transfer

 

Poverty, 1967-2012

 

 

 

Despite an increase in "market poverty" of 4.5 percentage points between 2007 and 2010, the actual SPM poverty rate rose only 0.5 percentage points due to the safety net. In fact, the 2009 Recovery Act reduced poverty by 1.7 percentage points in 2010 through its extensions to the safety net as discussed below. Overall, for the entire 45-year period, the poverty decline of 9.8 percentage points is almost entirely accounted for by the increased effectiveness of the safety net.²²

Figure 6-6 shows a similar analysis of the effect of the safety net on trends in deep poverty and highlights two important features of the safety net often overlooked. First, the safety net improves the well-being of many more individuals than is reflected in the standard accounting of how many individuals are lifted from poverty: in 2012, about one in twenty (5.3 percent) Americans lived in deep poverty, yet without government transfers the number would be closer to one in five (18.8 percent).

Second, the safety net almost entirely eliminates cyclical swings in the prevalence of deep poverty. Figure 6-6 shows that despite large increases in deep market poverty driven by the business cycle, there is little if any rise in actual deep poverty due to the supports provided by the safety net.

Again, this phenomenon is especially visible during the Great Recession, when the prevalence of deep poverty ticked upward by only 0.2 percentage points despite an increase in market deep poverty of 3.3 percentage points. This corresponds to more than 9.5 million men, women, and children prevented from living below half the poverty line during the Great Recession. Over the 45 years shown in Figure 6-6, deep "market poverty" actually rises from 14.9 to 18.8 percent. Despite that increase, the fraction of those in deep poverty fell from 8.2 to 5.3 percent.

The section then refutes the concern critics have raised that the existence of safety net programs may undermine growth in market incomes as well as our efforts to fight poverty. The social safety net has increasingly been designed to reward and facilitate work increasing participation rates -- in many cases, requiring work. Even where programs are not explicitly designed to require work, the highest-quality studies suggest that adverse earnings effects of safety net programs are nonexistent or very small, in part due to reforms over the past two decades that, for example, have phased out benefits gradually with increases in earnings to minimize disincentives to work.

Finally, this section presents findings on the level of economic mobility of individuals born into poverty in the United States, and the results of recent research showing the potentially large returns to social spending in terms of long-term outcomes of children in families receiving support.

Antipoverty Effects of Specific Programs

This section illustrates the role that various antipoverty programs have had in improving the well-being of different populations, and how the relative impacts of these programs have evolved since the War on Poverty began. It is based on an effectively "static" analysis that zeroes out income derived from various public programs to ask what effect this would have on poverty. The next subsection considers some of the broader impacts that these programs have on employment and earnings.

 

Figure 6-6

 

Trends in Market and Post-Tax, Post-Transfer

 

Deep Poverty, 1967-2012

 

 

 

Table 2 shows the impact of various safety net programs on overall poverty, and for three separate age groups: children, adults age 19 to 64, and elderly adults age 65 and over. The safety net program with the single greatest impact is Social Security, which provides income to the elderly, people with disabilities, and surviving spouses and children, reducing the overall poverty rate by 8.6 percentage points in 2012. The program's impacts on elderly poverty are profound: without Social Security income, the poverty rate of the elderly would be 54.7 percent, rather than its rate of 14.8 percent in 2012. On the other end of the age spectrum, refundable tax credits like the EITC and the Child Tax Credit (CTC) have large impacts on child poverty -- reducing the fraction of children in poverty by 6.7 percentage points.

Tax credits also reduce the poverty rates of nonelderly adults by 2.3 percentage points. The Supplemental Nutrition Assistance Program also has a dramatic effect on poverty, reducing child poverty by 3.0 percentage points and overall poverty rates by 1.6 percentage points.

Finally, unemployment insurance (UI) reduced poverty by 0.8 percent overall in 2012. This effect, as with the effects for other programs, was less than at the height of the Great Recession when more people were without work: in 2010, for example, unemployment insurance reduced poverty by 1.5 percentage points overall (Short 2012).

As noted, all of these estimates ignore the incentives to alter work behavior created by government programs and are not definitive causal estimates of the impact of the different programs on poverty. For example, Social Security may affect market incomes by changing retirement and savings incentives. Similarly, the estimates do not take into account the role that UI plays in keeping people attached to the labor force, or that the EITC plays in incentivizing additional hours of work and participation in the labor force. The importance of these considerations is discussed below.

Even programs with a small impact on overall poverty rates may be very effective in reducing poverty for certain populations or in alleviating hardship without lifting individuals out of poverty. For example, SSI reduces poverty rates by 1.1 percentage point overall, but this represents a large poverty reduction concentrated among a relatively small number of low-income recipients who are elderly or have a disability. TANF and General Assistance (state programs which typically provide limited aid to very poor individuals not qualifying for other aid) have only a small impact on the overall poverty rate at 0.2 percentage points, as TANF benefits are generally insufficient to bring people above the poverty level. However, by raising the incomes of those in poverty these programs have a much greater impact on reducing deep poverty.

Based on their historical estimates of SPM poverty rates, Wimer et al. (2013) conduct similar analyses to those in Table 2 for each year since 1967 for different groups of safety net programs. Their results shed light on how the safety net has changed over the past 50 years.

In aggregate, the antipoverty effects of three types of federal aid programs -- support through cash programs like Social Security, SSI, and TANF; in-kind support like SNAP and housing assistance; and tax credits like the EITC and CTC -- all increase over time, driving down overall poverty rates. The steady increase in the effect of each type of program masks some differences across the populations served by each program. For the elderly, for example, the trend is dominated by the growing real value of Social Security payments steadily driving down the elderly poverty rate.

For children, however, there is a shift in importance of the safety net's different components. Figure 6-7 shows the effect of eliminating various components of the safety net resources on the SPM poverty rate in three years corresponding to recession-driven peaks of the poverty rate. In the early 1970s recession, AFDC, and food stamps (now known as Supplemental Nutrition Assistance Program or SNAP) to a lesser extent, played the most important role in alleviating childhood poverty; and the EITC had not yet been introduced. Both in-kind transfers and the EITC had a small impact on child poverty in the early 1990s recession, but cash transfer programs still outweighed the poverty-reducing impact of both programs put together. During the Great Recession, however, we see that in-kind aid, cash-assistance, and tax credits all played similarly important roles in reducing poverty for families with children. This shift reflects both the large structural change away from cash welfare assistance during the 1990s, and expansion of both SNAP and tax credits through the Recovery Act.

Figure 6-8 is similar to Figure 6-7, only showing the impact of various transfer programs on deep poverty, or the fraction of individuals with incomes below 50 percent of the poverty threshold. This figure shows that for deep poverty, in-kind transfers have become the most important safety net program over time, and tax credits are less effective due to the paucity of work among families with very low resources.

 

Table 6-2

 

Poverty Rate Reduction from Government Programs, 2012

 

 

[ Editor's Notes: To view the table,

 

see Doc 2014-5547 , page 248.]

 

 

 

In his remarks before signing the cornerstone legislation of the War on Poverty, the Economic Opportunity Act, President Johnson declared: "Our American answer to poverty is not to make the poor more secure in their poverty but to reach down and to help them lift themselves out of the ruts of poverty and move with the large majority along the high road of hope and prosperity. The days of the dole in our country are numbered." He went on to describe the need to provide the poor with the means to lift themselves out of poverty through job training and employment services, among other strategies. In the past 20 years, this emphasis on promoting work through antipoverty programs intensified and we made dramatic changes to cash welfare and other programs that shifted the focus toward requiring or rewarding work.

The most important shift began with the introduction of the EITC in 1975. The EITC was expanded multiple times in the 1980s, 1990s, and 2000s as the safety net shifted increasingly toward work-based support. The EITC was expanded most recently in the 2009 Recovery Act, with those improvements extended in 2010 and 2013. Since 1996, the EITC has accounted for more support for low-income households than traditional cash welfare. In 2012, the EITC and the partially refundable CTC totaled $90 billion annually, more than four times the expenditures on the Temporary Assistance for Needy Families program.

Because the EITC is a supplement to labor market earnings, the credit provides strong incentives for those with otherwise low labor force attachment to increase their hours of work. However, because the credit is reduced as earnings continue to rise, it may also provide those with earnings above the phase-out threshold -- above $22,870 for a married couple with three children -- an incentive to reduce their earnings. Several studies have addressed these incentive effects, and reach similar conclusions: the EITC is associated with increased labor force participation, especially among single mothers, but it does not appear to substantially alter the hours or earnings of those already working (Eissa and Liebman 1996, Liebman 1998, Meyer and Rosenbaum 2001, Hotz and Scholz 2003, and Eissa and Hoynes 2005).²³ Taken together, these results suggest EITC expansions played an important role in increasing labor force participation among single mothers, without adversely affecting hours worked by those already working.

Meanwhile, the 1996 welfare reform law replaced AFDC with TANF and significantly strengthened work requirements in the cash assistance program. The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 ended the entitlement to cash assistance and beneficiaries were generally required to work or participate in "work activities" to receive assistance. Moreover, the implicit tax rate on benefits in response to increased earnings -- the benefit reduction rate -- was reduced dramatically in many states. Matsudaira and Blank (2013) show that these changes increased the return to work, with potential income gains of over $1,842 (in 2000 dollars) for welfare recipients working 30 hours a week. At the same time, it should be noted that researchers have found these reforms may have increased hardship among groups with high barriers to employment, highlighting the continued importance of programs that serve the most disadvantaged (Blank 2007; Danziger, Turner, and Seefeldt 2006).

 

Figure 6-7

 

Percentage Point Impact on SPM Child

 

Poverty for Selected Years

 

 

 

Source: Wimer et al (2013).

 

Figure 6-8

 

Percentage Point Impact on Deep SPM Child

 

Poverty for Selected Years

 

 

 

Source: Wimer et al (2013).

While today's safety net has been reformed to promote work, it is important to note that careful research has shown that most assistance programs have only small, if any, disincentive effects on work. This suggests that the "static" estimates of the antipoverty effects of the programs shown above largely capture the actual full impact of these programs including any employment disincentives they may have -- and may understate the poverty-reducing impacts of programs that effectively reward and facilitate work and thus increase market earnings. For example, examining labor supply behavior of individuals in the Oregon Health Insurance Experiment, Baicker et al. (2013) find that Medicaid recipients are not less likely to be employed nor do they earn less than they otherwise would have. Similarly, Hoynes and Schanzenbach (2012) study the initial rollout of SNAP (then food stamps) and find only small effects on labor supply. While there are no studies of the employment effects of the TANF program with its work requirements, prior studies of AFDC and the Negative Income Tax experiments of the 1970's suggest only modest disincentive effects of a program without an emphasis on work. Burtless (1986) found that a dollar of benefits reduces work earnings by 20 cents (total income is increased by 80 cents), but other evidence suggests the disincentive effects may have been even smaller (SRI International 1983).

Evaluating the weight of the evidence for all programs, Ben-Shalom, Moffitt, and Sholz (2011) conclude that the work disincentive effects of antipoverty programs have "basically, zero" effect on overall poverty rates. Going program by program, they conclude the behavioral effects of TANF are likely zero, and that the work disincentives induced by disability insurance, Medicare, and Unemployment Insurance might reduce the estimated "static" antipoverty effects of those programs by one-eighth or less. Although housing assistance provides significant benefits to some of the poorest households including the homeless, its effects on labor supply among those of working-age and free of disabilities are relatively modest. Shroder (2010) reports that the net negative impact of rental assistance on labor supply appears to vary among subgroups, may change over time, and seems rather small relative to the amounts paid out in subsidy. Jacob and Ludwig (2012) find that receipt of housing vouchers in Chicago during the welfare reform era reduced employment by 3.6 percentage points among able-bodied working-age individuals and reduced earnings an average 19 cents for each dollar of subsidies. Carlson et al. (2011) find employment effects in the same range for Wisconsin voucher recipients. Ben-Shalom, Moffitt, and Scholz estimate that the poverty rate among housing assistance recipients is 66.0 percent, and use the Jacob and Ludwig data to estimate that housing assistance lowers the poverty rate among recipients by 8.2 percentage points rather than the static estimate of 14.9 points. Finally, a recent study by Chetty, Friedman, and Saez (2012) of the EITC on the earnings distribution also finds that "the impacts of the EITC . . . come primarily through its mechanical effects." They do find, however, that behavioral responses reinforce the effects of the safety net on deep poverty, so that the overall impact of the EITC might be somewhat greater than implied by the "static" estimates because of the increased work induced by the program.²⁴

Economic Mobility

When economic mobility is high, individuals and families can lift themselves out of poverty by taking advantage of opportunities to improve their economic well-being. When economic mobility is low, it is difficult to change one's economic status and people may become stuck in poverty. Mobility can be measured either as relative mobility -- the likelihood of moving up or down the income distribution -- or absolute mobility, which is the likelihood of improving economic well-being in general without necessarily moving up in the income distribution (reflected in the saying, "a rising tide lifts all boats"). When there is low relative mobility, there are few opportunities for poor people to improve their standing in society by moving up the economic spectrum, and children from poor families are likely to continue to have low economic and social status as they become adults, even if their material well-being improves. When there is low absolute mobility, poor people and their children find it hard to escape the economic and material hardships of poverty.

While economic mobility in the United States allows some people to escape poverty, many do not. About half of the poorest individuals remain in the lowest one-fifth of the income distribution after 10 and 20 years, and no more than one-fourth make it to one of the top three income quintiles (Acs and Zimmerman 2008; Auten, Gee, and Turner 2013). Those who were in poor families as children are estimated to have 20 to 40 percent lower earnings as adults compared to those who did not grow up in poverty (Mayer 1997; Corcoran and Adams 1997; Corcoran 2001; Duncan et al. 2012). Among the children of low-earning fathers, about two-fifths of sons and one-fourth of daughters remained in the lowest earnings quintile when they became their father's age (Jäntti et al. 2006). Consistent with this, more than one-third of children who grew up in the lowest family income quintile were in the lowest quintile when they became adults (Isaacs 2008). Among children born in 1971 with parents in the lowest income quintile, 8.4 percent made it to the top quintile by age 26, compared to 9.0 percent of those born in 1986, indicating that the chance of moving up the income distribution has remained low and fairly stable over the past few decades (Chetty, Hendren, Kline, Saez, and Turner 2014).²⁵ More generally, studies find that about one-third to slightly less than one-half of parents' incomes are reflected in their children's incomes later in life (Chetty, Hendren, Kline, and Saez 2014, Black and Devereux 2011, Lee and Solon 2009), indicating that parents heavily influence children's economic fortunes.

The results of these studies imply strong lingering effects from growing up in poverty. The influence of poverty on future economic prospects stem not just from one's own family status, but from growing up in high-poverty neighborhoods that often have lower-quality schools, lower-paying jobs, higher crime rates, and other conditions that can create disadvantages (Sharkey 2009). Among children whose families were in the top three quintiles of family income, growing up in a high-poverty neighborhood raises the likelihood of downward mobility (falling at least one quintile) by 52 percent (Sharkey 2009). A comparison across geographic regions in the United States indicates that economic mobility is higher in areas with a larger middle class, less residential segregation between low-income and middle-income individuals, higher social capital, and lower rates of teen birth, crime, divorce, and children raised by single parents (Chetty, Hendren, Kline, and Saez 2014).

The above results mostly reflect relative mobility (moving up or down the economic spectrum). While absolute mobility (improving economic well-being without moving up the economic spectrum) has generally risen in the United States, the pace has slowed in recent decades. Comparing cohorts of men in their 30's, median personal income went up 5 percent from 1964 to 1994, but down 12 percent from 1974 to 2004 (Sawhill and Morton 2007). Counting all family income, income went up 32 percent between 1964 and 1994, and only 9 percent between 1974 and 2004. So the recent improvement in family income, which reflects the rising employment of women, was only one-third as large as in the earlier period.²⁶

 

Figure 6-9

 

Real Per Capita Expenditures on Select Programs, 1967 -2012

 

 

 

Despite the popular view that the United States is the land of opportunity, this Nation appears to have lower economic mobility than many other developed countries. Measuring mobility by the strength of the dependence of children's incomes on parents' incomes, the United States has similar mobility as the United Kingdom and Italy, but lower mobility than other European countries as well as Japan, Australia, and Canada (Solon 2002; Jäntti et al. 2006; Corak 2006, 2011).

Can anything be done to increase mobility? A study of data on families over time (including comparisons of twins and other siblings) found that genetics and shared upbringing play a statistically significant, but quantitatively minor role, in explaining adult earnings differences, indicating that environmental factors other than upbringing are largely responsible (Björklund, Jäntti, and Solon 2005). The large variation in mobility across countries and across regions (Chetty, Hendren, Kline, and Saez 2014) is further evidence for the fact that institutions and other potentially changeable factors can have a large impact on mobility.

Education appears to be one of the key factors that drive economic mobility. As shown in Figure 6-10, among families in the bottom fifth of the income distribution, almost half (45 percent) of the children who did not obtain college degrees remained in the poorest fifth as adults, while only one-sixth (16 percent) of the children who obtained college degrees remained in the poorest fifth (Isaacs, Sawhill, and Haskins 2008: 95).

The importance of education is also shown by the findings that economic mobility is higher in countries that have a greater public expenditure on education (Ichino, Karabarbounis, and Moretti 2009) and areas of the United States that have a higher-quality K-12 education system (Chetty, Hendren, Kline, and Saez 2014), and is improved for children in the poorest one-third of families when states increase spending on elementary and secondary education (Mayer and Lopoo 2008).

Intergenerational Returns

While much of the literature on mobility presented above is correlational, a handful of well-crafted studies that track the long-run outcomes of children exposed to safety net programs highlight the potential for investments in these programs to generate large returns.

Early childhood education has been found by many researchers to have dramatic, super-normal returns in terms of more favorable adult outcomes. The Head Start program created early in the War on Poverty has been heavily researched and the combined results show that it can "rightfully be considered a success for much of the past fifty years" (Gibbs, Ludwig, and Miller 2013: 61). Studies following children over time, and accounting for the influence of family background by comparing siblings, found that Head Start participants were more likely to complete high school and attend college (Garces, Thomas, and Currie 2002), and scored higher on a summary index of young adult outcomes that also included crime, teen parenthood, health status, and idleness (Deming 2009). The latter study found that Head Start closed one-third of the gap in the summary outcome index between children in families at the median and bottom quartiles of family income. Using a regression discontinuity research design that compared access to Head Start across counties, Ludwig and Miller (2007) found positive impacts of Head Start on schooling attainment, the likelihood of attending college, and mortality rates from causes that could be affected by Head Start. Gibbs, Ludwig, and Miller (2013) suggest the combination of the benefits due to Head Start might produce a benefit-cost ratio in excess of seven.

Randomized experiments studying the Perry Preschool Project, Abecedarian Project, Chicago Child-Parent Centers, Early Training Project, and Project CARE programs largely confirm these findings. A variety of well-done analyses find that the adult benefits for children -- especially girls -- who participated included higher educational attainment, employment, and earnings along with other benefits (Schweinhart et al. 2005; Anderson 2008; Campbell et al. 2008; Heckman and Masterov 2007; Heckman et al. 2010; Heckman et al. 2011). Heckman et al. (2009) find that the returns to one preschool program (Perry) exceed the returns to equities.

 

Figure 6-10

 

Economic Mobility for Children from First Income Quintile

 

 

 

Source: Isaacs, Sawhill, and Haskins (2008).

In the past decade, researchers have identified long-run linkages between early childhood (including exposure in-utero) health interventions and long-term outcomes. For example, Almond, Chay, and Greenstone (2006) document that the Johnson Administration used the threat of withheld Federal funds for the newly introduced Medicare program to force hospitals to comply with the Civil Rights Act mandate to desegregate, resulting in dramatic improvements in infant health and large declines in the black-white gap in infant mortality in the 1960s. Chay, Guryan, and Mazumder (2009) show these improvements in access to health care and health soon after birth had echoes in the form of large student achievement gains for black teenagers in the 1980's, contributing to the decline in the black-white test score gap. Their results suggest that improved health care access and better early childhood health improve test scores by between 0.7 and 1 standard deviations -- a very large impact that implies large increases in lifetime earnings. For example, studying a different intervention, Chetty, Friedman, and Rockoff (2011) find the financial value (the net present value at age 12 of the discounted increase in lifetime earnings) of a standard deviation increase in test scores to be $46,190 per grade.

While it may not be surprising that human capital interventions have long-run returns, recent studies have also found intergenerational effects on child outcomes from tax or near-cash transfers to their parents. That is, recent evidence suggests that government transfers that ameliorate child poverty by increasing family income have lasting, long-run benefits in terms of better child outcomes. For example, Hoynes, Schanzenbach, and Almond (2013) study the initial rollout of the Food Stamp (now SNAP) program between its initial pilot in 1961 and 1975. While Food Stamps are distributed as vouchers for food purchases, since their amount is generally less than households spend on food, the vouchers likely affect family behavior in the same manner as increased cash income (Hoynes and Schanzenbach 2009). Hoynes, Schanzenbach, and Almond find that exposure to Food Stamps led to improvements in adult health (reductions in the incidence of high blood pressure and obesity) and, for women, increased economic self-sufficiency. Similarly, Dahl and Lochner (2012) and Chetty, Friedman, and Rockoff (2011) find that family receipt of additional income from refundable tax credits improves the achievement test scores of children in the family. Chetty, Friedman, and Rockoff estimate that the implied increase in adult earnings due to improved achievement as a child is on the same order of magnitude, and probably greater, than the value of the tax expenditures.

The results discussed above highlight the crucial fact that government expenditures on the safety net have a strong economic justification. Not only do they help to propel struggling adults back onto their feet and protect them and their families from hardship, they improve opportunity and the adult outcomes of their children. As such, the poverty-reducing impact of these programs constitutes an important investment opportunity. To give a sense of the magnitude of this opportunity, Holzer et al. (2008) estimate the cost of childhood poverty at about $500 billion (in 2007 dollars) or about 4 percent of gross domestic product (GDP) annually in terms of foregone earnings, increased costs of crime, and higher health expenditures and lower health.

While the Holzer et al. study is correlational (though it attempts to correct for hereditary components of the intergenerational income correlation), the concern over bias in this estimate is overwhelmed by its magnitude. Based on Census Bureau estimates, the total poverty gap -- the shortfall between family resources and the SPM poverty thresholds -- among all families with children is about $59.8 billion in 2012, or 0.37 percent of GDP. Even if the Holzer et al. estimate was double the "true" causal effect of eliminating child poverty, the benefit would exceed the added costs five-fold. These numbers create a powerful case for renewed efforts to fight poverty.

At the same time, we must realize that while antipoverty programs are doing heroic work to lift struggling families from poverty, there is broad consensus among economists that a strong national recovery in the short run, and stronger economic growth in the long run, are necessary to sustain progress in the fight against poverty. Indeed, as our social safety net reinforces the economic benefits of work by supplementing wages for low earners, a strong labor market with jobs available for all is an essential partner in the fight against poverty. Given the growing economic inequality in the past few decades, we must strive for balanced growth that benefits all Americans. To do so, we must ensure that an economic expansion encompasses everyone, and commit to giving all Americans opportunities for lifelong learning and skills development to ensure a broad base of human capital that will be rewarded by good wages.

This section documents the Obama Administration's record in continuing the fight to expand opportunity and reduce poverty, and discusses the Administration's proposals to strengthen the safety net and to improve human capital and increase labor market earnings. These comprise a key part of the broader economic strategy to further the recovery and increase growth -- all of which combat poverty.

Taking Immediate Action During the Economic Crisis

As the economy was sliding into the Great Recession, the Administration took action to strengthen the safety net and prevent millions of Americans from falling into poverty. The Recovery Act instituted a number of temporary antipoverty measures, including the creation of the Making Work Pay tax credit worth up to $800 for a married couple, a $250 Economic Recovery Payment for Social Security and SSI recipients; unemployment changes including an additional $25 a week (for up to 26 weeks) to regular UI beneficiaries, increased federal funding for the Extended Benefits program, incentives for states to modernize their UI system to reach part-time workers and recent workforce entrants, and reauthorized Emergency Unemployment Compensation; an increase in SNAP benefits; and an expansion of the Community Services Block Grant (CSBG). The Recovery Act also delivered nearly $100 billion to States, school districts, postsecondary institutions, and students to help address budget shortfalls and meet the educational needs of all. This total included $10 billion for the Title I Grants to Local Educational Agencies program, a flagship program of the War on Poverty's Elementary and Secondary Education Act of 1965, which currently serves more than 23 million students in high-poverty schools, helping ensure access to a high-quality public education.

In addition, the Recovery Act included expansions in tax credits that the Administration intends to make permanent. The EITC expansion increased the credit for families with three or more children, reflecting the fact that they have greater needs and higher poverty rates than families with two children. It also reduced the penalty that some low-income families faced when getting married. Together, these two provisions benefit about 6 million households a year by an average of $500. Further, the partial refundability of the CTC for working families was expanded, benefiting 12 million families by an average of $800 each. These changes were subsequently extended through 2017 and the President is proposing to make them permanent.

The impact of these emergency measures on poverty was dramatic. The Recovery Act played a large role in keeping Americans out of poverty during the recession, as shown in Figure 11.²⁷ In total, between 4.0 and 5.5 million people a year were kept out of poverty by these programs from 2009 to 2012. Without the Recovery Act, the Supplemental Poverty Rate would have been 1.8 percentage points higher in 2009 and 1.7 percentage points higher in 2010. Over the four years between 2009 and 2012, CEA estimates that 19.2 million person years were kept from poverty as a result of the expansions created by the Recovery Act alone (Figure 6-11). This calculation is conservative, in that it does not account for the impact of those expansions on employment through increased aggregate demand, and does not attempt to measure the impact of other components of the Recovery Act such as increased funding for Pell Grants and paying for COBRA for the unemployed.

Expanding Health Care Security

The Affordable Care Act (ACA) ensures that all Americans have access to quality, affordable health insurance and provides financial incentives to states to expand their Medicaid programs to adults who are in or near poverty. As of January 2014, 26 states have adopted the Medicaid expansion. For moderate-income Americans, the ACA provides tax credits for the purchase of insurance through marketplaces and cost-sharing reductions. The Congressional Budget Office (CBO) estimates that, by 2016, these measures will increase the number of Americans with health insurance by 25 million (CBO 2013).

Americans of all income levels are already benefiting from insurance market reforms that ensure access to preventive services and no lifetime coverage limits. Further, Americans buying insurance are no longer forced to pay a higher premium because of their gender or health status, and can be confident that their insurance provides adequate financial protection. The ACA has also begun reforming the way the United States pays for health care to promote more efficiency and quality in the health system, the early results of which are discussed in a recent CEA report (CEA 2013) and in Chapter 4 of this report.

Nutrition programs like SNAP are vital to the economic livelihood of many families and communities, especially in a recessionary period. Every time a family uses SNAP benefits to put healthy food on the table, the benefits extend widely beyond those individuals. In fact, the U.S. Department of Agriculture's Economic Research Service estimates that an additional $1 billion in SNAP benefits supports an additional 8,900 to 17,900 full-time-equivalent jobs (Hanson 2010).

Rewarding Work

Work that pays enough to support a family is the most central anti-poverty measure. In 2013, the Obama Administration finalized rules to extend minimum wage and overtime protections to nearly 2 million direct-care workers who provide care assistance to elderly people and people with illnesses, injuries, or disabilities. This will ensure our nation's health aides, personal care aides, and certified nursing assistants receive the same basic protections already provided to most U.S. workers, while improving the quality and stability of care for those who rely on them.

 

Figure 6-11

 

Recovery Act and Subsequent Extensions: Cumulative

 

Person-Years Kept from Poverty, 2008-2012

 

 

 

Minimum wage and overtime protections have been a bulwark of protection against poverty, but the minimum wage has not kept pace with inflation. Today, a worker trying to support a family on the minimum wage is still living in poverty. That is why the President signed an Executive Order in 2014 to raise the minimum wage for workers on new and replacement Federal service and construction contracts to $10.10 an hour. This step will to ensure that no worker who provides services to the Federal government will raise their families in poverty, and will make Federal procurement more economical and efficient. An extensive body of research suggests that giving a raise to lower-income workers reduces turnover and raises morale, and can thus lower costs and improve productivity. To help insure that work is rewarded for millions more Americans, the President supports the Harkin-Miller bill to increase the minimum wage to $10.10 by 2016 (Box 6-5).

To further enhance economic security and incentivize work, the President has proposed to double the childless EITC to be worth up to $1,000 and lower the age threshold from 25 to 21 to help more lower-income young people, while continuing protections to ensure that it does not benefit, for example, typical full-time students. A small EITC for childless households was established in 1993, but its maximum value is expected to be only $503 in 2015 and is fully phased out for individuals making over $14,790 ($20,290 for married couples). This leaves a childless adult with wages equal to the poverty line with a federal tax burden (including income and payroll taxes) of $1,966 after receiving an EITC of $173, driving them deeper into poverty and making childless workers the sole demographic driven deeper into poverty by the federal tax system. Under the President's proposal, a household at the poverty line would see its EITC expand to $848, more than eliminating its income taxes -- although it would still pay net taxes on earnings when including payroll taxes.

To prepare Americans for the jobs of the future, we have to strengthen our investments in the Nation's educational system. The Administration has invested in coordinated state systems of early learning and proposed new policies building on evidence of how to create a foundation for success in the formative early years of life. High-quality early learning and development programs can help level the playing field for children from lower-income families on vocabulary, social and emotional development, and academic skills while helping students to stay on track and stay engaged in the early elementary grades. These programs generate a significant return on investment for society through a reduced need for spending on other services, such as remedial education, grade repetition, and special education, as well as increased later productivity and earnings. The Administration's comprehensive early-learning agenda invests in early childhood education, care, and development for our nation's youngest learners. In partnership with the states, the Preschool for All initiative would provide high-quality preschool for four-year olds from low- and moderate-income families, while encouraging states to serve additional four-year olds from middle-income families. The Administration has also proposed investments in high-quality early learning for infants and toddlers through new Early Head Start-Child Care Partnerships, as well as an extension and expansion of evidence-based voluntary home visiting programs that allow nurses, social workers, educators, and other professionals to connect pregnant women and vulnerable families with young children to tools that positively impact the child's health, development, and ability to learn.

Over the past 50 years, improvements in education at all levels have produced large returns for many Americans and played a key role in the economic mobility of children born to poor families. Because economic progress and educational achievement are inextricably linked, educating every American student to graduate from high school prepared for college and for a career is a national imperative. The President has articulated a goal for America to once again lead the world in college completion by the year 2020, and the Administration's education efforts aim toward this overarching objective. To provide a high-quality education to all American children, the Administration, in partnership with states, has advanced reforms of the Nation's K-12 education system to support higher standards that will prepare students to succeed in college and the workplace; efforts to recruit, prepare, develop, and advance effective teachers and principals; efforts to eliminate discrimination on the basis of race, color, national origin, gender, and disability in public school; efforts to ensure the use of data in the classroom; and a national effort to turn around our chronically lowest-achieving schools. The School Improvement Grants (SIG) program has invested over $5 billion in 1500 of the nation's lowest performing schools. The Promise Neighborhood Program, established by this Administration in 2010, has funded 58 neighborhoods across the country to design comprehensive projects that include a continuum of services and designed to combat the effects of poverty and improve education and life outcomes, from birth through college to career. The Administration has also put forward proposals to redesign the Nation's high schools to better engage students and to connect 99 percent of students to high-speed broadband and digital learning tools within the next five years. Continued investments and reforms are needed to ensure that all students have access to a high-quality education that prepares them for college and a career, and for success in today's global economy.

With the average earnings of college graduates at a level twice as high as that of workers with only a high school diploma, higher education is now the clearest pathway into the middle class. Our Nation suffers from a college attainment gap, as high school graduates from the wealthiest families are almost certain to continue on to higher education, while just over half of high school graduates in the poorest one-quarter of families attend college. This gap has increased over the past several decades. And while more than half of college students graduate within six years, the completion rate for low-income students is around 25 percent. To achieve the President's goal for college completion, ensure that America's students and workers receive the education and training needed for the jobs of today and tomorrow, and provide greater security for the middle class, the Administration is working to make college more accessible, affordable, and attainable for all American families. Under President Obama, Pell Grant funding was increased to serve over 3 million additional low-income students and the average grant was increased by more than $900. The Administration also created the American Opportunity Tax Credit to ease college costs for over 9 million families, and championed comprehensive reform of student loans that will save taxpayers $68 billion over the next decade. Finally, the Administration has launched an array of policies to contain college costs, and make it easier for students to manage their student debt through income-based repayment reforms, which limit student loan payments to a percentage of their income so that young workers will not be swamped by debt payments as they are establishing their households and careers.

Given the critical importance of education in expanding skills and opportunities, the Administration implemented several policies designed to increase college access and affordability for low-income students. Recognizing that the opportunity to acquire the skills to get and keep a good job starts early and through education, the President is also proposing to modernize America's high schools for real-world learning. The goal is to provide challenging, relevant experiences, and reward schools that develop new partnerships with colleges and employers, and that create classes that focus on technology, science, engineering, and other skills today's employers are demanding to fill jobs now and in the future.

Creating Jobs and Growing Our Economy

Building on the evidence that well-designed training programs can improve employment and earnings (Andersson et al. 2013), this Administration has proposed investing in subsidized employment and training opportunities for adults who are low-income or long-term unemployed. In 2009 and 2010, 372,000 low-income youth were placed into summer and year-round employment, and supported job opportunities were created for about 260,000 low-income individuals. In addition, the President continues to build public-private partnerships to provide opportunities for low-income youth.

The Administration is using all available tools to help people who have lost their jobs to find new work or to train for new careers in growth fields that will provide better jobs and paths to viable careers. This includes supporting training opportunities that lead directly to a job, and making sure our unemployment system promotes re-employment through wide-ranging reforms to the unemployment insurance program, some of which were adopted in the Middle Class Tax Relief and Jobs Creation Act of 2012, and continued investment in reemployment services, which have proven effective in speeding the return to work.

The President has proposed to build on these successes by further investing in creating job and work-based training opportunities for the long-term unemployed and youth seeking skills and wanting to get into the workplace.

This Administration has already invested $1.5 billion in community college-business partnerships in all 50 states to build capacity and develop curricula to train workers for jobs in growing industries. President Obama has proposed to build on these successes with further investments that will transform community college education and support Americans in getting training to enter skilled jobs.

Investing in and Rebuilding Hard-Hit Communities

Living in a high-poverty area presents various challenges, including crime, limited access to quality education, and scarcity of good jobs. Since these issues often interact with each other and compound the problems they create individually, it is very difficult for people, particularly children, to overcome the disadvantages created by and associated with poverty.

A child's zip code should never determine his or her destiny. To help provide ladders of opportunity so that every child has a chance to succeed, the Administration is working with State and local governments to focus public and private resources on transforming areas of high need into communities of opportunity.

The Administration's Promise Zones initiative focuses existing government resources on competitively selected communities and leverages private investment to create jobs, improve public safety, increase educational opportunities, and provide affordable housing. The Administration will designate 20 communities over the next several years with this intense and layered approach to community revitalization.

That approach includes working with local leadership, and bringing to bear the resources of the President's signature revitalization initiatives from the Department of Education, the Department of Housing and Urban Development, the Department of Agriculture, and the Department of Justice, to ensure that federal programs and resources support the efforts to turn around 20 of the highest poverty urban, rural and tribal communities across the country.

The Promise Zones initiative will build on existing programs, including HUD's Choice Neighborhoods and the Department of Education's Promise Neighborhoods grant programs. The Administration has invested $244 million in Choice and $157 million in Promise since 2010. For every federal dollar spent, Choice Neighborhoods has attracted eight dollars of private and other investment and has developed nearly 100,000 units of mixed-income housing in 260 communities, ensuring that low-income residents can afford to continue living in their communities. Promise Neighborhoods grants are supporting approximately 50 communities representing more than 700 schools. To help leverage and sustain grant work, 1,000 national, state, and community organizations have signed-on to partner with a Promise Neighborhood site. By expanding these programs, the Administration continues to support local efforts to transform low-income urban, rural, and tribal communities across the country.

The most important lesson from the War on Poverty is that government programs and policies can lift people from poverty; indeed they have for the past 50 years. Poverty rates fell from 25.8 percent in 1967 to 16 percent in 2012 -- a decline of nearly 40 percent. In 2012 alone, the combined effect of all federal tax, cash and in-kind aid programs was to lift approximately 14.5 percent of the population -- over 45 million people -- out of poverty.

But another lesson is that we cannot afford to simply embrace any program that purports to achieve this goal or attempt to freeze them in time. Instead, our antipoverty efforts have benefited from enormous changes in public policy since the 1960s, informed by a wealth of research on both successful and failed programs that provide important insights into what does and does not work in fighting poverty. Our safety net has evolved to put more emphasis on rewarding and supporting work, such as by providing greater support to working families through the EITC and refundable CTC, while also making it easier for them to get help from programs like SNAP and Medicaid. In 1967, we spent $19 billion in today's dollars on what was then called AFDC and nothing on the EITC. Today the EITC and partially refundable CTC are 3.8 times the size of the TANF program.²⁸ Meanwhile, the Affordable Care Act advances the goal of providing quality affordable health care to all Americans, with financial incentives to states to expand their Medicaid programs to adults who are in or near poverty and generous tax credits for moderate-income households. Our safety net remains imperfect, but these reforms and improvements represent important progress -- and they also help many families work and raise the rewards to that work.

Nearly 50 million Americans still live in poverty, however, including 13.4 million children, and so there remains a need to do more to help the poor. The 1964 Economic Report of the President estimated that the total shortfall in income necessary to bring all poor families up to the poverty level was $11 billion (about $71 billion in today's dollars), or about 1.6 percent of the country's annual GDP. Our nation has grown much richer since, and today the total shortfall below poverty is only 0.6 percent of GDP. Continuing to make progress in closing that shortfall will require not just defending the programs that have helped reduce poverty but also continuing the efforts to strengthen the economy, increase growth and ensure that growth is reflected in broad-based wage gains so that families can lift themselves out of poverty.

² These figures come from Singh and Kogan (2007). The analysis compares birth outcomes in counties in the top and bottom quintiles of a socioeconomic deprivation index based on Census information on education, occupation, wealth, income distribution, unemployment, poverty, and housing quality. The data are broken out into 5-year periods -- the statistics cited compare 1969-74 to 1995-2000.

³ See Bailey and Danziger (2013) for an authoritative assessment of the various components of the War on Poverty from economists' perspectives.

⁴ A similar poverty measure was adopted internally by the Office of Economic Opportunity in 1965.

⁵ It should be noted that Orshansky herself noted many flaws with her measure, and believed it understated poverty. She argued it measured income inadequacy rather than adequacy, stating "if it is not possible to state unequivocally 'how much is enough,' it should be possible to assert with confidence how much, on an average, is too little" (Orshansky 1965).

⁶ While the term "family" is used here, the SPM differs in its definition of a "family unit" when assuming the unit of individuals over which resources are shared. Most importantly, the SPM includes all related individuals living at the same address, but also cohabiting individuals and co-residing children in their care.

⁷ More accurately, the thresholds are based on average expenditures on food, clothing, shelter, and utilities between the 30^th and 36^th percentiles of that distribution, multiplied by 1.2 to account for other necessary expenses and adjusted for geographic differences in cost of living and family size and structure.

⁸ Korenman and Remler (2013) argue that the SPM's treatment of medical-out-of-pocket expenses actually does a poor job of capturing even deprivation of non-health care goods and services. They argue that households that are able to spend a large amount on health care are frequently those with substantial savings or other resources to draw on, and they present evidence that households with high out-of-pocket expenses frequently score lower on "direct" measures of hardship, like food insecurity.

⁹ So, for example, if families across the income spectrum spend more on, say, housing because preferences for or the ability to pay for space or bathrooms change then what is considered necessary for minimum housing will change.

¹⁰ The SPM is a hybrid, "quasi-relative" measure such that when spending on necessities increases, the threshold defining who is poor also increases. It is unlikely to rise at the same rate as income, as with relative poverty measures, but will adjust more slowly since spending on necessities grows more slowly than income. Eliminating poverty under this quasi-relative definition is possible, depending on how a country's spending on necessities evolves as its income increases.

¹¹ This Table uses the official Supplemental Poverty Measure statistics published by the Census Bureau (Short 2013), whereas for historical comparisons below we rely on historical estimates produced by Wimer et al. (2013), described below. The series from Wimer et al. anchors its poverty measure at the 2012 SPM thresholds, so their estimated poverty rates for 2012 are very similar to those in Table 1.

¹² Accounting for out-of-pocket medical expenses raises measured poverty for nonelderly adults and children by 2.9 and 3.1 percentage points, respectively.

¹³ See http://www.ers.usda.gov/topics/rural-economy-population/rural-poverty-well-being/geography-of-poverty.aspx#.UurSXhBdXA0

¹⁴ This comparison may be affected by significant differences among geographic areas in costs other than housing, such as transportation. In addition, there may be differences in housing quality that are not captured by the differences in housing costs.

¹⁵ The figure updates an analysis in Lemieux (2008), who graciously shared data.

¹⁶ Retrieved from http://www.bls.gov/cps/cpslutabs.htm, January 30, 2014.

¹⁷ Like all decompositions, the reference period matters. If the decomposition is performed using 2009 poverty rates rather than those in 1970, the predicted increase is 2.8 percentage points.

¹⁸ Research suggests the increase is driven primarily by increased sentencing severity rather than increases in criminal activity (Caplow and Simon 1999, Nicholson-Crotty and Meier 2003).

¹⁹ The "anchored poverty" measure allows progress to be measured against a constant definition of living standards. Using the SPM methodology of updating poverty thresholds each year to reflect rising expenditures on food, clothing, shelter, and utilities shows less of a decline in poverty since the real value of the poverty thresholds rise over time (Fox et al. 2013). Data constraints prevent Wimer et al. (2013) from following the SPM methodology exactly. The most important discrepancy from the Census procedure is that Wimer et al. do not adjust the poverty thresholds for geographic differences in living costs. It is worth noting that an alternative measure anchors poverty thresholds based on necessary expenditures in 1967, and adjusts for inflation each year afterwards. Both measures show similar declines in poverty, but using expenditures in 2012 gives a higher level of poverty in every period since increased real spending on necessities over time has led to a higher SPM poverty threshold.

²⁰ While this is a "static" exercise in that it assumes that individual earnings themselves are not affected by the existence of safety net programs, this chapter later reviews research on the effects of programs on employment and earnings and finds that such effects are generally small where they exist, and not large enough to meaningfully alter the conclusions from this simplification (Ben-Shalom, Moffitt, and Scholz 2010).

²¹ There are two counterfactuals estimated by Wimer et al. that are used in this report to discuss the impact of the safety net. For most estimates of the impact of the safety net we use a counterfactual poverty rate that strips away ("zeroes out") all cash and in-kind transfers, as well as refundable tax credits but continues to subtract any "normal" tax liability from family resources. In this section, we define "market poverty" similarly, only this measure additionally zeroes out all tax liabilities. Since poor families near the poverty line tend to have positive tax liabilities, market poverty rates are slightly -- about 1.8 percentage points in 2012 -- lower (since we assume families can keep the taxes they in fact must pay).

²² As described earlier, the underreporting of government income and benefits means that this is likely to be a conservative lower-bound estimate of the effect of the safety net.

²³ For example, Meyer and Rosenbaum (2001) suggest that EITC expansion accounts for about 60 percent of the roughly 10 percentage point rise in single mothers' employment rates relative to single mothers without children between 1984 and 1996.

²⁴ Unlike the earlier literature, they also find evidence of a slight downward adjustment of earnings for workers near 200 percent of the poverty line. This is consistent with the predictions of static labor supply theory since that level of earnings is in the phase-out region of the credit. This negative impact is small, however, and Chetty, Friedman, and Saez emphasize it is dominated by the (also small) positive impacts at low earnings levels.

²⁵ The stability of intergenerational mobility is also shown by comparing the correlation of parent and child income ranks over this period (Chetty, Hendren, Kline, Saez, and Turner 2014).

²⁶ Using a different method to assess absolute mobility, about half of individuals moved out of the bottom income quintile in the 1984-94 and 1994-2004 periods when the quintile thresholds were fixed at the beginning of the period (Acs and Zimmerman, 2008).

²⁷ Estimates of the effects of the Recovery Act on poverty have been updated relative to our January 2014 report to reflect minor methodological improvements.

²⁸ Based on CEA calculations using data from http://www.whitehouse.gov/sites/default/files/omb/budget/fy2014/assets/hist.pdf.

This Administration has supported the use of rigorous, high-quality "impact" evaluations to measure changes in a variety of outcomes targeted by Federal programs, ranging from earnings to health to electricity usage. Many factors affect whether Federal programs achieve their goals, and identifying impacts specifically attributable to programs is challenging. An impact evaluation is a particular type of program evaluation, and aims to measure the causal effect of a program or intervention on important program outcomes. This chapter focuses on impact evaluations. "Process" evaluations (another type of program evaluation) and performance measurement also contribute to building evidence about how well programs are working, but differ in important ways from impact evaluations (Box 7-1).

Building on the efforts of previous administrations, the Obama Administration is working to reform the Federal Government's approach to improving program performance. In addition to emphasizing transparency and accountability in tracking progress toward agencies' priority goals, this new approach also aims to complement and to draw on the Administration's program evaluation efforts. For example, the Administration this year is establishing strategic reviews within agencies to strengthen the use of evidence in strategic and budget decisions.

This chapter provides an overview of the implementation and use of impact evaluation in Federal programs, with a special focus on the lessons learned so far in this Administration. It begins with a discussion of some challenges inherent in conducting rigorous impact evaluations in government programs. The chapter then focuses on Administration efforts to build and to use evidence, including actions taken on lessons learned from completed evaluations, launching new evaluations in areas where not enough is known, and creating a culture of evidence-building in Federal programs, especially grant programs. The final section identifies opportunities for further progress: for example, through increasing legislative support and removing legislative barriers, embedding evaluation into routine program operations, and using existing program data to measure outcomes and impacts.

• Program managers use many approaches to assess how programs operate and how well they work. Impact evaluations aim to identify the causal effects of a program or intervention on some outcome or outcomes of interest. Impact evaluations are distinct from other types of program evaluation and performance measurement. For example:

• Process evaluations analyze the effectiveness of how programs deliver services relative to program design, professional standards, or regulatory requirements. For instance, a process evaluation might focus on whether a program is reaching the target number of participants or whether caseworkers are consistently following a specified protocol for providing services. Process evaluations help ensure that programs are running as intended, but in general these evaluations do not directly examine whether programs are achieving their outcome goals (GAO 2011).

• Performance measurement is a broader category that encompasses "the ongoing monitoring and reporting of program accomplishments, particularly progress toward pre-established goals" (GAO 2011). Typically, performance measures provide a descriptive picture of how a program is functioning and how participants are faring on various "intermediate" outcomes, but do not attempt to rigorously identify the causal effects of the program. For instance, performance measures for a job training program might capture how many individuals are served, what fraction complete the training, and what fraction are employed a year later. But these measures will not answer the question of how much higher these individuals' employment rates are as a result of having completed the training. Nonetheless, performance measures serve as important indicators of program accomplishments and can help establish that a program is producing apparently promising (or troubling) outcomes.

• While process evaluations and performance measurement are useful at all stages of a program's maturity, they can be particularly useful for providing evidence about how programs are working in the early years of a program's history when impacts on program outcomes may not be detectable and rigorous, high-quality impact evaluations are not possible. A logic model -- a tool that depicts the intended links between program investments and outcomes and helps to ensure program activities will achieve desired outcomes -- can facilitate agency efforts to develop high-quality "intermediate" indicators of impacts as well as an understanding of alternative causal channels that can affect important program outcomes.

One basic approach to answering questions like these is to look at outcomes before and after the "treatment" -- for instance, before and after taking a drug, before and after a new marketing strategy is rolled out, or before and after participation in an education program. Another straightforward approach is to compare outcomes for program participants with outcomes for non-participants. In complex policy environments, however, these simple approaches will often give the wrong answers. Take, for example, a job training program designed to help unemployed workers get jobs. The data may show that program participants were much more likely to be employed a year after the training program than before they entered the program. But if the unemployment rate has fallen substantially over the course of the program, then the gains may be due to the improving economy, not to the training program. Similarly, a government program offering start-up assistance to new businesses may appear to boost success rates. But if capable entrepreneurs are more likely than less capable ones to participate in the program, then self-selection of program participants, not the program itself, may be driving those better outcomes.

A strong impact evaluation needs a strategy for constructing more valid comparisons -- specifically, for identifying "treatment" and "control" groups for which differences in outcomes can reasonably be attributed to the program or intervention rather than to some other factor. Impact evaluations conducted using rigorous, high-quality methods provide the greatest confidence that observed changes in outcomes targeted by the program are indeed attributable to the program or intervention. It is well recognized within Congress and other branches of government (for example, GAO 2012, National Research Council 2009), in the private sector (Manzi 2012), in non-governmental research organizations (Coalition for Evidence-Based Policy 2012, Walker et al. 2006), and in academia (for example, Imbens 2010; Angrist and Krueger 1999; Burtless 1995) that evaluations measuring impacts on outcomes using random assignment provide the most definitive evidence of program effectiveness.

Although the classic impact evaluation design entails random assignment of recipients into treatment and control groups as part of the experiment, the goal of constructing valid comparisons sometimes can be achieved by taking advantage of natural variation that produces as-if randomness, an approach referred to as a quasi-experiment. Quasi-experiments can sometimes be much less expensive than traditional large-scale random assignment experiments, and are discussed further below.

Estimation of Causal Effects of a Program or Intervention

The starting point for estimating the causal effect of a program or intervention is being precise about what constitutes a causal effect. Consider a treatment delivered at the individual level: either the individual received the treatment, or did not. The difference between the potential outcome if the individual received the treatment and the potential outcome if the individual did not is the effect of the treatment on the individual.¹ The challenge of estimating this treatment effect stems from the fact that any given individual either receives the treatment or does not (for example, a child either does or does not attend preschool). Thus, for any given person only one of two potential outcomes can be observed. The fact that we cannot directly observe the counterfactual outcome (for example, the earnings a person who went to preschool would have had if they had, in fact, not gone to preschool) implies that we cannot directly measure the causal effect. This problem of observing only one of the potential outcomes for any given individual is the fundamental problem of causal inference (Holland 1986).

Randomization provides a solution to the problem of not observing the counterfactual outcome. If individuals are randomly assigned to treatment and control groups, then on average the individuals in the two groups are likely to be the same in terms of other characteristics that might affect outcomes. As a result, one can safely assume that ex-post differences between the groups are the result of the treatment. To take the preschool example, simply comparing test scores of all U.S. elementary school children who had attended preschool to all U.S. elementary school children who had not would not provide confidence that higher test scores for the first group were an effect of preschool. The scores might reflect differences in family background, elementary school resources, or other important factors between the two groups. On the other hand, if a group of three-year olds are randomly assigned to attend or not attend preschool, and the preschool group has higher test scores in third grade, we can attribute the test score gains to attending preschool because the two groups would not be systematically different along other dimensions that might impact learning.

In most cases, simple comparisons of treated and untreated individuals without random assignment will not produce valid comparisons because treatment status will be correlated with other important factors. For example, if potential preschool enrollees were initially screened so that those with the least learning opportunities outside school were placed in the program, then we might find that the treatment group (enrollees) has worse outcomes than the control group. However, the reason for this finding is that enrollees are more disadvantaged than non-enrollees. The variation between treatment and control groups affects ultimate outcomes both through the treatment and the differences in learning opportunities outside school. Thus, any comparison of outcomes between treatment and control groups would measure the combined effects of both the treatment and those differences in learning opportunities.

Because randomized experiments can be expensive or infeasible, researchers have also developed methods to use as-if random variation in what is known as a quasi-experiment. The necessary condition for a high-quality quasi-experimental design is that people are assigned to a treatment or control group in a way that mimics randomness. This can be done by forming treatment and control groups whose individuals have similar observable characteristics, and exploiting some rule that governs assignment to the treatment and control groups in a way that is plausibly unrelated to the outcome of interest.

One example of a quasi-experimental design that lends itself to estimating impacts of programs or interventions is when eligibility is determined based on one or more variables in a way that individuals who (just) qualify for the program are very much like those who (just) do not. If so, and if both eligible and ineligible applicants are tracked, then a method called regression discontinuity design can be used to compare the outcomes for individuals on the two sides of the threshold, controlling for other observable differences between the two groups.

Another example of quasi-experimental design is when a program varies across units for reasons unrelated to the program outcomes. Rothstein (2011), for example, exploits the fact that, due to different business cycle patterns combined with policy variation created by expirations and renewals of the Emergency Unemployment Compensation (EUC) program during the Great Recession, the number of available weeks of benefits available to job-seekers varied dramatically from month to month in differing ways across states. After controlling for local economic conditions, the haphazard nature of the changes in EUC benefit levels across states enabled estimation of EUC benefits on job-finding rates.

Describing the whole range of quasi-experimental approaches is beyond the scope of this chapter.² Quasi-experiments require stronger assumptions than randomized experiments and the debate around those assumptions makes it harder for quasi-experiments to be convincing, especially to non-experts. However, if the quasi-experimental variation used is plausibly unrelated to the outcomes of interest except through the treatment, quasi-experimental evidence can be convincing, with some methods and applications being nearly as compelling as randomized trials and others leaving more room for doubt.

Other Criteria for High-Quality, Successful Impact Evaluations

A strong impact evaluation also needs to address questions that are actionable and relate to outcomes that matter. In some cases, the actionable information might identify if a program is or is not effective. In other cases, the actionable information might identify which interventions are best at achieving important program outcomes, so that programs can be improved by adopting successful interventions more broadly. However, if there are legal or other impediments to expanding an evaluated small-scale intervention, then learning that the intervention works does not directly lead to an action that can improve a program at a national level. In such cases, it may be better to allocate scarce evaluation resources to testing more modest interventions or ways to run the program more effectively.

For the second of these criteria -- outcomes that matter -- the long-term goals of a program must be considered. For a preschool program, the number of students enrolled is an easy-to-measure intermediate outcome. However, preschool enrollment may or may not be related to ultimate outcomes, such as high school graduation rates, employment rates, or income. It is also important to consider program size and stage of development, as programs or interventions must be sufficiently mature, and treatment and control groups sufficiently large, to obtain credible estimates of impact.

Other issues must also be addressed to conduct policy-relevant impact evaluations in government programs. At the most practical level, rigorous evaluation requires adequate funding, staff expertise, and often cooperation across different parts of an agency (or across multiple agencies). Rigorous evaluation also requires support from top agency management and program managers. Further, many Federal programs have multiple goals, which can make it hard to take action on evaluation findings when the results support some goals but not others.

An important part of evaluating a program is remaining open to the findings, regardless of the outcome, to inform the best course of action to improve outcomes going forward. Findings of positive impacts provide important feedback that may indicate whether additional investment is warranted. Findings of no impact, either for all participants and program goals or for important subsets of the participants and program goals, also send valuable signals that modifications -- including reallocating program funding to other strategies that could better achieve outcomes -- are needed.

Lower-Cost Ways for Impact Evaluations to Facilitate Real-Time Learning

Large-scale random-assignment studies of social programs have been very influential, but also can be quite expensive, and their expense has been a major impediment to wide-scale adoption of learning and program improvement through randomization. For this reason, researchers have focused on lower-cost methods for learning about program effectiveness.

One lower-cost method is to build randomization into the design of the program, so that data on program performance can be tracked and evaluated on an ongoing basis. This strategy has been pioneered as a management tool in the private sector for ongoing product and process improvement. Indeed, some companies run thousands of randomized studies annually: by 2000, Capital One was running 60,000 studies annually using randomization methods, as they experimented with different strategies to determine what works. Google has also run randomized experiments in the tens of thousands in some years (Manzi 2012).

In the public sector, Federal agencies are also finding ways to conduct high-quality evaluation strategies at lower cost, including ways that employ the lessons learned from behavioral economics (Box 7-2). The U.S. Department of Agriculture's Food and Nutrition Service is conducting a range of rigorous demonstration projects to further develop the evidence base of effective strategies for programs that address food insecurity and improve nutrition among children; one such project implements low-cost environmental changes in lunchrooms to encourage students to make healthier food choices. One demonstration found that merely placing fruit in a colorful bowl in a convenient part of the lunch line can lead to an increase in fruit sales of up to 102 percent (Wansink, Just, and Smith 2011). Funding research for these simple, evidence-based interventions allows for the development of effective strategies to strengthen the nutrition and hunger safety net for the more than 30 million children fed by the National School Lunch Program.

Utilizing existing data and independent programmatic changes to measure outcomes is another strategy that agencies are using to minimize evaluation costs. For example, the Department of Justice's National Institute of Justice conducts impact evaluations of interventions that can help inform the approximately 18,000 local law enforcement agencies that do not individually have the resources to test interventions on their own. Hawaii's Opportunity Probation with Enforcement (HOPE) program was established as a demonstration pilot for drug-involved probationers in Hawaii. The pilot tested the efficacy of "swift and certain" sanctions against probationers who fail to meet the conditions of their probation. The randomized controlled experiment found that after one year, probationers who received very frequent drug testing (every other day) and -- if they failed the drug test -- an immediate court date and a modest but certain sanction (a night in jail), were 72 percent less likely to use drugs, 61 percent less likely to skip meetings with their supervisory officers, 55 percent less likely to be arrested for a new crime, and 53 percent less likely to have their probation revoked. These reductions led to HOPE participants being sentenced to an average of 48 fewer days in prison than those in the control group who received the traditional delayed but more severe sentence (National Institute of Justice). Because of the high costs associated with servicing inmates in prison, any intervention that reduces prison time can generate large savings. By making use of available administrative data, evaluations employing quasi-experimental and randomized controlled trial designs were implemented at a cost of only $150,000 and $230,000, respectively.³ A follow-up analysis is examining the long-term impacts of the intervention; this model is also being piloted in four other locations.

Uses of Evaluation

Agencies have used impact evaluations to inform policy and program decisions in a wide variety of ways.

Making the Unemployment Insurance System More Effective. Unemployment insurance (UI) provides an important safety net for workers who become unemployed. Occasionally, concerns are raised that UI payments could reduce an unemployed worker's incentive to find employment. While the evidence suggests that any such effects are small (Council of Economic Advisers and the Department of Labor 2013), the Federal Reemployment and Eligibility Assessment (REA) initiative started providing funds in 2005 to states and sought to reduce UI duration by combining in-person UI eligibility reviews with (1) labor market information, (2) developing a reemployment plan, and (3) offering a referral to reemployment services. The Department of Labor funded research using a randomized design that showed the REA initiative was effective in reducing the duration of UI (Benus et al. 2008). However, these studies focused on measuring reduced duration on UI and associated costs and not on other outcomes, such as return to employment or increased wages. These studies were followed by another randomized controlled trial which showed that the REAs were also effective at reducing joblessness when eligibility assessments were personalized and more closely integrated with the delivery of reemployment services (Poe-Yamagata et al. 2011). Consequently, the Administration proposed in the American Jobs Act to create a requirement that all Emergency Unemployment Compensation claimants receive both an REA and reemployment services; this was enacted in the Middle Class Tax Relief and Job Creation Act of 2012.⁴ Evidence from rigorous evaluations is playing a role in making the REA initiative more effective and getting unemployed Americans back to work faster, and the Administration has sought to expand it to cover more workers. A modest increase in funding for REAs was included in the recently enacted Consolidated Appropriations Act of 2014.

The study's findings helped spur many important policy changes. Most notably, students and their families now have the option to pre-populate the FAFSA with the income information they have already provided the Internal Revenue Service on their tax returns, similar to the arrangement the researchers tested with H&R Block. This simplifies FAFSA completion for students, lowers the risk of errors, and as such should increase access to college among socioeconomically disadvantaged students and should lead to gains in college enrollment. As a complement, the Department of Education has also simplified the FAFSA application to make it easier to complete for all applicants, but especially for low-income students. In 2012, the Department of Education awarded a second grant to the research team to test the effects of FAFSA simplification at scale. The evaluation will use an experimental design and involve 9,000 tax-filing sites across the United States.

Institutionalizing Evidence-Based Decision-making in Grant Programs. In many programs, funds are distributed to states and local entities through competitive and formula grants. Grants to State and local governments have constituted roughly one-third of total outlays over the last 20 years, so increasing the use of evidence in informing policy in grant-supported programs could improve outcomes for a significant portion of outlays (Figure 7-1). Many effective program structures treat evaluation as an essential element in the decision-making framework, while also building in opportunities to scale up approaches that work and scale back or eliminate approaches that do not. As stated by then-OMB Director Peter Orszag, new initiatives should ideally have "evaluation standards built into their DNA" (OMB 2009b). The Administration has experimented with several models that embed both evidence building and evidence-based decisionmaking into the "DNA" of grant programs.

During this Administration, several initiatives have adopted a "tiered evidence" approach that embeds evidence-based decision making into program structure. Tiered evidence programs tie grant funding to the evidence base behind proposed interventions. In these programs, interventions that provide better evidence of success move to higher tiers and become eligible to receive more funding for expanded implementation and evaluation. The built-in mechanism for scaling up interventions that work helps prevent the troubling problem of not investing in programs with proven high returns.

A successful example of a three-tier approach is the Investing in Innovation program at the Department of Education. This program provides seed development grants of up to $3 million for high-potential and relatively untested interventions, validation grants of up to $12 million for interventions based on only a moderate amount of evidence, and scale-up grants of up to $20 million for potentially high-impact, transformative education interventions. Evidence of effectiveness is an "entry requirement" for validation and scale-up grants, and all grantees are expected to conduct an evaluation that will add to the evidence base on effectiveness. For the scale-up and validation grants, the grantee must make the data from their evaluations available to third-party researchers, consistent with applicable privacy requirements (Department of Education 2013a).

 

Figure 7-1

 

Outlays for Grants to State and Local Governments, 1992-2012

 

 

 

Source: Office of Management and Budget, FY 2014 Budget, Tables 3.1 and 12.2.

Similarly, the Maternal, Infant, and Early Childhood Home Visiting Program in the Department of Health and Human Services (HHS) was an early Administration initiative that uses a two-tiered evidence structure. Implemented in 2010 as part of the Affordable Care Act, this voluntary home visiting program uses trained professionals and paraprofessionals to provide support to vulnerable pregnant women and parents of young children to improve health, development, and well-being outcomes for at-risk children and their families. The Act required that at least 75 percent of the home visiting program funds be spent on proven, evidence-based approaches and allowed for the remainder to be spent on promising approaches as long as they are rigorously evaluated. Currently, 14 home visiting models meet the HHS criteria of "evidence-based approaches," and have been evaluated with a mix of randomized experiments and quasi-experiments using multiple measures of key outcomes (Paulsell et al. 2013). While the Act funded the home visiting program through 2014, the Administration has proposed to continue funding and expand the availability of voluntary evidence-based home visiting programs to reach additional families in need as part of a continuum of early childhood interventions.

In addition to tiered evidence structures, agencies have begun using other designs in competitive grant programs that encourage the use of evidence-based practices. One such design is the "Pay for Success" approach. In this performance-based model, philanthropic and private funding is leveraged and the government provides payment only after targeted outcomes are achieved. In 2012 and 2013, the Administration started supporting programs that use a Pay for Success model to fund preventive services, and which had outcomes that could be measured with credible evaluation methodologies. The first Pay for Success awards were for projects to prevent prison recidivism.⁵ The Consolidated Appropriations Act of 2014 authorized up to $21.5 million for Pay for Success projects.

Even in more traditionally structured grant programs where funding is provided upfront, agencies are embedding more rigorous evaluation requirements into funding requirements. For example, upfront grants in the Department of Labor's Workforce Innovation Fund, first issued in 2011, fund promising but untested employment and training service and administrative strategies. These grants also fund well-tested ideas being adapted to new contexts as a way to significantly increase evidence about interventions that generate long-term improvements in public workforce system performance, such as reduced duration of unemployment. Grantees are required to conduct rigorous evaluations, and a national evaluation coordinator works with grantee evaluators to ensure consistent and high-quality evaluations (Department of Labor 2011).

Ending or Reducing Funding for Interventions or Programs. The Administration's commitment to evidence-based evaluation means terminating or reducing funding for a program when a body of evidence consistently shows that the program is not achieving its stated goals, helping to reduce the use of taxpayer dollars on ineffective programs. The FY 2012 Budget took this approach with the Mentoring Children of Prisoners (MCP) program run by the Department of Health and Human Services. Rigorous evaluations show that high-quality mentoring relationships lasting for at least 12 months can have positive impacts on youth, while relationships that do not last more than three months can actually have harmful effects on youth (Grossman and Rhodes 2002). According to the MCP program performance data, fewer than half of program participants each year were in matches that lasted at least 12 months and, in 2008 alone, as many as 27 percent of matches that ended prematurely ended within three months. An evaluation of one MCP-funded program suggested that premature terminations were the result of program performance and were independent of the demographics of the participants (Schlafer et al. 2009).

Interpreting the MCP performance data in light of the evidence from impact evaluations of other mentoring programs, the Administration concluded that the MCP was not as effective as it should be. As a result, the Administration proposed to reduce funding for the MCP, noting that other competitive grant programs could serve the youth targeted by the MCP, and that some of those programs, such as Promise Neighborhoods, utilize evidence-based practices. Congress ultimately eliminated funding for the program in the Continuing Appropriations Act of 2011.

Even Start, originally designed to improve family literacy in disadvantaged populations, was another program not meeting its stated goals that the Administration took steps to replace. While the literacy levels of Even Start children and parents improved, multiple national randomized experiments showed that parents and children in control groups who did not participate in Even Start (one-third of whom received other early childhood education or adult education services) had comparable improvements (see for example St. Pierre et al. 2003). The President's FY 2012 Budget proposed, and Congress approved, the elimination of separate funding for Even Start. The Administration has proposed incorporating it and other narrowly focused literacy programs into the newly created literacy component of the Effective Teaching and Learning program that would support competitive grants to states for high-quality, evidence-based literacy programs.

Building Evidence when Existing Evidence is Limited

In many of the examples highlighted above, evidence existed on what programs or interventions were most effective, and the key challenge facing policymakers was to act on that evidence. However, not enough is known about what works in many other important areas, and so the first step in evidence-based policymaking is to invest in developing evidence.

Reducing Electricity Use. Experts have long suggested time-varying pricing (more costly at times of peak demand) as a way of increasing the efficiency of electricity use, including reducing electricity demand. Such time-varying pricing could increase efficiency, defer investments in expensive new power plants, and reduce pollution. However, most electricity delivery systems have not invested in the in-home technologies necessary to allow residential consumers to respond to time-varying prices. In addition, regulators have been hesitant to approve varying rates, and private companies have been reluctant to invest in modernizing their systems without knowing whether time-varying pricing will significantly impact consumer behavior. In recent years, the Federal Government, in partnership with states and utilities, invested in evaluating the impact of time-varying pricing on consumer behavior so that this information would be available to utilities, regulators, and states. These consumer behavior studies were implemented with American Recovery and Reinvestment Act funds and use randomized controlled experimental methods. Deciding which type of pricing strategy to use falls within State jurisdiction, rather than Federal, so these studies will allow State and local public utilities to make more informed decisions on pricing models (Cappers et al 2013). While these studies are still ongoing, two utilities and their regulators have decided to implement time-varying rates across their service territories based on the results observed to date. Such efforts can serve as an impetus to get more public utilities to adopt time-varying pricing.

Improving Health Care Delivery. In another example, the Affordable Care Act made a number of major investments in understanding how to improve quality and reduce cost in health care delivery, in addition to expanding access to affordable health insurance coverage. As described earlier in this chapter, the Center for Medicare and Medicaid Innovation (the "Innovation Center"), created by the Act, is using high-quality evaluation approaches to test innovative payment and service delivery models designed to reduce expenditures while preserving or enhancing quality of care for Medicare, Medicaid and Children's Health Insurance Program beneficiaries. Several ongoing Innovation Center payment reform initiatives -- and early results from those initiatives -- are discussed in Chapter 4. The Innovation Center will use the results of such model evaluations and actuarial data to identify best practices and determine which successful models could be implemented more broadly.

Better Outcomes for Youth with Disabilities. The Administration is also testing many different approaches aimed at youth with disabilities. The Promoting Readiness of Minors in Supplemental Security Income (PROMISE) is a joint initiative of the departments of Education, Health and Human Services, Labor, and the Social Security Administration. PROMISE aims to improve the education and employment outcomes for youth with disabilities who receive Supplemental Security Income (SSI) and their families, by improving coordination of services such as those available through the Individuals with Disabilities Education Act, the Vocational Rehabilitation State Grants program, Medicaid health and home and community based services, Job Corps, Temporary Assistance for Needy Families, and Workforce Investment Act programs. The PROMISE program allows grantees (states or consortia of states) to design their own intervention models to serve youth and their families for three years with a two-year extension option, provided they include a minimum set of services. Grantees may also apply for waivers of funding restrictions or rules in individual programs that they believe will constrain their ability to achieve outcomes. Grantees agree to enroll a large number of youth (around 2,000) who are eligible to be served by a PROMISE intervention, and to allow random assignment to be used to assign half of eligible youth to the treatment group and the remaining youth to a control group that receives the services that child SSI recipients normally receive. The first grants were awarded in September 2013. To evaluate whether PROMISE can help child SSI recipients achieve better outcomes, a national evaluation will be conducted of all grantees to analyze intervention impacts on educational attainment, employment credentials and outcomes, and whether the interventions reduce long-term reliance on public benefits, and SSI payments in particular (Social Security Administration 2013).

Improving Outcomes For At-Risk Youth. The Administration also is working to identify approaches that help at-risk youth. The National Guard Youth Challenge (ChalleNGe) program, which has been rigorously evaluated, is designed to provide opportunities for adolescents who have dropped out of school but demonstrate a willingness to turn their lives around. Using random assignment, Millensky et al. (2011) found significant benefits to program participation in addition to higher earnings, as ChalleNGe graduates were more likely than the control group to have obtained a high school diploma or GED, to have earned college credits, and to be working three years after completing the program. Participation was projected to increase discounted lifetime earnings by over $40,000 (in 2010 dollars) (Perez-Arce et al. 2012). After considering education costs to the student and other non-earnings benefits, the ChalleNGe program was estimated to generate $2.66 for every dollar of program cost (Perez-Arce et al. 2012). The Administration now plans to test the application of the ChalleNGe model to adjudicated youth, through the Department of Labor's Reintegration of Ex-Offenders program.

Reducing Homelessness. Sharply reducing homelessness is a key focus of the Administration.⁶ Although once considered an intractable problem, a broad body of research (including rigorous evaluations) documented that there are models that effectively serve individuals experiencing chronic homelessness. The Department of Housing and Urban Development (HUD) has invested heavily in promoting these evidence-based approaches, and has re-oriented the Homelessness Assistance Grant Program away from such traditional approaches as transitional housing and toward more-effective permanent supportive housing (Figure 7-2). Because research on interventions that are effective for homeless families does not yet exist at the same level of rigor as for homeless individuals (Culhane et al. 2007), HUD has undertaken an experimental study of family homelessness called the Family Options Study. This study will compare several combinations of housing assistance and services in a multi-site experiment to determine which interventions work best to promote housing stability, family preservation, child well-being, adult well-being, and self-sufficiency. In addition to usual care, defined as remaining in emergency shelter and accessing whatever resources that would normally be available to families in shelter, three interventions are being studied: 1) subsidy only (a voucher primarily), 2) transitional housing, 3) rapid re-housing.⁷

Instilling a culture of evidence-based decision making within agencies, and building the foundations that enable rigorous evaluations to guide new investments and drive policy, is neither quick nor easy. Evaluations of particular interventions or entire programs should not be isolated exercises that occur on an ad hoc basis, but rather planned in advance. Challenges always accompany efforts to enact significant change, but addressing several key elements can greatly facilitate agency efforts to improve the collection and use of evidence. While not a comprehensive list, these issues represent several major areas that provide either useful opportunities, or serve as barriers, in agency efforts to build and advance an evidence-based agenda.

 

Figure 7-2

 

 

Inventory of Beds for Homeless and Formerly Homeless People,

 

2007-2012

 

 

 

Legislative Support for Evaluation

Authorizing legislation and appropriations bills can direct how agencies should use program funds for a wide variety of activities. Legislation can encourage stronger and more cost-effective evaluation in many ways. One is through language that recognizes the importance of conducting rigorous evaluations. Another is by making sure already-collected program data are made available for such statistical and analytical purposes.

Legislative Support for Rigorous Evaluations. Two ways that legislation can support rigorous evaluation is through set-asides and support for evaluation of demonstration programs. In recent years, with support of top management within agencies and the Administration, several agencies have had set-asides for evaluation specified in program legislation and appropriations. For example, the Consolidated Appropriations Act of 2012 first enabled the Secretary of Labor to reserve up to 0.5 percent of specific Department of Labor (DOL) appropriations for evaluations. Also, a set-aside of 5 percent of competitive grant funds in the Teacher Incentive Fund allows the Department of Education to conduct a rigorous national evaluation of the program and to share with grantees the results of current, rigorous research to help facilitate ongoing improvement. Additionally, authority to set-aside a percentage of program funds for evaluation is specified in some HHS and HUD programs, including several that received additional funding through the Affordable Care Act of 2010 and the American Recovery and Reinvestment Act of 2009.⁸

Many programs are funded through annual appropriations, which generally require obligation of those funds within a given fiscal year. However, the DOL set-aside for evaluation in the Consolidated Appropriations Act extends the deadline by which the DOL must obligate transferred evaluation funds to two years. Because designing rigorous evaluations takes time, a window beyond the standard one-year for obligating evaluation funds can in some cases enable agencies to plan and execute more thorough, higher-quality evaluations.

Legislation that specifies funding for demonstration pilots also provides important support for developing an evidence base. The legislatively authorized demonstrations being conducted by the Innovation Center are recent examples that illustrate the value of legislative support for evidence building.⁹ As another example, the Department of Health and Human Services' child welfare waiver authority allows states to design and demonstrate a wide range of approaches for reforming child welfare and improving outcomes, including decreased first-time entries and reentries into foster care, and improvements in various aspects of child developmental, behavioral, and social functioning. States are required to conduct rigorous impact evaluations as well as process evaluations as part of their waiver agreements.¹⁰ In addition, the Administration is proposing to restore demonstration authority for the Disability Insurance program, while also providing new authority for the Social Security Administration and partner agencies to test early-intervention strategies that would help people with disabilities remain in the workforce.

Legislation can also encourage stronger evaluations through explicit language requiring grantees to participate in evaluations and by requiring use of proven interventions. The Healthy, Hunger-Free Kids Act, for example, included a nondiscretionary provision that requires State and local grant recipients in a number of nutrition assistance programs, including the National School Lunch Program, the Special Supplemental Nutrition Program for Women, Infants and Children, and other programs authorized in the National School Lunch Act and the Child Nutrition Act to cooperate in evaluations conducted by or on behalf of the Department of Agriculture.¹¹ This Act also reformed the structure of the nutrition education provided through the Supplemental Nutrition Assistance Program, one of the Nation's main anchors of the social safety net that provides nutrition assistance to eligible low-income individuals and families. It established a new and improved Nutrition Education and Obesity Prevention Grant Program that requires a greater emphasis on evidence-based, outcome-driven interventions, with a focus on preventing obesity and coordinating with other programs for maximum impact and cost-effectiveness.

Legislative Support for Access to Data for Statistical Purposes, Including Evaluations. Existing laws can be explicit or implicit regarding whether information collected as part of administering programs can be used for statistical purposes integral to evaluation. Explicit and supportive laws can save significant time and effort in negotiating agreements to provide data for evaluations and can facilitate more and better analysis. For example, the Social Security Act explicitly states that one of the agency's datasets can be used for statistical and research activities conducted by Federal and State agencies.

Some legislation provides the agency head with broad authority to determine appropriate uses of program data. Given that the statistical uses of data in program evaluation often inform the context, policies, and operations of the same programs authorized by a given statute, agencies sometimes determine that their general statutory authority can grant sufficient authorization to provide administrative data to other Federal agencies for statistical purposes. For example, the Social Security Administration provides certain datasets for statistical and research purposes as described in its implementing regulations.

Multiple legitimate goals must be balanced when determining appropriate access to data, including reducing the burden of data collection on individuals and institutions and protecting personal privacy. Even so, careful crafting of legislative language can achieve those aims while still making data available for Federal researchers to rigorously evaluate and to improve government programs. Key considerations include: avoiding vague or unclear authority to determine appropriate uses of program data; avoiding narrowly written statutory language that only allows access to data for narrowly defined programmatic reasons; or restricting a Federal agency's ability to collect data from grantees.

The information needs in programs managed at the State level could theoretically be addressed through non-Federal data systems, but this is not always possible. States or other grantees may not voluntarily develop comprehensive data systems in ways that are comparable across states, or have the capacity or incentive to make data available to researchers. When no feasible solutions exist to alleviate these issues, legislation may be warranted to authorize creation of Federal datasets accessible to researchers, or to establish requirements for State-held datasets that enable data exchange and comparability across states and to ensure access by researchers.

Building Evaluation into the Design of Programs

Many of the examples described earlier demonstrate the ways in which agencies are designing programs to facilitate evaluation. But agencies can still do more to embed rigorous evaluation designs into both new programs and existing programs.

New programs. The benefits of adopting evidence-based program designs, like the tiered evidence structure in the Investing in Innovation program and in HHS' home visiting program, include the ability to guide competitive grant funds to the strategies with a strong evidence base, while also requiring grantees to conduct evaluations where no evidence is yet available. Even without such a program structure, agencies implementing new programs over the past five years have increasingly required grantees to collect data and develop administrative data systems that can improve comparability and facilitate evaluation in addition to meeting program operating needs. For example, the Department of Education's Promise Neighborhoods initiative implemented in 2010 requires grantees to collect and track outcome data in an individual-level longitudinal data system to facilitate rigorous evaluation. This initiative aims to improve educational and developmental outcomes of children and youth in distressed communities.¹² To assist grantees in collecting high-quality and comparable data, the Education Department is providing grantees with extensive guidance on data collection and reporting (Comey et al. 2013).

Other benefits of considering evaluation needs in the design of new programs include creating opportunities to save time and money by identifying evaluation data up front, minimizing burden on program respondents, and avoiding the loss of information all together that cannot be created too long after the fact. When not considered in the earliest stages of program design, a typical alternative to collecting the information needed for evaluation is to conduct surveys, which requires identifying expertise to design and test the surveys, gaining approval for their use, and then administering them to collect data, often long after the fact. Surveys add to the time and cost to build evidence, due to the time and skill involved in developing survey instruments that will yield high-quality data, the requirements for obtaining needed approvals, and the actual implementation of the survey.¹³ Careful planning can help limit the need for evaluation-related surveys to data that cannot be obtained in any other way, such as information on post-program choices, earnings, or jobs necessary for identifying longer-term impacts of a program or intervention.

One of the most important ways the design of a program can facilitate evidence building is through careful consideration of how treatment and control groups can be established to facilitate impact evaluation. As discussed earlier, randomly assigning potential program participants to treatment and control groups enables the most credible impact evaluations. Several mechanisms exist for creating good comparison groups that allow for experimental or quasi-experimental techniques to be employed to produce high-quality estimates of program effectiveness.

Several options for enrolling potential participants in a program or intervention, presented in order of the rigor of evaluation they might support, are as follows:

1. Random assignment by lottery when capacity is limited. In many instances, due to limited funds or other constraints, a program or intervention cannot serve every person or entity that is eligible to apply. In such cases, rather than "first-come, first-serve" or other nonrandom devices, implementing a lottery to select which applicants may participate in a program or intervention generates a low-cost randomized experiment. This strategy has been used recently to determine the impact of Medicaid access (Baicker et al. 2013), charter school attendance (Abdulkadiroglu et al. 2011), and small business entrepreneurship training (Benus et al. 2009). Note that the losers of the lottery need to be followed to track their outcome data.

2. Assignment based on a continuous "need score." A common objection to random assignment is that resources should be targeted to those with the greatest need, or those most likely to benefit. In this situation, program assignment might lend itself to a strong evaluation if it incorporates some sort of explicit, continuous, ranking of applicant need (or likely benefit), and bases program eligibility on some cutoff in need. For example, Ludwig and Miller (2007) study the effect of participating in Head Start on mortality rates for children by exploiting the fact that the Office of Economic Opportunity provided technical assistance to the 300 poorest counties in 1965. This created lasting differences in Head Start funding rates for counties with poverty levels just below and just above the poverty rate of the 300th poorest county. With this type of assignment rule, a regression discontinuity design can be used to study the impact of the program. The logic of the design is that individuals with "scores" just above and just below the threshold -- in Ludwig and Miller, living in a county with a poverty rate just above or below the poverty rate of the 300th poorest county's rate -- are likely to be similar to each other in ways that affect their outcomes, except that those just below receive the treatment (in this case, participation in Head Start). This design can deliver estimates of the effect of the program that are similar to randomized experiments.¹⁴

3. Staging the rollout of a large program. If a program will be introduced that will ultimately serve many participants spread across different geographic areas, or schools, or other natural groupings, staggering the rollout across time and space, with the rollout sequence chosen randomly, makes it much easier to evaluate. For example, suppose a mentoring program aimed at increasing college attendance will be introduced in a group of schools and the government hopes to learn about the effect of the program by estimating the change in college enrollment among students at the school before and after the program is introduced. If the program is introduced in only one school district, then any other changes that the school district introduced around the same time might affect the change in outcomes and bias the conclusion. Similarly, if the program is introduced in many different schools but all in the same year, then any other changes in policy, the economic climate, or other macro-economic conditions may be confounded with the treatment effect and thus may "bias" the estimate of the treatment effect. Staggering the rollout of the program over time and space, using randomization and possibly further matching treatment and control units based on observable characteristics, helps to control for these potential biases, and thus allows for better estimates of a program's impact. This strategy has been used by Rothstein (2010) to study the effect of extended unemployment benefits.

The three strategies above create experiments or quasi-experiments that lend themselves to high-quality impact evaluations. In the absence of such devices, evaluators need to acknowledge the differences that do exist between program (or intervention) participants and non-participants and to use statistical techniques like multivariate regression and matching to control for these differences. Since these strategies all attempt to compare the outcomes of program participants and non-participants with similar characteristics, the success of the evaluation will be determined by the availability of good information on the characteristics of the population that are most predictive of the outcome under study, as well as the reasons why individuals choose to participate in a program. However, for these strategies to work, the variation between the treatment and control groups after using statistical or matching techniques to control for differences between these groups must be plausibly unrelated to the outcomes of interest, except through the effect of the treatment. There needs to be some part of that variation between the treatment and control groups that operates like randomization.

Existing Programs. Designing programs to facilitate evaluation may be relatively simpler in new programs than in existing programs, due to program manager reluctance in the latter to trying new strategies, concerns about equity among participants if the control group receives no services, and other reasons. But experiences at several agencies demonstrate these barriers can be overcome. Lotteries for oversubscribed programs are as applicable in longstanding programs as in new ones (see for example the Jacob and Ludwig (2012) study on impacts of housing vouchers). However, increasing opportunities for evidence-based decisionmaking in programs that allocate funds to states on the basis of formulas remain especially challenging, because evaluations and evidence-based funding allocations are not a requirement of States receiving the funds. Waiver authorities or other mechanisms to incentivize evaluations in these programs are only available in a few instances. A control that could prompt State and local grant recipients to do evaluations in these types of programs is a legislated requirement that a certain portion of funds be set aside for evidence-based grants or models of delivery. For example, in the Senate Appropriations Bill for FY 2014, the Substance Abuse and Mental Health Services Administration mental health block grant programs included language defining a 5 percent set-aside for evidence-based grants.¹⁵

There is still work to be done to embed evaluation and evidence-based decision making into more programs. Agencies can focus evaluation efforts in those programs that can help ensure that the agency's most critical program and policy questions are addressed.¹⁶

Developing the Capacity to Link to Other Administrative and Survey Data Sources

Increasingly, agencies are seeking opportunities to improve their evaluation approaches by supplementing their administrative program data with other available government data, where appropriate and while ensuring strong privacy protections. Using pre-existing data collected for other reasons, while maintaining strong privacy protections, provides a number of benefits. Several challenges arise when doing so, and the Administration is taking steps to address these challenges.

Benefits of using existing data resources. Using pre-existing administrative data collected for other reasons, while maintaining strong privacy protections, can help agencies answer important policy questions that could not otherwise easily be addressed with a single program database or survey. Administrative data provides the most complete and accurate source of information on program participation and can provide more accurate data on earnings, test scores, and other outcomes of interest. Indeed, the benefits of using pre-existing administrative data for evaluation and other statistical purposes have been widely acknowledged for some time. Data from multiple sources have been used in a number of impact evaluations, primarily to identify the characteristics of treatment and control groups, identify outcome variables which indicate the impacts of treatments, reduce study costs and reduce the burden on study participants by avoiding the need to collect the data via another survey (Coalition for Evidence-Based Policy 2012; Finkelstein et al. 2012; Bettinger et al. 2009; Jacob and Ludwig 2012). Linked datasets are also facilitating current evidence-building efforts in various agencies, such as in the Department of Health and Human Service's Office of Child Support Enforcement, which is currently implementing a child support-led employment services demonstration project with a random assignment impact evaluation (where treatment consists of extra services under the program, and the control group receives regular services that are available) and a cost-benefit analysis. The planned evaluation will draw on unemployment insurance wage and benefit records, as well as State administrative data on benefits in the Supplemental Nutrition Assistance Program and other public assistance programs, criminal justice system data and other data to more cost-effectively and accurately determine the effectiveness and the true costs and benefits of the program. As another example, HUD and HHS are pilot-testing links between HUD administrative data and HHS Medicare/Medicaid data, to build evidence on opportunities to improve the health of Medicare/Medicaid beneficiaries in HUD-assisted housing as well as the impact of housing assistance on health.

Challenges and Solutions. Nevertheless, accessing administrative data for these statistical uses is challenging. These data are collected to facilitate day-to-day program operations, including developing performance measures. Unless evaluation needs are considered in the database design stage, however, the meaningfulness of administrative data for conducting rigorous evaluation may be limited. Also, data definitions can vary dramatically across datasets; especially with State-level data, the definitions often vary across states and even counties. Aside from definitional differences, the quality of programmatic data -- its completeness and accuracy -- can vary dramatically across datasets. Significant data-quality gaps or errors can compromise analysis. It can also be costly to negotiate access to data on a state-by-state basis.

One key practical challenge is that agencies, in an attempt to be privacy sensitive, may not include in program databases unique identifiers for program applicants and participants. Such unique identifiers facilitate linking to data provided by subjects through other programs or even for the same program over time. Linking datasets through name and address matching or matching on other less unique variables can introduce bias and render the linked data unusable for rigorous analysis. While some agencies have an established history of allowing use of data (including identifying information) for statistical purposes, in many cases access to such data is not readily available due to real or perceived legal, policy, or operational barriers.¹⁷ In some cases, extensive negotiations with the agency responsible for the data are needed to gain access to the data for use in evaluation studies; sometimes the efforts are not successful even after months or years of negotiations.

To help address these barriers, the OMB recently issued guidance to assist both program and statistical agencies (and statistical components within agencies) in increasing the opportunities to use administrative data for statistical purposes, which includes evaluation.¹⁸ In part, this guidance requires government departments to engage both program and statistical agencies in identifying administrative datasets of potential value for statistical purposes; communicating the importance to staff of promoting the use of administrative data for statistical purposes; and identifying several datasets with the most value for statistical purposes but which are not currently being provided, along with descriptions of critical barriers that appear to preclude providing access for statistical purposes. The guidance also offers tools to help agencies in these tasks, including guidance in understanding relevant legal requirements, a tool to facilitate more efficient interagency agreements, and a tool to assess administrative data quality developed under the auspices of the Federal Committee on Statistical Methodology. Departments must also report to the OMB on their efforts to foster collaboration and increase access to administrative data for statistical purposes.

Facilitating Researcher Access to Federal Data while Protecting Privacy

Some agencies have developed ways for researchers to access Federal data for statistical purposes in secure research environments that preserve the confidentiality of individual records. The Census Bureau and National Center for Health Statistics operate secure research data centers, in which qualified researchers with approved projects can use micro-data files for statistical research. The Retirement Research Consortium is a key tool that the Social Security Administration uses to facilitate policy-relevant research on retirement and Social Security. The consortium comprises three competitively selected research centers based at the University of Michigan, Boston College, and the National Bureau of Economic Research. The centers perform valuable research and evaluation of retirement policy, disseminate results, provide training and education awards, and facilitate the use of SSA's administrative data by outside researchers. Nonetheless, due to confidentiality restrictions, uneven interpretations of laws governing privacy of data provided to the government, and other reasons, many data sets remain hard for researchers to access for statistical uses, and opportunities to link to researcher-collected data remain limited.

This Administration is committed to improving opportunities for researcher access in ways that fully maintain privacy protections of Federal program participants. HHS' Centers for Medicare and Medicaid Services' (CMS) Virtual Research Data Center is an innovative example of ways agencies are working to improve access to Federal agencies for their own use and for their grantees carrying out federally sponsored research activities. In late 2013, the Virtual Research Data Center began providing users with a dedicated workspace where they can upload external files and use them with CMS data to run analyses and download aggregate statistical files to their workstations. This model is a more-efficient, less-expensive, more-flexible and more-secure way for researchers to access a variety of Medicare and Medicaid program data, relative to the existing approach that entails cutting, encrypting, and shipping large quantities of information.

Over the last five years, Federal agencies have increasingly used rigorous impact evaluations to inform program decisions, including how to improve programs. Agencies are trying new approaches when the evidence indicates existing strategies are not yielding sufficiently positive impacts on important outcomes. They are restructuring programs to increase their effectiveness when evidence shows new strategies produce better results, and are developing evidence where an insufficient evidence base exists. And they are scaling up approaches that work, improving public policy and people's lives. As part of this effort, agencies are improving the collection and comparability of data to provide new opportunities for evaluation. They are also using cutting-edge technology to improve data access to other Federal agencies and to outside researchers while protecting privacy -- strategies that can enable evaluations to be done more rapidly and at lower cost. The Administration continues to support these efforts to affect change. By using rigorous evaluation strategies to identify what works, and by taking steps to make needed modifications, agencies and taxpayers will have the greatest confidence that scarce resources are being used as efficiently as possible in meeting priority goals.

² For more extensive introductions to impact evaluation (both randomized experiments and quasi-experiments), see Angrist and Pischke (2008, ch. 1) and Stock and Watson (2010, ch. 13). Shadish, Cook, and Campbell (2002) and Berk and Rossi (1998) provide more advanced textbook treatments, and Imbens and Wooldridge (2009) provide a survey of recent methodological developments in the field.

³ Cost estimates supplied by the Department of Justice's National Institute of Justice.

⁴ Public Law 112-96.

⁵ For example, the Department of Labor allocated nearly $24 million in Workforce Innovation Fund grants to pilot Pay for Success grants to increase employment and reduce recidivism among formerly incarcerated individuals (United States Interagency Council on Homelessness, 2013a). DOL required the grantees to employ rigorous evaluation methods in gauging impacts on outcomes, which was defined in the grant solicitation as an experimental or credible quasi-experimental evaluation design.

⁶ Spurred in part by the Homeless Emergency Assistance and Rapid Transition to Housing Act of 2009, the Obama Administration released Opening Doors: The Federal Strategic Plan to End Homelessness in 2010. The plan establishes ambitious goals to end veterans' and chronic homelessness as well as homelessness among youth and families. The U.S. Interagency Council on Homelessness serves to coordinate action by 19 member agencies (United States Interagency Council on Homelessness, 2013b).

⁷ A summary of the study, as well as the Interim Report (which documents the study design process of randomization, and characteristics of the study population) can be found here: http://www.huduser.org/portal/family_options_study.html

⁸ While set-asides within programs are useful, some have noted that department-wide set-asides may have advantages over program-level set-asides by providing agencies with more flexibility over maximizing the return to evaluation investments. Also, set-asides will be used most effectively when agencies have a demonstrated capacity to manage evaluation funds.

⁹ Prior to passage of the ACA, existing demonstration payment waiver authority allowed HHS to conduct Medicare demonstrations of the impacts of new service delivery methods and new payment approaches. However, due to statutory restrictions these demonstrations tended to be relatively small. The ACA provided the Secretary with more flexible authority for testing payment and delivery system innovations, and expanding them based on evidence. This work is conducted under the auspices of the CMMI.

¹⁰ Child and Family Services Improvement and Innovation Act, Title II, Sec. 201, P.L. 112-34.

¹¹ U.S. Department of Agriculture, Food and Nutrition Service Final Rule, Cooperation in USDA Studies and Evaluations, and Full Use of Federal Funds in Nutrition Assistance Programs Nondiscretionary Provisions of the Healthy, Hunger-Free Kids Act of 2010, Public Law 111-296. Federal Register Vol 76, No. 125, June 29, 2011.

¹² This program is based on the Harlem Children's Zone model, which was found to increase earnings for students, decrease the probability of committing crimes and decrease health disability probabilities -- with the potential for providing large public benefits (Dobbie and Fryer 2011).

¹³ For example, the Paperwork Reduction Act (PRA), first enacted in 1980 and amended in 1995 (44 U.S.C., Chapter 35), requires Federal agencies to obtain OMB approval when an agency plans to collect information from ten or more persons using identical reporting, recordkeeping, or disclosure requirements. Among the PRA's goals are ensuring the greatest possible public benefit from and maximizing the utility of information created, collected, maintained, used, shared and disseminated by or for the Federal government and minimizing the burden for persons resulting from the collection of information by or for the Federal government. As a further example, some data collections are subject to review by Institutional Review Boards, in order to protect the rights of the human subjects of such research, a requirement under (42 USC 289) under 45 CFR 46.

¹⁴ On the other hand, the estimates from an RDD strictly pertain only to the types of participants "near" the cutoff. To the extent the impact varies across participants with different levels of need, this can be a limitation.

¹⁵ S. 1284, Report No. 113-71.

¹⁶ Recognizing that agencies operate with scarce resources, OMB has encouraged agencies to adopt such a focus (OMB 2013b).

¹⁷ One legal barrier is that when a program's authorizing statute is silent about whether access to data can be provided for statistical purposes (which includes evaluation), agencies need to make a determination about allowable uses. In such cases, agencies may conservatively interpret the lack of an express authority as a prohibition on providing data to another agency. However, as discussed in OMB memorandum M-14-06, agencies may be able to provide the data under their general statutory authority (OMB 2014).

¹⁸ Statistical purposes is defined in footnote 2 of the OMB memorandum M-14-06 (OMB 2014): [it] refers to "the description, estimation, or analysis of the characteristics of groups, without identifying the individuals or organizations that comprise such groups," (PL-107-347, Title V -- Confidential Information Protection and Statistical Efficiency Act (CIPSEA), Section 502(9)(A)). Statistical purposes exclude "any administrative, regulatory, law enforcement, adjudicatory, or other purpose that affects the rights, privileges, or benefits of a particular identifiable respondent" (PL-107-347, Title V -- CIPSEA, Section 502(5)(A)).

Bernanke, Ben S. 2012. "Monetary Policy since the Onset of the Crisis." Speech at the Federal Reserve Bank of Kansas City Economic Symposium. (http://www.federalreserve.gov/newsevents/speech/ bernanke20120831a.htm).

____. 2014. "The Federal Reserve: Looking Back, Looking Forward." Speech at the Annual Meeting of the American Economic Association. (http://www.federalreserve.gov/newsevents/speech/bernanke20140103a.htm).

(CBO) Congressional Budget Office. 2013a. "Automatic Reductions in Government Spending -- aka Sequestration." (http://www.cbo.gov/publication/43961).

_____. 2013b. "The Economic Impact of S. 744, the Border Security, Economic Opportunity, and Immigration Modernization Act." (http://www.cbo.gov/publication/44346).

Greenspan, Alan. 2013. The Map and the Territory: Risk, Human Nature, and the Future of Forecasting. The Penguin Press.

Piketty, Thomas and Emmanuel Saez. 2003. "Income Inequality in the United States 1913-1998." Quarterly Journal of Economics 118, no. 1: 1-39.

______. 2013. Data update to "Income Inequality in the United States 1913-1998." September (http://elsa.berkeley.edu/~saez/TabFig2012prel.xls).

Reinhart, Carmen M., and Kenneth S. Rogoff. Forthcoming. "Recovery from Financial Crises: Evidence from 100 Episodes." American Economic Review.

Romer, Christina D. 2009. "Back from the Brink." Speech at the Federal Reserve Bank of Chicago. (http://www.whitehouse.gov/assets/documents/Back_from_the_Brink2.pdf).

Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and Jane Waldfogel. 2013.

"Trends in Poverty with an Anchored Supplemental Poverty Measure." Working Paper. New York: Columbia Population Research Center. (http://socialwork.columbia.edu/sites/default/files/file_manager/pdfs/News/Anchored%20SPM.December7.pdf).

Bank of Japan, Financial Markets Department. 2013. Outline of Outright Purchases of Japanese Government Bonds. Financial Market Department 1-3. April.

Blanchard, Oliver J., and Peter A. Diamond. 1994. "Ranking, Unemployment Duration, and Wages." Review of Economic Studies 61, no. 3: 417-34.

Board of Governors of the Federal Reserve System. 2009. "FRB/US Equation Documentation."

______. 2012. "Press Release: December 12, 2012." (http://www.federalreserve.gov/newsevents/press/monetary/20121212a.htm).

______. 2013a. "The October 2013 Senior Loan Officer Opinion Survey on Banking Lending Practices."

______. 2013b. "Press Release: June 19, 2013." (http://www.federalreserve.gov/newsevents/press/monetary/20130619a.htm).

______. 2013c. "Press Release: December 18, 2013." (http://www.federalreserve.gov/newsevents/press/monetary/20131218a.htm).

______. 2014. "Press Release: January 29, 2014." (http://www.federalreserve.gov/newsevents/press/monetary/20140129a.htm).

(BEA) Bureau of Economic Analysis. 2013. "Comprehensive Revision: 1929 Through First Quarter 2013." U.S. Department of Commerce.

______. 2014. "Technical Note: Gross Domestic Product Fourth Quarter of 2013 (Advance Estimate)." U.S. Department of Commerce. January.

(CBO) Congressional Budget Office. 2012. "Economic Effects of Policies Contributing to Fiscal Tightening in 2013." November.

______. 2013a. "Automatic Reductions in Government Spending aka Sequestration." February.

______. 2013b. "The Economic Impact of S. 744, the Border Security, Economic Opportunity, and Immigration Modernization Act." June.

(CEA) Council of Economic Advisers. 1997. Economic Report of the President. February.

______. 2013. "Comprehensive GDP Revision and Advance Estimate for the Second Quarter of 2013." July (http://www.whitehouse.gov/ blog/2013/07/31/comprehensive-gdp-revision-and-advance-estimate-second-quarter-2013).

Database of State Incentives for Renewables and Efficiency. 2013. "Renewable Electricity Production Tax Credit (PTC)." October (http://dsireusa.org/incentives/incentive.cfm?Incentive_Code=US13F).

(DOT) Department of the Treasury. 2012. "Description of the Extraordinary Measures." December (http://web.archive.org/web/20131019163723/http:/www.treasury.gov/connect/blog/Documents/Sec%20Geithner%20LETTER%2012-26-2012%20Debt%20Limit.pdf).

(EIA) Energy Information Administration. 2013. "Industrial Sector Natural Gas Use Rising." June (http://www.eia.gov/todayinenergy/detail.cfm?id=11771).

______. 2013b. "International Energy Outlook." April (http://www.eia.gov/forecasts/ieo/pdf/ieorefngtab_1.pdf).

______. 2013c. "Oil and Gas Industry Employment Growing Much Faster than Total Private Sector Employment." August (http://www.eia.gov/todayinenergy/detail.cfm?id=12451).

(EPA) Environmental Protection Agency. 2011. "Paving the Way Toward Cleaner, More Efficient Trucks." August.

European Central Bank. 2012. "Press Release: Technical Features of Outright Monetary Transactions." September (http://www.ecb.europa.eu/press/pr/date/2012/html/pr120906_1.en.html).

______. 2013. "Monetary Policy." (http://www.ecb.europa.eu/mopo/html/index.en.html).

European Commission. 2013. "European Economic Forecast: Autumn 2013." November.

Executive Office of the President. 2013. "The President's Climate Action Plan." The White House. June.

Fuhrer, Jeffrey C., and Giovanni P. Olivei. 2010. "The Role of Expectations and Output in the Inflation Process: An Empirical Assessment." Federal Reserve Bank of Boston Public Policy Brief, no. 10-2: 1-39.

Gordon, Robert. 1990. "U.S. Inflation, Labor's Share, and the Natural Rate of Unemployment." Working Papers 2585. Cambridge, MA: National Bureau of Economic Research.

______. 2013. "The Phillips Curve is Alive and Well: Inflation and the NAIRU During the Slow Recovery." Working Paper 19390. Cambridge, MA: National Bureau of Economic Research.

Layard, Richard, Stephen Nickells, and Richard Jackman. 1991. Unemployment: Macroeconomic Performance and the Labor Market. New York: Oxford University Press.

Krueger, Alan B., and Andreas I. Mueller. 2011. "Job Search and Job Finding in a Period of Mass Unemployment: Evidence from High-Frequency Longitudinal Data." Working Papers 1295. Princeton, NJ: Center for Economic Policy Studies.

Nalewaik, Jeremy J. 2010. "The Income- and Expenditure-Side Measures of U.S. Output Growth." Brookings Papers on Economic Activity 41, no. 1: 71-127.

(OMB) Office of Management and Budget. 2013. "Impacts and Costs of the Government Shutdown." (http://www.whitehouse.gov/blog/2013/11/07/impacts-and-costs-government-shutdown).

Office of the Press Secretary. 2012. "Obama Administration Finalizes Historic 54.5 MPG Fuel Efficiency Standards." The White House. (http://www.whitehouse.gov/the-press-office/2012/08/28/obama-administration-finalizes-historic-545-mpg-fuel-efficiency-standard).

(USTR) Office of the United States Trade Representative. 2013a. "Acting Deputy U.S. Trade Representative Endy Cutler Discussed Japan and the TPP at the Peterson Institute for International Economics." (http://www.ustr.gov/about-us/press-office/blog/2013/November/Cutler-TPP-Japan-PIIE).

______. 2013b. "Ambassador Froman Discusses the Transatlantic Trade and Investment Partnership at the Munich Security Conference." (http://www.ustr.gov/about-us/press-office/blog/2013/November/Froman-Munich-Security-Conference).

______. 2013c. "Statement of the Ministers and Heads of Delegation for the Trans-Pacific Partnership Countries." (http://www.ustr.gov/tpp).

______. 2013d. "Weekly Trade Spotlight: The Benefits of the WTO Trade Facilitation Agreement to Small Business." (http://www.ustr.gov/about-us/press-office/blog/2013/December/Benefits-of-WTO-Trade-Facilitation-Agreement-to-Small-Business).

______. 2013e. "White House Fact Sheet: Transatlantic Trade and Investment Partnership (T-TIP)." (http://www.ustr.gov/about-us/press-office/fact-sheets/2013/june/wh-ttip).

(OECD) Organization for Economic Cooperation and Development. 2013. "General Assessment of the Macroeconomic Situation." OECD Economic Outlook, no. 2: 9-66.

Polk. 2013. "Polk Finds Average Age of Light Vehicles Continues to Rise." (https://www.polk.com/company/news/polk_finds_average_age_of_light_vehicles_continues_to_rise).

Stock, James H. 2011. "Discussion of Ball and Mazumder, 'Inflation Dynamics and the Great Recession." Brookings Papers on Economic Activity 42, no. 1: 387-402.

Stock, James H., and Mark W. Watson. 2010. "Modeling Inflation After the Crisis." Working Papers 16488. Cambridge, MA: National Bureau of Economic Research.

Zindler, Ethan. 2013. "Burst of Construction in December Delivers Record Year for US Wind." Bloomberg New Energy Finance. January 18.

Alesina, Alberto,and Silvia Ardagna. 2010. "Large Changes in Fiscal Policy: Taxes versus Spending." Tax Policy and the Economy 24: 35-68.

Angrist, Joshua D., and Jörn-Steffen Pischke. 2010. "The Credibility Revolution in Empirical Economics: How Better Research Design Is Taking the Con out of Econometrics." Journal of Economic Perspectives 24, no. 2: 3-30.

Auerbach, Alan J., and Daniel Feenberg. 2000. "The Significance of Federal Taxes as Automatic Stabilizers." Journal of Economic Perspectives 14(3): 37-56.

Auerbach, Alan, and Yuriy Gorodnichenko. 2012. "Measuring the Output Responses to Fiscal Policy." American Economic Journal: Economic Policy 4(2): 1-27.

Barro, Robert J. 1974. "Are Government Bonds Net Wealth?" Journal of Political Economy 82(6): 1095-1117.

Ball, Laurence M. 2009. "Hysteresis in Unemployment: Old and New Evidence." Working Paper 14818. Cambridge, MA: National Bureau of Economic Research.

Bastiat, Frédéric. 1848. "Selected Essays on Political Economy." Seymour Cain, trans. 1995. Library of Economics and Liberty. 3 February 2014.

Blanchard, Olivier, Giovanni Dell'Ariccia, and Paolo Mauro. 2010. "Rethinking Macroeconomic Policy." IMF Staff Position Note. Washington: International Monetary Fund.

Blanchard, Olivier, and Daniel Leigh. 2013. "Growth Forecast Errors and Fiscal Multipliers." IMF Working Paper.

Blanchard, Olivier, and Roberto, Perotti. 2002. "An Empirical Characterization of the Dynamic Effects of Changes in Government Spending and Taxes on Output." Quarterly Journal of Economics.

Blanchard, Olivier, and Lawrence H. Summers. 1986. "Hysteresis and the European Unemployment Problem." NBER Macroeconomics Annual, 1, 15-90. Cambridge, MA: National Bureau of Economic Research.

Blinder, Alan S., and Mark Zandi. 2010. "How the Great Recession was Brought to an End." Princeton University and Moody's Analytics.

Chetty, Raj. 2008. "Moral Hazard vs Liquidity and Optimal Unemployment Insurance." Working Paper 13967. Cambridge, MA: National Bureau of Economic Research.

Chodorow-Reich, Gabriel, Laura Feiveson, Zachary Liscow, and William Gui Woolston. 2012. "Does State Fiscal Relief During Recessions Increase Employment? Evidence from the American Recovery and Reinvestment Act." American Economic Journal: Economic Policy 4, no. 3: 118-145.

Christiano, Lawrence, Martin Eichenbaum, and Sergio Rebelo. 2011. "When Is the Government Spending Multiplier Too Large?" Journal of Political Economy 119(1): 78-121.

Coenen, Gunter, et al. 2012. "Effects of Fiscal Stimulus in Structural Models."

American Economic Journal: Macroeconomics 4(1): 22-68.

Cogan, John F., et al. 2010 "New Keynesian versus old Keynesian government spending multipliers." Journal of Economic Dynamics and Control 34(3): 281-295.

(CBO) Congressional Budget Office. 2009a. "Health Information Technology for Economic and Clinical Health Act."

______. 2009b. "Cost Estimate: H.R. 1, American Recovery and Reinvestment Act of 2009."

______. 2009c. "Cost Estimate: H.R. 3548, Worker, Homeownership, and Business Assistance Act of 2009."

______. 2010a. "The Budget and Economic Outlook: Fiscal Years 2010 to 2020."

______. 2010b. "The Budget and Economic Outlook: An Update."

______. 2010c. "Cost Estimate: H.R. 4691, the Temporary Extension Act of 2010, As Introduced on February 25, 2010."

______. 2010d. "Cost Estimate: Budgetary Effects of Hiring Incentives to Restore Employment Act, as Introduced by Senator Reid on February 11, 2010."

______. 2010e. "Cost Estimate: Amendment No. 3721 to H.R. 4851, the Continuing Extension Act, 2010, as Proposed by Senator Baucus."

______. 2010f. "Cost Estimate: Budgetary Effects of Senate Amendment 4425, the Unemployment Compensation Extension Act of 2010."

______. 2010g. "Cost Estimate: CBO Estimate of Changes in Revenues and Direct Spending for Senate Amendment 4594 in the Nature of a Substitute to H.R. 5297, the Small Business Jobs and Credit Act of 2010."

______. 2010h. "Cost Estimate: CBO Estimate of Changes in Revenues and Direct Spending for S.A. 4753, an amendment to H.R. 4852, the Tax Relief, Unemployment Insurance Reauthorization, and Job Creation Act of 2010."

______. 2010i. "Cost Estimate: Budgetary Effects of Senate Amendment 4575, containing proposals related to education, state fiscal relief, the Supplemental Nutrition Assistance Program, rescissions, and revenue offsets."

______. 2011a. "The Budget and Economic Outlook: Fiscal Years 2011 to 2021."

______. 2011b. "Cost Estimate: Budgetary Effects of Senate Amendment 927 to H.R. 674, as proposed by Senator Reid for Senator Tester."

______. 2011c. "Cost Estimate: H.R. 5297, the Small Business Jobs Act of 2010."

______. 2011d. "Cost Estimate: Budgetary Effects of the Temporary Payroll Tax Cut Continuation Act of 2011, as Posted on the Website of the House Committee on Rules on December 22, 2011."

______. 2012a. "The Budget and Economic Outlook: Fiscal Years 2012 to 2022."

______. 2012b. "Cost Estimate: Budgetary effects of the Conference Agreement for H.R. 3630, the Middle Class Tax Relief and Job Creation Act of 2012, as Posted on the Web Site of the House Committee on Rules on February 16, 2012."

______. 2012c. "Cost Estimate: Budgetary effects of the Conference Agreement for H.R. 3630, the Middle Class Tax Relief and Job Creation Act of 2012, as Posted on the Web Site of the House Committee on Rules on February 16, 2012."

______. 2013a. "Cost Estimate: Estimate of Budgetary Effects of H.R. 8, the American Taxpayer Relief Act of 2012, as passed by the Senate on January 1, 2013."

______. 2014a. "The Budget and Economic Outlook: 2014 to 2024."

______. 2014b. "Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output in 2013."

Conley, Timothy, and Bill Dupor. 2013. "The American Recovery and Reinvestment Act: Public Sector Jobs Saved, Private Sector Jobs Forestalled." Journal of Monetary Economics.

(CEA) Council of Economic Advisers. 2009a. "Estimates of Job Creation from the American Recovery and Reinvestment Act of 2009."

______. 2009b. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." First Quarterly Report. September.

______. 2010a. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." Second Quarterly Report. January.

______. 2010b. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." Third Quarterly Report. April.

______. 2010c. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." Fourth Quarterly Report. July.

______. 2011. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." Sixth Quarterly Report. March.

______. 2013a. "The Economic Impact of the American Recovery and Reinvestment Act of 2009." Ninth Quarterly Report. February.

______. 2013b. "Economic Report of the President." March.

Council of Economic Advisers and the Department of Labor. 2014. "The Economic Benefits of Extending Unemployment Insurance." January.

Delong, J. Bradford, and Lawrence H. Summers. 2012. "Fiscal Policy in a Depressed Economy." Brookings Papers on Economic Activity. Spring.

Department of Education. 2011. "FY 2012 Department of Education Justifications of Appropriation Estimates to the Congress: Student Financial Assistance" in President's FY 2012 Budget Request for the U.S. Department of Education.

Department of the Treasury. 2010a. "The Case for Temporary 100 Percent Expensing: Encouraging Business to Expand now by Lowering the Cost of Investment."

______. 2010b. "The American Opportunity Tax Credit."

______. 2011. "Treasury Analysis of Build America Bonds Issuance and Savings."

Department of the Treasury and Council of Economic Advisers. 2012. "A New Economic Analysis of Infrastructure Investment." March.

Eggertson, Gauti B. 2001. "Real Government Spending in a Liquidity Trap." New York Federal Reserve.

Elmendorf, Douglas W., and Jason Furman. 2008. "If, When, How: A Primer on Fiscal Stimulus." Washington: Brookings Institution.

(EPA) Environmental Protection Agency. 2013. "American Recovery and Reinvestment Act Quarterly Performance Report -- FY2013 Quarter 4 Cumulative Results as of September 30, 2013."

Executive Office of the President and Office of the Vice President. 2010. "The Recovery Act: Transforming the American Economy through Innovation".The White House.

Farhi, Emmanuel, and Ivan Werning. 2012. "Fiscal Multipliers: Liquidity Traps and Currency Unions." Working Paper 18321. Cambridge, MA: National Bureau of Economic Research.

Favero, Carlo, and Francesco Giavazzi. 2012. "Measuring Tax Multipliers: The Narrative Method in Fiscal VARs." American Economic Journal: Economic Policy 4(2): 69-94.

Fernald, John G. 1999. "Roads to Prosperity? Assessing the Link Between Public Capital and Productivity," The American Economic Review, Vol. 89, No. 3: 619-638.

Feyrer, James, and Bruce Sacerdote. 2011. "Did the Stimulus Stimulate? Real Time Estimates of the Effects of the American Recovery and Reinvestment Act." Working Paper 16759. National Bureau of Economic Research.

Financial Crisis Inquiry Commission. 2011. The Financial Crisis Inquiry Report. New York: PublicAffairs.

Follette, Glenn, and Byron Lutz. 2010. "Fiscal Policy in the United States: Automatic Stabilizers, Discretionary Fiscal Policy Actions, and the Economy." Finance and Economics Discussion Series. Washington: Federal Reserve Board.

Guajardo, Jaime, Daniel Leigh, and Andrea Pescatori. Forthcoming. "Expansionary Austerity? New International Evidence." Journal of the European Economic Association.

Hall, Bronwyn H., Jacques Mairesse, Pierre Mohnen. 2009. "Measuring the Returns to R&D." Working Paper 15622. Cambridge, MA: National Bureau of Economic Research.

House, Christopher L., and Matthew D. Shapiro. 2008. "Temporary Investment Tax Incentives: Theory with Evidence from Bonus Depreciation." American Economic Review 98(3): 737-768.

Ilzetzki, Ethan, Enrique G. Mendoza, and Carlos A. Vegh. 2011. "How Big (Small?) are Fiscal Multipliers?" IMF Working Paper.

(IMF) International Monetary Fund. 2009 "What's the Damage? Medium Term Dynamics after Financial Crises." World Economic Outlook October 2009: Sustaining the Recovery Chapter 4, p. 121-151.

______. 2012. World Economic Outlook April 2012: Growth Resuming, Dangers Remain."

Johnson, David S., Jonathan A. Parker, and Nicholas S. Souleles. 2006. "Household Expenditure and the Income Tax Rebates of 2001." American Economic Review 96(5): 1589-1610.

Joint Committee on Taxation. 2009. "Estimated Budget Effects of the Revenue Provisions Contained in the Conference Agreement for H.R.1, The 'American Recovery and Reinvestment Tax Act of 2009.'"

______. 2010a. "Estimated Budget Effects of the Revenue Provisions Contained in Senate Amendment #4594 to H.R. 5297, The 'Small Business Jobs Act of 2010,' Scheduled for Consideration by the United States Senate on September 16, 2010."

______. 2010b. "Estimated Revenue Effects of the House Amendment to the Senate Amendment to H.R. 4853, The 'Middle Class Tax Relief Act of 2010,' Scheduled for Consideration by the House of Representatives on December 2, 2010."

______. 2012. "Estimated Revenue Effects of H.R. 8, The 'Job Protection and Recession Prevention Act of 2012.'"

Laeven, Luc, and Fabian Valencia. 2012. "Systemic Banking Crises Database: An Update."Working Paper WP/12/163. Washington: International Monetary Fund.

Ljungqvist, Lars, and Thomas J. Sargent. 1998. "The European Unemployment Dilemma." Journal of Political Economy 106(3): 514-550.

Mertens, Karel, and Morten O. Ravn. 2012. "Empirical Evidence on the Aggregate Effects of Anticipated and Unanticipated US Tax Policy Shocks." American Economic Journal: Economic Policy 4(2): 145-181.

Mertens, Karel, and Morten Ravn. 2013. "The Dynamic Effects of Personal and Corporate Income Tax Changes in the United States." American Economic Review. June.

Munnell, Alicia H, 1992. "Infrastructure Investment and Economic Growth," Journal of Economic Perspectives, vol. 6(4), pages 189-98, Fall. Pittsburg, PA: American Economic Association.

Nakamura, Emi, and Jón Steinsson. 2011. "Fiscal Stimulus in a Monetary Union: Evidence from U.S. Regions." Unpublished paper, New York: Columbia University.

Office of Science and Technology Policy, and The National Economic Council. 2013. "Four Years of Broadband Growth." The White House.

Parker, Jonathan A. 2011. "On Measuring the Effects of Fiscal Policy in Recessions." Working Paper 17240. Cambridge, MA: National Bureau of Economic Research.

Parker, Jonathan A., Nicholas S. Souleles, David S. Johnson, and Robert McClelland. 2011. "Consumer Spending and the Economic Stimulus Payments of 2008." Working Paper 16684. Cambridge, MA: National Bureau of Economic Research.

Perotti, Roberto. 2011. "The Austerity Myth: Gain Without Pain?" Working Paper 17571. Cambridge, MA: National Bureau of Economic Research.

Phelps, Edmund. 1972. "Inflation policy and unemployment theory." New York: WW Norton and Company.

Poterba, James M. 1994. "State Responses to Fiscal Crises: The Effects of Budgetary Institutions and Politics." Journal of Political Economy 102(4): 799-821.

Ramey, Valerie A. 2011a. "Identifying Government Spending Shocks: It's All in the Timing" Quarterly Journal of Economics 126, no. 1: 1-50.

______. 2011b. "Can Government Purchases Stimulate the Economy?" Journal of Economic Literature 49, no. 3: 673-85.

Ramey, Valerie A, and Matthew Shapiro. 1998. "Costly Capital Reallocation and the Effects of Government Spending." Carnegie-Rochester Conference on Public Policy 48: 145-94.

Reichling, Felix, and Charles Whalen. 2012. "Assessing the Short-Term Effects on Output Changes in Federal Fiscal Policies." Working Paper 2012-08. Congressional Budget Office.

Reinhart, Carmen M., and Kenneth S. Rogoff. 2009. This time is different: Eight Centuries of Financial Folly. Princeton, NJ: Princeton University Press.

______. Forthcoming. "Recovery from Financial Crises: Evidence from 100 Episodes." American Economic Review.

Reifschneider, Dave, William L. Wascher, and David Wilcox. 2013. "Aggregate Supply in the United States: Recent Developments and Implications for the Conduct of Monetary Policy." 14^th Jacques Polak Annual Research Conference. Washington: International Monetary Fund.

Romer, Christina D. 2011. "Back from the Brink." In The International Financial Crisis: Have the Rules of Finance Changed? Edited by Asli Demirgüç-Kunt, Douglas D. Evanoff, George G. Kaufman, pp. 15-31. World Scientific Publishing Company.

______. 2012. "Fiscal Policy In the Crisis: Lessons and Policy Implications." University of California-Berkeley, Department of Economics.

Romer, Christina D., and David H. Romer. 2010. "The Macroeconomic Effects of Tax Changes: Estimates Based on a New Measure of Fiscal Shocks." American Economic Review 100(3): 763-801.

Rothstein, Jesse. 2011. "Unemployment Insurance and Job Search in the Great Recession." Brookings Papers on Economic Activity, Fall.

Shoag, Daniel. 2013. "Using state pension shocks to estimate fiscal multipliers since the Great Recession." The American Economic Review 103(3): 121-124.

Sims, Christopher A. 2010. "But Economics Is Not an Experimental Science." Journal of Economic Perspectives 24, no. 2: 59-68.

Smets, Frank, and Rafael Wouters. 2007. "Shocks and Frictions in US Business Cycles: a Bayesian DSGE Approach." American Economic Review 97(3): 586-606.

Sperling, Gene. 2007. "Ways to Get Economic Stimulus Right This Time." Bloomberg.com. December 17.

Stock, James H. 2010. "The Other Transformation in Econometric Practice: Robust Tools for Inference." Journal of Economic Perspectives 24, no. 2: 83-94.

Stock, James H. and Mark W. Watson. 2012. "Disentangling the Channels of the 2007-09 Recession," Brookings Papers on Economic Activity. Spring.

Stock, James H., and Mark W. Watson. 2010. Introduction to Econometrics. 3rd ed. Boston, MA: Addison-Wesley.

Suarez Serrato, Juan Carlos, and Philippe Wingender. 2011. "Estimating the Incidence of Government Spending."

Summers, Lawrence H. 2007. "The State of the US Economy." Presentation at Brookings Institution Forum on December 19^th, 2007.

______. 2008. Speech at the Wall Street Journal-CEO Council Conference. Washington, D.C., November 19.

Taylor, John B. 2011. "An Empirical analysis of the Revival of Fiscal Activism in the 2000s." Journal of Economic Literature 49(3): 686:702.

______. Forthcoming. "The Role of Policy in the Great Recession and the Weak Recovery." American Economic Review, Papers and Proceedings.

Transportation Research Board of the National Academies. 2013. "Transportation Investments in Response to Economic Downturns, Special Report 312."

Wilson, Daniel J. 2012. "Fiscal Spending Jobs Multipliers: Evidence from the 2009 American Recovery and Reinvestment Act." American Economic Journal: Economic Policy.

Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and Jane Waldfogel. 2013. "Trends in Poverty with an Anchored Supplemental Poverty Measure." Working Paper 1-25. New York: Columbia Population Research Center.

Woodford, Michael. 2011. "Simple Analytics of the Government Expenditure Multiplier." American Economic Journal: Macroeconomics (3), p. 1-35.

Aitken, Murray, Ernst R. Berndt, and David M. Cutler. 2009. "Prescription Drug Spending Trends in the United States: Looking Beyond the Turning Point." Health Affairs 28, no. 1: 151-160.

Altarum Institute. 2014. "Insights from Monthly National Health Expenditures Estimates through November 2013." (http://altarum.org/sites/default/files/uploaded-related-files/CSHS-Spending-Brief_January%202014.pdf).

ASPE (Assistant Secretary for Planning and Evaluation, Office of). 2013. "Health Insurance Marketplace Premiums for 2014." Health and Human Services. (http://aspe.hhs.gov/health/reports/2013/Market-placePremiums/ib_marketplace_premiums.cfm).

Baicker, Katherine and Amitabh Chandra. 2006. "The Labor Market Effects of Rising Health Insurance Premiums." Journal of Labor Economics 24, no. 3: 609-634.

Chandra, Amitabh, Jonathan Holmes, and Jonathan Skinner. 2013. "Is This Time Different? The Slowdown in Healthcare Spending." Brookings Panel on Economic Activity. Washington: Brookings Institution. (http://www.brookings.edu/~/media/Projects/BPEA/Fall%202013/2013b%20chandra%20healthcare%20spending.pdf).

Centers for Medicare and Medicaid Services Office of the Actuary. 2013. "Projections of National Health Expenditures: Methodology and Model Specification."

(CMS) Centers for Medicare and Medicaid Services. 2013a. "New Data Shows Affordable Care Act Reforms Are Leading to Lower Hospital Readmission Rates for Medicare Beneficiaries." (http://blog.cms.gov/2013/12/06/new-data-shows-affordable-care-act-reforms-are-leading-to-lower-hospital-readmission-rates-for-medicare-beneficiaries/).

______. 2013b. "Pioneer Accountable Care Organizations succeed in improving care, lowering costs." News Release. (http://www.cms.gov/Newsroom/MediaReleaseDatabase/Press-Releases/2013-Press-Releases-Items/2013-07-16.html).

______. 2013c. "The Affordable Care Act: A Stronger Medicare Program." (http://www.cms.gov/apps/files/Medicarereport2012.pdf).

Clemens, Jeffrey and Joshua D. Gottlieb. 2013. "Bargaining in the Shadow of a Giant: Medicare's Influence on Private Payment Systems." Working Paper 19503. Cambridge, MA: National Bureau of Economic Research.

(CBO) Congressional Budget Office. 2010a. "Cost Estimate of H.R. 4872, Reconciliation Act of 2010 (Final Health Care Legislation)." (http://www.cbo.gov/publication/21351).

______. 2010b. "The Budget and Economic Outlook: An Update." (http://www.cbo.gov/publication/21670).

______. 2011. "The Budget and Economic Outlook: An Update." (http://www.cbo.gov/publication/41586).

______. 2012a. "Updated Budget Projections: Fiscal Years 2012 to 2022." (http://www.cbo.gov/publication/43119).

______. 2012b. "Letter to the Honorable John Boehner providing an estimate for H.R. 6079, the Repeal of Obamacare Act." (http://www.cbo.gov/publication/43471).

______. 2012c. "Estimates for the Insurance Coverage Provisions of the Affordable Care Act Updated for the Recent Supreme Court Decision." (http://www.cbo.gov/sites/default/files/cbofiles/attachments/43472-07-24-2012-CoverageEstimates.pdf).

______. 2012d. "An Update to the Budget and Economic Outlook: Fiscal Years 2012 to 2022." (http://www.cbo.gov/publication/43539).

______. 2013a. "Updated Budget Projections: Fiscal Years 2013 to 2023." (http://www.cbo.gov/publication/44172).

______. 2013b. "The 2013 Long-Term Budget Outlook." (http://www.cbo.gov/publication/44521).

______. 2013c. "Health-Related Options for Reducing the Deficit: 2014 to 2023." (http://www.cbo.gov/sites/default/files/cbofiles/attachments/44906-HealthOptions.pdf).

______. 2014. "The Budget and Economic Outlook: Fiscal Years 2014 to 2024." (http://www.cbo.gov/publication/45010).

(CEA) Council of Economic Advisers. 2009. "The Economic Case for Health Care Reform." (http://www.whitehouse.gov/assets/documents/CEA_Health_Care_Report.pdf).

______. 2013. Economic Report of the President. (http://www.whitehouse.gov/administration/eop/cea/economic-report-of-the-President/2013).

Cuckler, Gigi A., et al. 2013. "National Health Expenditure Projections, 2012-22: Slow Growth Until Coverage Expands And Economy Improves." Health Affairs 32 (September): 1-12.

Cutler, David. 2004. Your Money or Your Life: Strong Medicine for America's Health Care System. New York: Oxford University.

Cutler, David and Neeraj Sood. 2010. "New Jobs Through Better Health Care: Health Care Reform Could Boost Employment by 250,000 to 400,000 a Year this Decade." Washington: Center for American Progress. (http://www.americanprogress.org/issues/2010/01/pdf/health_care_jobs.pdf).

Cutler, David and Nikhil R. Sahni. 2013. "If Slow Rate of Health Care Spending Growth Persists, Projections May Be Off by $770 Billion." Health Affairs 32 (September): 841-850.

Daly, Mary, Bart Hobijn, and Brian Lucking. 2012. "Why Has Wage Growth Stayed Strong?" Federal Reserve Board of San Francisco. (http://www.frbsf.org/economic-research/publications/economic-letter/2012/april/strong-wage-growth/).

Dranove, David. 1988. "Pricing by Non-Profit Institutions." Journal of Health Economics 7 (March): 47-57.

Elmendorf, Douglas. 2013. "The Slowdown in Health Care Spending." Presentation to the Brookings Panel on Economic Activity. Washington: Brookings Institution. September 19 (http://www.cbo.gov/publication/44596).

Frakt, Austin B. 2011. "How Much Do Hospitals Cost Shift? A Review of the Evidence." The Milbank Quarterly 89, no. 1: 90-130.

______. 2013. "The End of Hospital Cost Shifting and the Quest for Hospital Productivity." Health Services Research 49 (September): 1-10.

Gerhardt, Geoffrey, et al. 2013. "Medicare Readmission Rates Showed Meaningful Decline in 2012." Medicare & Medicaid Research Review 3, no. 2: E1-E12.

Gruber, Jonathan and Alan B. Krueger. 1991. "The Incidence of Mandated Employer-Provided Insurance: Lessons From Workers' Compensation Insurance." In Tax Policy and the Economy, Volume 5, edited by David Bradford, pp. 111-144. Cambridge, MA: MIT Press.

Gruber, Jonathan. 1994. "The Incidence of Mandated Maternity Benefits." American Economic Review 84, no. 3: 622-641.

He, Daifeng and Jennifer M. Mellor. 2012. "Hospital volume responses to Medicare's Outpatient Prospective Payment System: Evidence from Florida." Journal of Health Economics 31: 730-743.

Holahan, John and Stacey McMorrow. 2012. "Medicare, Medicaid, and the Deficit Debate." Washington: Urban Institute. (http://www.urban.org/UploadedPDF/412544-Medicare-Medicaid-and-the-Deficit-Debate.pdf).

IMS Institute for Healthcare Informatics. 2013. "Declining Medicine Use and Costs: For Better or Worse?" (http://static.correofarmaceutico.com/docs/2013/05/20/usareport.pdf).

Kaiser Family Foundation. 2013a. "2013 Employer Health Benefits Survey." (http://kff.org/private-insurance/report/2013-employer-health-benefits).

______. 2013b. "Medicaid Enrollment: June 2012 Data Snapshot." (http://kaiserfamilyfoundation.files.wordpress.com/2013/08/8050-06-medicaid-enrollment.pdf).

Kaiser Family Foundation and Altarum Institute. 2013. "Assessing the Effects of the Economy on the Recent Slowdown in Health Spending." (http://kff.org/health-costs/issue-brief/assessing-the-effects-of-the-economy-on-the-recent-slowdown-in-health-spending-2).

Levine, Michael and Melinda Buntin. 2013. "Why Has Growth in Spending for Fee-for-Service Medicare Slowed?" Working Paper 2013-06. Congressional Budget Office.

Martin, Anne B., et al. 2014. "National Health Spending in 2012: Rate of Health Spending Growth Remained Low for the Fourth Consecutive Year." Health Affairs 33, no. 1: 1-11.

McClellan, Mark, et al. 1998. "Are Medical Prices Declining? Evidence from Heart Attack Treatments." The Quarterly Journal of Economics 113, no. 4: 991-1024.

McWilliams, J. Michael, Bruce Landon, Michael E. Chernew. 2013. "Changes in Health Care Spending and Quality for Medicare Beneficiaries Associated With a Commercial ACO Contract." Journal of the American Medical Association 310, no. 8: 829-836.

MedPAC (Medicare Payment Advisory Commission). 2009. "Improving Incentives in the Medicare Program." (http://www.medpac.gov/documents/jun09_entirereport.pdf).

Morrissey, Michael A. 1994. Cost Shifting in Health Care: Separating Evidence from Rhetoric. Washington: American Enterprise Institute.

Ryu, Alexander J., et al. 2013. "The Slowdown in Health Care Spending in 2009-11 Reflected Factors Other than the Weak Economy and Thus May Persist." Health Affairs 32, no. 5: 835-840.

Sisko, Andrea, et al. 2009. "Health Spending Projections Through 2018: Recession Effects Add Uncertainty to the Outlook." Health Affairs 28, no. 2: 346-357.

Sood, Neeraj, Arkadipta Ghosh and Jose J. Escarce. 2009. "Costs, Use and Outcomes: Employer-Sponsored Insurance, Health Care Cost Growth, and the Economic Performance of U.S. Industries." Health Services Research 44, no. 5: 1449-1464.

Sommers, Benjamin D. 2005. "Who Really Pays for Health Insurance? The Incidence of Employer-Provided Health Insurance with Sticky Nominal Wages," International Journal of Health Care Finance and Economics 5, no. 1: 89-118.

Song, Zirui, et al. 2012. "The 'Alternative Quality Contract,' Based on a Global Budget, Lowered Medical Spending and Improved Quality." Health Affairs 31, no.8: 1885-1894.

Spiro, Topher and Jonathan Gruber. 2013. "The Affordable Care Act's Lower-Than-Projected Premiums Will Save $190 billion." Washington: Center for American Progress. October 21 (http://www.americanprogress.org/issues/healthcare/report/2013/10/23/77537/the-affordable-care-acts-lower-than-projected-premiums-will-save-190-billion/).

(SSA) Social Security Administration. 2013. "Annual Statistical Supplement, 2013." (http://www.ssa.gov/policy/docs/statcomps/supplement/).

Summers, Lawrence H. 1989. "Some Simple Economics of Mandated Benefits." The American Economic Review 79, no. 2: 177-183.

White, Chapin. 2007. "Health Care Spending Growth: How Different is the United States from the Rest of the OECD." Health Affairs, 26 no. 1: 154-161.

______. 2013. "Contrary To Cost-Shift Theory, Lower Medicare Hospital Payment rates For Inpatient Care Lead To Lower Private Payment Rates." Health Affairs 32, no. 5: 935-943.

White, Chapin and Vivian Yaling Wu. 2013. "How Do Hospitals Cope with Sustained Slow Growth in Medicare Prices?" Health Services Research 49: 1-21.

Yamamato, Dale H. 2013. "Health Care Costs From Birth to Death." Health Care Cost Institute Independent Report Series. (http://www.health-costinstitute.org/files/Age-Curve-Study_0.pdf).

Acemoglu, Daron, and David Autor. 2011 "Skill, Tasks and Technologies: Implications for Employment Earnings." In The Handbook of Labor Economics, edited by Orley Ashenfelter and David Card, vol. 4b. Amsterdam: Elsevier.

Acemoglu, Daron and James Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity and Poverty. New York: Crown Publishers.

Acemoglu, Daron and David Autor. 2012. "What Does Human Capital Do? A Review of Goldin and Katz's Race between Education and Technology." Journal of Economic Literature, 50(2): 426-463.

Agrawal, N. 2010. "Review on just in time techniques in manufacturing systems." In Advances in Production Engineering & Management 5, no.2: 101-110

Akcigit, Ufuk, Douglas Hanley, and Nicolas Serrano-Velarde. 2013. "Back to Basics: Basic Research Spillovers, Innovation Policy and Growth." Working Paper 19473. Cambridge, MA.: National Bureau of Economic Research.

Alvaredo, Facundo, Anthony B. Atkinson, Thomas Piketty, and Emmanuel Saez. 2013. "The Top 1 Percent in International and Historical Perspective." Journal of Economic Perspectives, 27(3): 3-20.

Autor, David. 2010. "The Polarization of Job Opportunities in the U.S. Labor Market, Implications for Employment and Earnings." Washington: Center for American Progress and The Hamilton Project.

Autor, David, Lawrence F. Katz, and Melissa S. Kearney. 2006. "The Polarization of the U.S. Labor Market." Working Paper 11986. Cambridge, MA: National Bureau of Economic Research.

Autor, David, Frank Levy, and Richard J. Murnane. 2003. "The Skill Content of Recent Technological Change: an Empirical Exploration." Quarterly Journal of Economics 116 (4): 1279-1333.

Ayres, Ian, and Peter Cramton. 1996. "Deficit Reduction Through Diversity: How Affirmative Action at the FCC Increased Auction Competition," Stanford Law Review 48: 761-815.

Bailey, Diane E., and Nancy B. Kurland. 1999. "The Advantages and Challenges of Working Here, There, Anywhere, and Anytime."

Organizational Dynamics (Autumn): 1-16.

Bailey, Diane E., and Nancy B. Kurland. 2002. "A Review of Telework Research: Findings, New Directions, and Lessons for the Study of Modern Work." Journal of Organizational Behavior 23, no. 4: 383-400.

Baily, Martin N., and Robert J. Gordon. 1988. "The Productivity Slowdown, Measurement Issues, and the Explosion of Computer Power," Brookings Papers on Economic Activity, vol. 19(2): 347-432.

Bakia, Marianne, Karla Jones, Barbara Means, Robert Murphy, and Yukie Toyama. 2010. "Evaluation of Evidence-Based Practices in Online Learning: A Meta-Analysis and Review of Online Learning Studies." U.S. Department of Education Office of Planning, Evaluation, and Policy Development Policy and Program Studies Service.

Banerjee, Abhijit V., and Esther Duflo. 2003. "Inequality and Growth: What Can the Data Say?" Journal of Economic Growth 8, no. 3: 267-299.

Banerjee, Abhijit, Shawn Cole, Esther Duflo, and Leigh Linden. 2007. "Remedying Education: Evidence from Two Randomized Experiments in India." The Quarterly Journal of Economics, 1235-1264.

Barrow, Lisa, Lisa Markman, and Cecilia Rouse. 2009. "Technology's Edge: The Educational Benefits of Computer-Aided Instruction." The American Economic Journal: Economic Policy, (1), 52-74.

Basu, Fernald, Oulton, and Srinivasan. 2004. "The Case of the Missing Productivity Growth, or Does Information Technology Explain Why Productivity Accelerated in the United States but Not in the United Kingdom?" Working Paper 10010. Cambridge, MA: National Bureau of Economic Research.

Basu, Susanto, John Fernald, and Miles Kimball. 2006. "Are Technology Improvements Contractionary?" American Economic Review. 96(5), 1418-48.

Berman, Eli, John Bound, and Zvi Griliches. 1994. "Changes in the Demand for Skilled Labor within U.S. Manufacturing: Evidence from the Annual Survey of Manufactures." The Quarterly Journal of Economics 109(2), 367-397.

Bernanke, Ben. 2013. "Economic Progress for the Long Run." Speech at Bard College. May 18.

Bernstein, Jared. 2013. "The Impact of Inequality on Growth," Washington: Center for American Progress.

Bloom, Kretschmer, and Van Reenen. 2006. "Work-Life Balance, Management Practices and Productivity." Centre for Economic Performance (January): 1-45.

Bloom, Nicholas, Mark Schankerman, and John Van Reenen. 2012. "Identifying Technology Spillovers and Product Market Rivalry." Centre for Economic Performance (December): 1-81.

Bloom, Liang, Roberts, and Ying. 2013. "Does Working from Home Work? Evidence from a Chinese Experiment." Working Paper. London: Centre for Economic Performance.

BLS (Bureau of Labor Statistics). "Private Nonfarm Business Sector: Multifactor Productivity"

______. "Private Nonfarm Business Sector: Output per Hour of all Persons"

______. "Private Nonfarm Business Sector: Sources of Productivity Growth."

______. "Private Nonfarm Business Sector: Real Output Per Hour."

______. "Private Nonfarm Business Sector: Real Compensation Per Hour."

Bound, John and George Johnson. 1995. "What are the Causes of Rising Wage Inequality in the United States?" Economic Policy Review 1, no. 1: 9-17.

Bureau of Labor Statistics, Office of Productivity and Technology. 2013. Net Multifactor Productivity and Cost, 1948-2012.

Busch, Emily, Jenna Nash, and Bradford S. Bell. 2011. "Remote Work: An Examination of Current Trends and Emerging Issues." Center for Advanced Human Resource Studies, Cornell University (Spring): 1-12.

Card, David, and John DiNardo. 2002. "Skill Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles." Journal of Labor Economics 20(4), 733-783.

Carew, Diana, G. and Michael Mandel. 2013. Progressive Policy Institute. "U.S. Investment heroes of 2013: The Companies Betting on America's Future".

Carillo, Ponce, Mercedes Onofa, and Juan Ponce. 2010. "Information Technology and Student Achievement: Evidence from a Randomized Experiment in Ecuador." Washington: Inter-American Development Bank.

CBO (Congressional Budget Office). 2011a. "Trends in the Distribution of Household Income between 1979 and 2007."

______. 2011b. "S. 27, Preserve Access to Affordable Generics Act."

______. 2013. The Economic Impact of S. 744, the Border Security, Economic Opportunity, and Immigration Modernization Act. Government Printing Office.

Charles, Dustin, et al. 2013. "Adoption of Electric Health Record Systems among U.S. Nonfederal Acute Care hospitals: 2008-2012." ONC Data Brief no. 9. The Office of the National Coordinator for Health Information Technology, U.S. Department of Health and Human Resources.

Chien, Colleen V. 2012. "Reforming Software Patents." Legal Studies Research Paper Series. Santa Clara, CA: Santa Clara University School of Law.

Coase, Ronald H. 1959. "The Federal Communications Commission," Journal of Law and Economics 2: 1-40.

Council of Economic Advisers, the National Economic Council, and the Office of Science & Technology Policy. 2013. "Patent Assertion and U.S. Innovation." The White House.

CEA (Council of Economic Advisers). ______. 2010. "Work-Life Balance and the Economics of Workplace Flexibility."(March).

______. 2012 "The Economic Benefits of New Spectrum for Wireless Broadband."

Cramton, Peter and Jesse A. Schwartz. 2000. "Collusive Bidding: Lessons from the FCC Spectrum Auctions," Journal of Regulatory Economics 17: 229-252.

______. 2002. "Collusive Bidding in the FCC Spectrum Auctions," Contributions to Economic Analysis & Policy 1: 1-18.

Deardorff, Alan V. 1991. "Welfare Effects of Global Patent Protection." Economica 59 (May): 35-51.

Delgado, Mercedes, Christian Ketels, Michael E. Porter, and Scott Stern. 2012. "The Determinants of National Competiveness." Working Paper 18249. Cambridge, MA.: National Bureau of Economic Research (July).

Department of Justice and United States Patent and Trademark Office. 2013. Policy Statement on Remedies for Standard-Essential Patents Subject to Voluntary FRAND Commitments. Government Printing Office.

DiNardo, John, Nicole M Fortin, and Thomas Lemieux. 1996. "Labor Market Institutions and the Distribution of Wages, 1973-1992: A Semiparametric Approach." Econometrica 64(5), 1001-1044.

Evenson, Robert, and Sunil Kanwar. 2003. "Does intellectual property spur technological change?" Oxford Economic Papers 55: 235-264.

Fairris, David, and Mark Brenner. 2001. "Workplace Transformation and the Rise in Cumulative Trauma Disorders: Is There a Connection?" Journal of Labor Research XXII, no.1 (Winter): 15-28.

Falvey, Rod, Neil Foster, and David Greenaway. 2006. "Intellectual Property Rights and Economic Growth." Review of Development Economics 10(4): 700-719.

Farrell, Joseph, John Hayes, Carl Shapiro, and Theresa Sullivan. 2007. "Standard Setting, Patents, and Hold-Up." Antitrust Law Journal 74, no. 3: 603-670.

Federal Trade Commission v. Actavis, Inc. et al. 570 U.S. __ (2013).

FTC (Federal Trade Commission). 2010. "Pay-for-delay: How drug company pay-offs cost consumers billions: An FTC staff study.".

Fernald, John. 2012. "Productivity and potential output before, during, and after the Great Recession." Working Paper 2012-18. Federal Reserve Bank of San Francisco.

Feyrer, James. 2007. "Demographics and Productivity." The Review of Economics and Statistics 89, 100-109.

______. 2011. "The U.S. Productivity Slowdown, the Baby Boom, and Management Quality." Journal of Population Economics 24, 267-284.

Fields, Gary. 2001. "Distribution and Development, A New Look at the Developing World." Journal of Development Economics, Vol. 70: 238-243.

Foellmi, R. and Zweimuller, J. 2006. "Income Distribution and Demand-Induced Innovations."

Review of Economic Studies 73(4): 941-960.

Freeman, Richard B., and Lawrence F. Katz, eds. 1995. Differences and Changes in Wage Structures. Chicago, IL: University of Chicago Press.

Furman, Jason. 2013. "Remarks at AEI's Center on Internet, Communications and Technology Policy." Washington, D.C., September 17.

Galor, Oded. 2011."Inequality, Human Capital Formation and the Process of Development," Working Paper 17058. Cambrige: MA: National Bureau of Economic Research.

Galor, Oded and Omer Moav. 2000. "Ability-Biased Technological Transition, Wage Inequality, and Economic Growth," Quarterly Journal of Economics, 115, 469-497.

______. 2004. "From Physical to Human Capital Accumulation: Inequality and the Process of Development." Review of Economic Studies, 71, 1001-1026.

Galor, Oded and Daniel Tsiddon. 1997. "The Distribution of Human Capital, Technological Progress, and Economic Growth." Journal of Economic Growth, 2, 93-124

GAO (Government Accountability Office). 2001. Spectrum Management: NTIA Planning and Processes Need Strengthening to Promote the Efficient Use of Spectrum by Federal Agencies. GAO-11-352.

______. 2012. Electronic Health Records: Number and Characteristics of Providers Awarded Medicaid Incentive Payments for 2011. Government Printing Office.

______. 2013a. Assessing Factors That Affect Patent Infringement Litigation Could Help Improve Patent Quality. GAO 13-465. GovernmentPrinting Office.

______. 2013b. Electronic Health Records: Number and Characteristics of Providers Awarded Medicare Incentive Payments for 2011-2012.Government Printing Office.

Gilbert, Richard J. 2010-2011. "Deal or No Deal? Licensing Negotiations in Standard-Setting Organizations." HeinOnline 77 855-888.

Goos, Martin and Alan M. Manning, 2007. "Lousy and Lovely Jobs: the rising polarization of work in Britain. Review of Economics and Statistics. 89(1): 118-133.

Goldin, Claudia, and Lawrence F. Katz. 2008. "The Race between Education and Technology." Cambridge, MA: The Belknap Press of Harvard University Press.

Gordon, Robert J. 2012. "Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds." Working Paper 18315. Cambridge, MA.: National Bureau of Economic Research.

Greenstone, M., R. Hornbeck, and E. Moretti. 2010. "Identifying Agglomeration Spillovers: Evidence from Winners and Losers of Large Plant Openings," Journal of Political Economy, 118, 536-598.

Hall, Robert E., and Charles I. Jones. 1999. "Why Do Some Countries Produce So Much More Output Per Worker Than Others?" The Quarterly Journal of Economics: 83-116.

Haskel, Jonathan, Robert Z. Lawrence, Edward E. Leamer, and Matthew J. Slaughter. 2012. "Globalization and U.S. Wages: Modifying Classic Theory to Explain Recent Facts." Journal of Economic Perspectives 26, no. 2: 119-40.

Heckman, James J. and Paul A. LaFontaine. 2010. "The American High School Graduation Rate: Trends and Levels." The Review of Economics and Statistics 92, no. 2: 244-262.

Heckman, James J., Rodrigo Pinto, and Peter A. Savelyev. 2012. "Understanding the Mechanisms Through Which an Influential Early Childhood Program Boosted Adult Outcomes." Working Paper 18581, Cambridge, MA.: National Bureau of Economic Research.

Hemphill, C. Scott, and Bhaven N. Sampat. 2011. "When Do Generics Challenge Drug Patents?." Journal of Empirical Legal Studies 8.4: 613-649.

HHS (Department of Health and Human Services). 2013. "Doctors and hospitals' use of health IT more than doubles since 2012." May 22.

Holl, Adelheid, Rafael Pardo, and Ruth Rama. 2010. "Inside and outside the factory: Just-in time manufacturing systems, subcontracting and geographic proximity." Regional Studies 44, no. 5: 519-533.

______. 2011. "Spatial Patterns of Adoption of Just-in-Time Manufacturing." Working Paper, No. 1. Spanish National Research Council.

Hong, Jun, George Huang, Pingyu Jiang, Ting Qu, Yingfeng Zahang, and Guanghui Zhou. 2010. "RFID-enabled real-time manufacturing information tracking infrastructure for extended enterprises." Journal of Intelligent Manufacturing 23 (November): 2357-2366.

Hsiao, Chun-Ju and Esther Hing. 2014. "Use and Characteristics of Electric Health Record Systems Among Office-based Physician Practices: United States, 2001-2013." NCHS Data Brief No. 143. U.S. Department of Health and Human Services.

Hur, Seung Min, Suho Jeong, and Suk-Hwan Suh. 2009. "An experimental approach to RFID system performance prediction model." International Journal of Computer Integrated Manufacturing 22, no. 7(July): 686-697.

Iowa State University Extension Service. 2013. Corn Production, Harvest and Yield.

JEC (U.S. Congress Joint Economic Committee). 2010a. The Pivotal Role of Government Investment in Basic Research. Government Printing Office.

______. 2010b. Challenges and Opportunities for Job Creation in the Aftermath of the Great Recession. Cong. 2 sess. Government Printing Office.

Jones, Charles J., and John C. Williams. 1998. "Measuring the Social Return to R&D." The Quarterly Journal of Economics (November): 1119-1135.

Jorgenson, Dale. 2001. "Information Technology and the US Economy." The American Economic Review. Vol 91 (No. 1). 1-32. March.

Jorgenson, Dale, Mun Ho, and Jon Samuels. 2012. "Information Technology and U.S. Productivity Growth." Industrial Productivity in Europe, 35-64. Northampton MA: Edward Elgar.

Jorgenson, Dale W. 1988. "Productivity and Postwar U.S. Economic Growth." Journal of Economic Perspectives." 2(4): 23-41.

Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce. 1993. "Wage Inequality and the Rise in Returns to Skill." Journal of Political Economy 101(3), 410-442.

Kahn, James and Robert Rich. 2011. "The Productivity Slowdown Reaffirmed." Federal Reserve Bank of New York.

Katz, Lawrence F. and Kevin M. Murphy. 1992. "Changed in Relative Wages, 1963-1987: Supply and Demand Factors." The Quarterly Journal of Economics 107, no. 1: 35-78.

Klein, Judith L.V., and William N. Parker. 1966. "Productivity Growth in Grain Production in the United States, 1840-60 and 1900-10." Working Paper 523-582. Cambridge, MA: National Bureau of Economic Research.

Krueger, Alan B. 1993. "How Computers Have Changed the Wage Structure: Evidence from Microdata, 1984-1989." The Quarterly Journal of Economics 108, no. 1: 33-60.

Kwerel, Evan R. and Gregory L. Rosston. 2000. "An Insiders' View of FCC Spectrum Auctions," Journal of Regulatory Economics 17: 253-289.

Lee, David. 1999. "Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?" Quarterly Journal of Economics. 114(3), 977-1023.

Lemieux, Thomas. 2008. "The Changing Nature of Wage Inequality." Journal of Population Economics 21, no. 1: 21-48.

______. 2006. "Post-Secondary Education and Increasing Wage Inequality," American Economic Review 96(2), 195-99.

Lemley, Mark A., and Carl Shapiro. 2005. "Probabilistic Patents." Journal of Economic Perspectives 19, no.2 (Spring): 75-98.

Mateyka, Peter J., and Melanie A. Rapino. 2012. "Home-Based Workers in the United States: 2010." Household Economic Studies, U.S. Census Bureau (October): 1-32.

Mishel, Lawrence, Heidi Shierholz, and John Schmitt. 2013. "Don't Blame the Robots: Assessing the Job Polarization Explanation of Growing Wage Inequality." Washington: Economic Policy Institute.

National Science Foundation. 2013. National Patterns of R&D Resources: 2010-11 Update.

National Telecommunications and information Administration. 2013. National Broadband Map. "US Broadband Availability Data."

Nelson, Richard, R. 1959. "The Simple Economics of Basic Scientific Research." Journal of Political Economy 297. 727-736.

Noonan, Mary C. and Jennifer L. Glass. 2012. "The hard truth about telecommuting." Monthly Labor Review, Statistics (June): 38-45. U.S. Bureau of Labor Statistics.

Nordhaus, William. 2004. "Retrospective on the Postwar Productivity Slow-down." Cowles Foundation Discussion Paper No. 494. Cambridge, MA: National Bureau of Economic Research.

Office of Science and Technology Policy and the National Economic Council. 2013. "Four Years of Broadband Growth." The White House. (June).

Oliner, Sichel. And Stiroh, 2007. "Explaining a Productive Decade," Finance and Economics Discussion Series 2007-63, Board of Governors of the Federal Reserve System.

Pham, Duc Truong, and Paulette Pham. 2009. "FIT Manufacturing: Linking manufacturing, marketing and product innovation strategies to achieve long term economic sustainability." Wales, UK: Cardiff University.

Piketty, Thomas and Emmanuel Saez. 2003. "Income Inequality in the United States, 1919 1998." The Quarterly Journal of Economics118, no. 1: 1-39.

_____. 2006. "The Evolution of Top Incomes: A Historical and International Perspective." Working Paper 11955. Cambridge, Mass: national Bureau of Economic Research (January).

Presidential Memorandum. 2010. "Unleashing the Broadband Wireless Revolution." June 10.

President's Council of Advisors on Science and Technology. 2012. Report to the President: Realizing the Full Potential of Government-Held Spectrum to Spur Economic Growth. The White House.

Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. "Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development." Journal of Economic Growth 9: 131-165.

Scott Morton, Fiona, and Carl Shapiro. 2013. "Strategic Patent Acquisitions." Working Paper. University of California at Berkeley. (July).

Shackleton, Robert. 2013. "Total Factor Productivity Growth in Historical Perspective." Working Paper 2013-01. Congressional Budget Office

Sim, Khim Ling, and Hian Chye Koh. 2003. "An Empirical Examination of Management Control Systems in Just-In-Time Manufacturing." The Review of Business Information Systems 7, no.3: 71-82.

Stalk, George. 1989. "Time-the next source of competitive advantage." The McKinsey Quarterly (Spring): 28-50.

Tayal, S.P. 2012. "Just In Time Manufacturing." International Journal of Applied Engineering Research 7, no.11.

USDA. 2013. Farm Computer Usage and Ownership. Government Printing Office.

U.S. House of Representatives. Committee on the Judiciary. 2011. America Invents Act. Report 112-98. Part 1 Cong.1 sess. Government Printing Office.

U.S. Patent and Trademark Office. 2013. US Patents by Technological Category.

Zhang, Yingfeng, Pingyu Jiang, George Huang, Ting Qu, Guangui Zhou, Jun Hong. 2012. "RFID-enabled real-time manufacturing information tracking infrastructure for extended enterprises."

Almond, Douglas, Kenneth Y. Chay, and Michael Greenstone. 2006. "Civil Rights, the War on Poverty, and Black-White Convergence in Infant Mortality in the Rural South and Mississippi." MIT Department of Economics Working Paper 07-04.

Anderson, Michael L. 2008. "Multiple Inference and Gender Differences in the Effects of Early Intervention: A Reevaluation of the Abecedarian, Perry Preschool, and Early Training Projects." Journal of the American Statistical Association 103, no. 484: 1481-1495.

Andersson, Fredrik, Harry J. Holzer, Julia I. Lane, David Rosenblum and Jeffrey Smith. 2013. "Does Federally-Funded Job Training Work? Nonexperimental Estimates of WIA Training Impacts Using Longitudinal Data on Workers and Firms," Working Paper 19446. Cambridge, MA: National Bureau of Economic Research.

Anzick, Michael A. and David A. Weaver. 2001. "Reducing Poverty Among Elderly Women." Working Paper Series Number 87. Office of Research, Evaluation, and Statistics, Social Security Administration.

Auten, Gerald, Geoffrey Gee, and Nicholas Turner. 2013. "Income Inequality, Mobility, and Turnover at the Top in the US, 1987-2010." American Economic Review 103 (May, Papers and Proceedings, 2012): 168-172.

Baicker, Katherine, Sarah L. Taubman, Heidi L. Allen, Mira Bernstein, Jonathan H. Gruber, Joseph P. Newhouse, Eric C. Schneider, Bill J. Wright, Alan M. Zaslavsky, and Amy N. Finkelstein. 2013. "The Oregon Experiment -- Effects of Medicaid on Clinical Outcomes." The New England Journal of Medicine 368: 1713-1722.

Bailey, Martha J. and Sheldon Danziger, eds. 2013. Legacies of the War on Poverty. New York: Russell Sage Foundation.

Ben-Shalom, Yonatan, Robert A. Moffitt, and John Karl Scholz. 2011. "An Assessment of Anti-Poverty Programs in the United States." Working Paper 17042. Cambridge, MA: National Bureau of Economic Research.

Björklund, Anders, Markus Jäntti, and Gary Solon. 2005. "Influences of Nature and Nurture on Earnings Variation: A Report on a Study of Various Sibling Types in Sweden." In Unequal Chances: Family Background and Economic Success, edited by Samuel Bowles,Herbert Gintis, and Melissa Osborne Groves, pp. 145-164. Princeton University Press.

Black, Sandra E. and Paul J. Devereux. 2011. "Recent Developments in Inter-generational Mobility." In Handbook of Labor Economics, Volume 4B, edited by Orley Ashenfelter and David Card, pp. 1487-1541. Amsterdam: North Holland Publishing Co.

Blank, Rebecca M. 1993. "Public Sector Growth and Labor Market Flexibility: The United States vs. the United Kingdom." Working Paper 4338. Cambridge, MA: National Bureau of Economic Research.

_____. 2000. "Fighting Poverty: Lessons from Recent U.S. History." Journal of Economic Perspectives 14, no. 2: 3-19.

_____. 2007. "Improving the Safety Net for Single Mothers Who Face Serious Barriers to Work." Future of Children 17, no. 2: 183-197.

Burtless, Gary. 1986. "Social Security, Unanticipated Benefit Increases, and the Timing of Retirement." Review of Economic Studies 53, no. 5: 781-805.

Campbell, Frances A., Barbara H. Wasik, Elizabeth Pungello, Margaret Burchinal, Oscar Barbarin, Kirsten Kainz, Joseph J. Sparling, and Craig T. Ramey. 2008. "Young Adult Outcomes of the Abecedarian and CARE Early Childhood Educational Interventions." Early Childhood Research Quarterly 23, no. 4: 452-466.

Cancian, Maria and Deborah Reed. 2009. "Family structure, childbearing, and parental employment: Implications for the level and trend in poverty." Focus 26, no.2: 21-26.

Caplow, Theodore and Jonathan Simon. 1999. "Understanding Prison Policy and Population Trends." Crime and Justice 26: 63-120.

Card, David E. and Steven Raphael. 2013. Immigration, Poverty, and Socio-economic Inequality. New York: Russell Sage.

Carlson, Deven, Robert Haveman, Thomas Kaplan, and Barbara Wolfe. 2011. "The Benefits and Costs of the Section 8 Housing Subsidy Program: A Framework and Estimates of First-Year Effects." Journal of Policy Analysis and Management 30, 2: 233-255.

Census Bureau. 2013. "Poverty -- Experimental Measures." http://www. census.gov/hhes/povmeas/data/nas/tables/index.html

Chay, Kenneth Y., Jonathan Guryan, and Bhashkar Mazumder. 2009. "Birth Cohort and the Black-White Achievement Gap: The Roles of Access and Health Soon After Birth." Working Paper No. 15078. Cambridge, MA, National Bureau of Economic Research.

Chetty, Raj. 2008. "Moral Hazard versus Liquidity and Optimal Unemployment Insurance." Journal of Political Economy 116, no. 2: 173-234.

Chetty, Raj, John N. Friedman, and Jonah E. Rockoff. 2011. "New Evidence on the Long-Term Impacts of Tax Credits." Statistics of Income Paper Series. Internal Revenue Service.

Chetty, Raj, John N. Friedman, and Emmanuel Saez. 2012. "Using Differences in Knowledge Across Neighborhoods to Uncover the Impacts of the EITC on Earnings." Working Paper 18232. Cambridge, MA: National Bureau of Economic Research.

Chetty, Raj, John N. Friedman, Soren Leth-Peterson, Torben Heien Nielson, and Tore Olsen. 2013. "Subsidies vs. Nudges: Which Policies Increase Saving the Most?" Issue Brief 13-3. Center for Retirement Research at Boston College.

Chetty, Raj, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez. 2014. "Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States." Working Paper 19843. Cambridge, MA: National Bureau of Economic Research.

Chetty, Raj, Nathaniel Hendren, Patrick Kline, and Emmanuel Saez, and Nick Turner. 2014. "Is the United States Still a Land of Opportunity? Recent Trends in Intergenerational Mobility." Working Paper 19844. Cambridge, MA: National Bureau of Economic Research.

(CBO) Congressional Budget Office. 2013. "Growth in Means-Tested Programs and Tax Credits for Low-Income Households." http://www.cbo.gov/sites/default/files/cbofiles/attachments/43934-Means-TestedPrograms.pdf.

Connelly, Rachel and Jean Kimmel. 2003. "The Effect of Child Care Costs on the Employment and Welfare Recipiency of Single Mothers." Southern Economic Journal 69, no. 3: 498-519.

Corak, Miles. 2006. "Do Poor Children Become Poor Adults? Lessons from a Cross Country Comparison of Generational Earnings Mobility." IZA Discussion Paper No. 1993.

_____. 2011. "Inequality from generation to generation: the United States in Comparison." Graduate School of Public and International Affairs, University of Ottawa.

Corcoran, Mary. 2001. "Mobility, Persistence, and the Consequences of Child Poverty for Children: Child and Adult Outcomes." In Understanding Poverty, edited by Sheldon H. Danziger and Robert H.Haveman, pp. 127-140. Cambridge, MA: Harvard University Press.

Corcoran, Mary and Terry Adams. 1997. "Race, Sex, and the Intergenerational Transmission of Poverty." In Consequences of Growing Up Poor, edited by Greg J. Duncan and Jeanne Brooks-Gunn, pp.461-517. New York: Russell Sage Foundation.

(CEA) Council of Economic Advisers. 1964. Economic Report of the President.

_____. 2013. "Trends in Health Care and Cost Growth and the Role of the Affordable Care Act."

Dahl, Gordon B. and Lance Lochner. 2012. "The Impact of Family Income on Child Achievement: Evidence from the Earned Income Tax Credit." American Economic Review 102, no. 5: 1927-1956.

Danziger, Sheldon and Peter Gottschalk. 1995. America Unequal. Cambridge, MA: Harvard University Press.

Danziger, Sheldon H., Lesley J. Turner, and Kristin S. Seefeldt. 2006. "Failing the Transition from Welfare to Work: Women Chronically Disconnected from Employment and Cash Welfare." Social Science Quarterly 87, no. 2: 227-249.

Deming, David. 2009. "Early Childhood Intervention and Life-Cycle Skill Development: Evidence from Head Start." American Economic Journal: Applied Economics 1, no. 3: 111-134.

DiNardo, John, Nicole M. Fortin, and Thomas Lemieux. 1996. "Labor Market Institutions and the Distribution of Wages, 1973-1992: A Semiparametric Approach." Econometrica 64, no. 5: 1001-1044.

DiNardo, John, and Thomas Lemieux. 1997. "Diverging Male Wage Inequality in the United States and Canada, 1981-1988: Do Institutions Explain the Difference?" Industrial and Labor Relations Review 50, no. 4: 629-651.

Doucouliagos, Hirstos and T.D. Stanley. 2009. "Publication Selection Bias in Minimum-Wage Research? A Meta-Regression Analysis." British Journal of Industrial Relations 47, no. 2: 406-428.

Dowd, Tim and John B. Horowitz. 2011. "Income Mobility and the Earned Income Tax Credit: Short-Term Safety Net or Long-Term Income Support." Public Finance Review 39, no. 5: 619-652.

Dube, Arindrajit, T. William Lester, and Michael Reich. 2010. "Minimum Wage Effects Across State Borders: Estimates Using Contiguous Counties." The Review of Economics and Statistics 92, no. 4: 945-964.

Dube, Arindrajit. 2013. "Minimum Wages and the Distribution of Family Incomes." University of Massachusetts, Amherst Working Paper. https://dl.dropboxusercontent.com/u/15038936/Dube_Minimum-WagesFamilyIncomes.pdf.

Duncan, Greg J., Katherine Magnuson, Ariel Kalil, Kathleen Ziol-Guest. 2012. "The Importance of Early Childhood Poverty." Social Indicators Research 108, no. 1: pp 87-98.

Eissa, Nada and Hilary W. Hoynes. 2005. "Behavioral Responses to Taxes: Lessons from the EITC and Labor Supply." Working Paper 11729. Cambridge, MA: National Bureau of Economic Research.

Eissa, Nada and Jeffrey B. Liebman. 1996. "Labor Supply Response to the Earned Income Tax Credit." The Quarterly Journal of Economics 111, no. 2: 605-637.

Engelhardt, Gary V., and Jonathan Gruber. 2006. "Social Security and the Evolution of Elderly Poverty." In Public Policy and the Income Distribution, edited by Alan J. Auerbach, David E. Card, and JohnM. Quigley, pp. 259-287. New York: Russell Sage Foundation.

Fisher, Gordon M. 1992. "The Development and History of the Poverty Thresholds." Social Security Bulletin 55, no. 4: 3-14.

Fox, Liana, Irwin Garfinkel, Neeraj Kaushal, Jane Waldfogel, and Christopher Wimer. 2013. "Waging War on Poverty: Historical Trends in Poverty Using the Supplemental Poverty Measure." Working Paper 13-01. New York: Columbia Population Research Center.

Fremstad, Shawn. 2009. "Half in Ten: Why Taking Disability into Account is Essential to Reducing Income Poverty and Expanding Economic Inclusion." Reports and Issue Briefs 2009-30. Washington: Center for Economic and Policy Research.

Garces, Eliana, Duncan Thomas, and Janet Currie. 2002. "Longer-Term Effects of Head Start." American Economic Review 92, no. 4: 999-1012.

Gibbs, Chloe, Jens Ludwig, and Douglas L. Miller. 2013. "Head Start Origins and Impacts." In Legacies of the War on Poverty, edited by Martha Bailey and Sheldon Danziger, pp. 39-65. New York: Russell Sage Foundation Press.

Gottschalk, Peter and Sheldon Danziger. 2003. "Wage Inequality, Earnings Inequality and Poverty in the U.S. Over the Last Quarter of the Twentieth Century." Working Papers in Economics 560. Boston College Department of Economics.

Hanson, Kenneth. 2010. "The Food Assistance National Income-Output Multiplier (FANIOM) Model and Stimulus Effects of SNAP." Economic Research Report No. 103. U.S. Department of Agriculture Economic Research Service.

Harrington, Michael. 1962. The Other America. New York: Simon and Schuster.

Heckman, James J. and Dimitriy V. Masterov. 2007. "The Productivity Argument for Investing in Young Children." Applied Economic Perspectives and Policy 29, no. 3: 446-493.

Heckman, James J., Seong Hyeok Moon, Rodrigo Pinto, Peter A. Savelyev, and Adam Yavitz. 2009. "The Rate of Return to the High/Scope Perry Preschool Program." Working Paper No. 15471. Cambridge, MA: National Bureau of Economic Research.

_____. 2010. "A New Cost-Benefit and Rate of Return Analysis for the Perry Preschool Program: A Summary." Working Paper No. 16180. Cambridge, Mass.: National Bureau of Economic Research.

Heckman, James J., Rodrigo Pinto, Azeem M. Shaikh, and Adam Yavitz. 2011. "Inference with Imperfect Randomization: the Case of the Perry Preschool Program." Working Paper No. 16935. Cambridge, MA: National Bureau of Economic Research.

Holzer, Harry J. 2007. "Collateral Costs: The Effects of Incarceration on the Employment and Earnings of Young Workers." IZA Discussion Paper No. 3118.

Holzer, Harry J., Diane Whitmore Schanzenbach, Greg J. Duncan, and Jens Ludwig. 2008. "The Economic Costs of Childhood Poverty in the United States." Journal of Children and Poverty 14, no. 1: 41-61.

Hotz, V. Joseph and John Karl Scholz. 2003. "The Earned Income Tax Credit." In Means-Tested Transfer Programs in the U.S., edited by Robert A. Moffitt, pp. 141-198. Chicago: University of Chicago Press.

Hoynes, Hilary W., Marianne E. Page, and Ann Huff Stevens. 2006. "Poverty in America: Trends and Explanations." Journal of Economic Perspectives 92, no. 3: 748-765.

Hoynes, Hilary W. and Diane W. Schanzenbach. 2009. "Consumption Reponses to In-Kind Transfers: Evidence from the Introduction of the Food Stamp Program." American Economic Journal: Applied Economics 1, no. 4: 109-139.

_____. 2012. "Work incentives and the Food Stamp Program." Journal of Public Economics 96, no. 1: 151-162.

Hoynes, Hilary W., Diane W. Schanzenbach, and Douglas Almond. 2013. "Long Run Impacts of Childhood Access to the Safety Net." Working Paper 18535. Cambridge, MA: National Bureau of Economic Research.

Ichino, Andrea, Loukas Karabarbounis, and Enrico Moretti. 2009. "The Political Economy of Intergenerational Income Mobility." IZA Discussion Paper No. 4767.

Isaacs, Julia B. 2008. "Economic Mobility of Families across Generations." In Getting Ahead or Losing Ground: Economic Mobility in America, edited by Julia Isaacs, Isabel Sawhill and Ron Haskins. Washington: The Pew Charitable Trusts.

Isaacs, Julia B., Isabel Sawhill, and Ron Haskins. 2008. Getting Ahead or Losing Ground: Economic Mobility in America. Washington: ThePew Charitable Trusts.

Jacob, Brian A. and Jens Ludwig. 2012. "The Effects of Housing Assistance on Labor Supply: Evidence from a Voucher Lottery." American Economic Review 102, no. 1: 272-304.

Jäntti, Markus, Bernt Bratsberg, Knut Røed, Oddbjørn Raaum, Robin Naylor, Eva Österbacka, Anders Björklund, and Tor Eriksson. 2006. "American Exceptionalism in a New Light: A Comparison of Inter-generational Earnings Mobility in the Nordic Countries, the United Kingdom and the United States." IZA Discussion Paper No. 1938.

Johnson, Rucker C. 2008. "Ever-increasing Levels of Parental Incarceration and the Consequences for Children." In Do Prisons Make us Safer?, edited by S. Raphael and M. Stoll, 177-206. New York: Russell Sage Foundation.

Korenman, Sanders and Dahlia Remler. 2013. "Rethinking Elderly Poverty: Time for a Health Inclusive Poverty Measure?" Working Paper 18900. Cambridge, MA: National Bureau of Economic Research.

Lee, Chul-In and Gary Solon, 2009. "Trends in Intergenerational Income Mobility," The Review of Economics and Statistics, 91, no. 4: 766-772

Lee, David S. 1999. "Wage Inequality in the United States During the 1980s: Rising Dispersion or Falling Minimum Wage?" Quarterly Journal of Economics 114, no. 3: 977-1023.

Lemieux, Thomas. 2008. "The Changing Nature of Wage Inequality." Journal of Population Economics 21, no. 1: 21-48.

Liebman, Jeffrey B. 1998. "The Impact of the Earned Income Tax Credit on Incentives and Income Distribution." In Tax Policy and the Economy, Volume 12, edited by James M. Poterba, pp. 83-120. Cambridge, MA: National Bureau of Economic Research.

Ludwig, Jens and Douglas Miller. 2007. "Does Head Start Improve Children's Life Chances? Evidence from a Regression Discontinuity Design." Quarterly Journal of Economics 122, no. 1: 159-208.

Matsudaira, Jordan D. and Rebecca M. Blank. 2013. "The Impact of Earnings Disregards on the Behavior of Low-Income Families." Journal of Policy Analysis and Management 33, no. 1: 7-35.

Mayer, Susan E. and Leonard M. Lopoo. 2008. "Government Spending and Intergenerational Mobility." Journal of Public Economics 92, no. 1-2: 139-158.

Mayer, Susan E. 1997. What Money Can't Buy: Family Income and Children's Life Chances. Cambridge, MA: Harvard University Press.

Meyer, Bruce D., Wallace K. C. Mok, and James X. Sullivan. 2009. "The Underreporting of Transfers in Household Surveys: Its Nature and Consequences." Working Paper 15181. Cambridge, MA: National Bureau of Economic Research.

Meyer, Bruce D. and Dan T. Rosenbaum. 2001. "Welfare, the Earned Income Tax Credit, and the Labor Supply of Single Mothers." Quarterly Journal of Economics 116, no. 3: 1063-1114.

Meyer, Bruce D. and James X. Sullivan. 2003. "Measuring the Well-Being of the Poor Using Income and Consumption." Journal of Human Resources 38, Supplement: 1180-1220.

_____. 2012a. "Winning the War: Poverty from the Great Society to the Great Recession." Brookings Papers on Economic Activity 45, no. 2: 133-200.

_____. 2012b. "Identifying the Disadvantaged: Official Poverty, Consumption Poverty, and the New Supplemental Poverty Measure." Journal of Economic Perspectives 26, no. 3: 111-136.

_____. 2013. "Winning the War: Poverty from the Great Society to the Great Recession." Working Paper 18718. Cambridge, MA: National Bureau of Economic Research.

Mishel, Lawrence, Josh Bivens, Elise Gould, and Heidi Shierholz. The State of Working America, 12th Edition. A forthcoming Economic PolicyInstitute book. Ithaca, NY: Cornell University Press.

Misra, Joya, Michelle Budig, and Irene Boeckmann. 2011. "Work-Family Policies and the Effects of Children on Women's Employment Hours and Wages." Community, Work and Family 14, no. 2: 139-157.

Nicholson-Crotty, Sean and Kenneth J. Meier. 2003. "Crime and Punishment: The Politics of Federal Criminal Justice Sanctions." Political Research Quarterly 56, no. 2: 119-126.

Orshansky, Mollie. 1965. "Counting the Poor: Another Look at the Poverty Profile." Social Security Bulletin 28, no. 1: 3-29.

Peri, Giovanni. 2013. "Immigrant Workers, Native Poverty and Labor Market Competition." Policy Brief, Center for Poverty Research 1, no. 3.

Peters, Alan H. and Peter S. Fisher. 2002. "State Enterprise Zones: Have They Worked?" Kalamazoo, MI: W.E. Upjohn Institute for Employment Research Press.

Piketty, Thomas and Emmanuel Saez. 2003. "Income Inequality in the United States, 1913-1998." Quarterly Journal of Economics 118, no. 1: 1-39.

Raphael, Steven. 2007. "Early Incarceration Spells and the Transition to Adulthood." In The Price of Independence: The Economics of Early Adulthood, edited by Sheldon Danziger and Cecilia Elena Rouse, pp.278-306. New York: Russell Sage Foundation.

Reed, Deborah and Maria Cancian. 2001. "Sources of Inequality: Measuring the Contributions of Income Sources to Rising Family Income Inequality." Review of Income and Wealth 47, no. 3: 321-333.

Sawhill, Isabel V. and John E. Morton. 2007. "Economic Mobility: Is the American Dream Alive and Well?" Washington: Economic Mobility Project, Pew Charitable Trusts.

Schur, Lisa A., Douglas L. Kruse, and Peter Blanck. 2013. People with Disabilities: Sidelined or Mainstreamed? Cambridge, England: CambridgeUniversity Press.

Schweinhart, Lawrence J., Jeanne Montie, Zongping Xiang, W. Steven Barnett, Clive R. Belfield, and Milagros Nores. 2005. Lifetime Effects: The High/Scope Perry Preschool Study Through Age 40. Monographs of the High/Scope Educational Research Foundation. Ypsilanti, MI: High/Scope Press.

Sen, Amartya. 2009. The Idea of Justice. London: Allen Lane.

Singh, Gopal K. and Michael D. Kogan. 2007. "Persistent Socioeconomic Disparities in Infant, Neonatal, and Postneonatal Mortality Rates in the United States, 1969-2001." Pediatrics 119, no. 4: 928-939.

Sharkey, Patrick. 2009. "Neighborhoods and the Black-White Mobility Gap." Washington: Economic Mobility Project, Pew Charitable Trusts.

She, Peiyun and Gina A. Livermore. 2007. "Material Hardship, Poverty, and Disability Among Working Age Adults." Social Science Quarterly 88, no. 4: 970-989.

Sherman, Arloc. 2013. "Official Poverty Measure Masks Gains Made Over Last 50 Years." Washington: Center on Budget and Policy Priorities. http://www.cbpp.org/files/9-13-13pov.pdf.

Short, Kathleen. 2012. "The Research Supplemental Poverty Measure: 2011" Current Population Reports.

_____. 2013. "The Research Supplemental Poverty Measure: 2012" Current Population Reports.

Shroder, Mark. 2010. "Housing Subsidies and Work Incentives." MPRA Paper 26019, University Library of Munich, Germany.

Solon, Gary. 2002. "Cross-Country Differences in Intergenerational Earnings Mobility." Journal of Economic Perspectives 16, no. 3: 59-66.

SRI International. 1983. "Final Report of the Seattle/Denver Income Maintenance Experiment. Vol 1: Design and Results." Menlo Park, CA.

(SSA) Social Security Administration. 2012. Income of the Population 55 or Older, 2010. SSA Publication No. 13-11871.

Western, Bruce. 2002. "The Impact of Incarceration on Wage Mobility and Inequality."

American Sociological Review 67: 526-46.

Western, Bruce and Becky Pettit. 2010. "Incarceration and Social Inequality." Daedalus 139: 8-19.

Wimer, Christopher, Liana Fox, Irwin Garfinkel, Neeraj Kaushal, and Jane Waldfogel. 2013. "Trends in Poverty with an Anchored Supplemental Poverty Measure." Working Paper 1-25. New York: Columbia Population Research Center.

Angrist, Joshua D., and Jörn-Steffen Pischke. 2008. Mostly Harmless Econometrics: An Empiricist's Companion. Princeton, NJ: Princeton University Press.

Angrist, Joshua. D., and Alan B. Krueger. 1999. "Empirical strategies in labor economics." In Handbook of Labor Economics, edited by O. Ashen-felter and D. Card, pp.1277-1366. Vol 3, Elsevier.

Baicker, Katherine, Amy Finkelstein, Jae Song, and Sarah Taubman. 2013. "The Impact of Medicaid on Labor Force Activity and Program Participation: Evidence from the Oregon Health Insurance Experiment." Working Paper. Cambridge, MA: National Bureau of Economic Research (October).

Benus, J. T. Shen, S. Zhang, M. Chan and B. Hansen. 2009. "Growing America Through Entrepreneurship: Final Evaluation of Project GATE." Columbia, MD: IMPAQ International LLC.

Benus, J., Poe-Yamagata, E., Wang, Y., & Blass, E. 2008. "Reemployment and Eligibility Assessment (REA) Study FY 2005 Initiative." Columbia, MD: IMPAQ International LLC.

Berk, Richard A., and Peter H. Rossi. 1998. Thinking about Program Evaluation 2. 2^nd ed. London: Sage Publications, Inc.

Bettinger, E.P., B.T. Long, P. Oreopoulos and L. Sanbonmatsu. 2009. "The Role Of Simplification and Information In College Decisions: Results From The H&R Block FAFSA Experiment." Working Paper 15361. Cambridge, MA: National Bureau of Economic Research (September).

Burtless, G. 1995. (The case for randomized field trials in economic and policy research.) The Journal of Economic Perspectives 9, no. 2 (Spring): 63-84.

Cappers, Peter, Charles Goldman, and Annika Todd. 2013. "Smart Grid Investment Grant Consumer Behavior Study Analysis: Summary of Utility Studies." Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory.

Card, David, Carlos Dobkin, and Nicole Maestas. 2009. "Does Medicare Save Lives?" The Quarterly Journal of Economics 124, no. 2: 597-636.

Coalition for Evidence-Based Policy. 2012. "Rigorous Program Evaluations on a Budget: How Low-Cost Randomized Controlled Trials are Possible in Many Areas of Social Policy." Washington, DC (March).

_____. 2013. "Practical Evaluation Strategies for Building a Body of Proven-Effective Social Programs." Washington, DC (October).

Comey, Jennifer, Kaitlin Franks, Lesley Freiman, Christopher Hayes, Reed Jordan, Peter A. Tatian, and Mary K. Winkler. 2013. "Measuring Performance: A Guidance Document for Promise Neighborhoods on Collecting Data and Reporting Results." Washington, DC: The Urban Institute (February).

Council Of Economic Advisers And Department Of Labor. 2013. "The Economic Benefits of Extending Unemployment Insurance" (December).

Culhane, D. P., S. Metraux, J.M. Park, M.A. Schretzman, M. A., and J. Valente. 2007. "Testing a Typology of Family Homelessness Based on Patterns of Public Shelter Utilization in Four U.S. Jurisdictions: Implications for Policy and Program Planning." Housing Policy Debate, 18(1), 1-28.

Department of Education. 2012. Education Improvement Programs: Fiscal Year 2012 Budget Request. http://www2.ed.gov/about/overview/budget/budget12/justifications/d-eip.pdf

_____. 2013a. Investing in Innovation Fund (i3) Program Guidance and Frequently Asked Questions (FAQs). http://www2.ed.gov/programs/innovation/faq.html

_____. 2013b. Promise Neighborhoods: Purpose. http://www2.ed.gov/programs/promiseneighborhoods/index.html

_____. 2013c. PROMISE: Promoting Readiness of Minors in Supplemental Security Income. http://www2.ed.gov/about/inits/ed/promise/index.html#about

Department of Labor. 2011. Employment and Training Administration Notice of Availability of Funds and Solicitation for Grant Applications for Workforce Innovation Fund Grants. SGA/DFA PY-11-05at http://www.doleta.gov/workforce_innovation/applicant_information.cfm

_____. 2013. Congressional Budget Justification: Employment and Training Administration. http://www.dol.gov/dol/budget/2013/PDF/CBJ-2013-V1-05.pdf

Departments of Labor, Health and Human Services, and Education, and Related Agencies Appropriations Act, 2006, Title V, Part D (P.L. 109-149). http://www2.ed.gov/programs/teacherincentive/legislation.html)

Dobbie, Will and Roland G. Fryer. 2011. "Are High Quality Schools Enough to Increase Achievement Among the Poor? Evidence from the Harlem Children's Zone." American Economic Journal: Applied Economics 3 (July):158-187.

Finkelstein, Amy, Sarah Taubman, Bill Wright, Mira Bernstein, Jonathan Gruber, Joseph P. Newhouse, Heidi Allen, Katherine Baicker, and the Oregon Health Study Group. 2012. "The Oregon Health Insur ance Experiment: Evidence from the First Year." Quarterly Journal of Economics 127(3): 1057-1106 (August).

General Accounting Office. 2011. Performance Measurement and Evaluation: Definitions and Relationships. GAO-11-646SP, May.

_____. 2012. "Designing Evaluations: 2012 Revision." U.S. Government Accountability Office, GAO-12-208G, January.

Grossman, J.B., and J.E. Rhodes. 2002. "The test of time: Predictors and effects of duration in youth mentoring programs." American Journal of Community Psychology (30), 199-219.

Jacob, Brian A., Jens Ludwig. 2012. "The Effects of Housing Assistance On Labor Supply: Evidence From A Voucher Lottery." American Economic Review 102(1): 272-304.

Holland, Paul W. 1986. "Statistics and Causal Inference." Journal of the American Statistical Association 81, no. 396: 945-960.

Imbens, Guido W. and Jeffrey M. Woolridge. 2009. "Recent Developments in the Econometrics of Program Evaluation." Journal of Economic Literature 47, no. 1: 5-86.

Imbens, Guido W. 2010. "Better LATE Than Nothing: Some Comments on Deaton (2009) and Heckman and Urzua (2009)." Journal of Economic Literature 48 (June 2010): 399-423.

Ludwig, J. and D. L. Miller. 2007. "Does Head Start improve children's life chances? Evidence From a Regression Discontinuity Design." Quarterly Journal of Economics 122 no. 1: 159-208.

Manzi, Jim. 2012. Uncontrolled: The Surprising Payoff of Trial-and-Error for Business, Politics, and Society. New York: Basic Books.

Millensky, Megan, et al. 2011. "Staying on Course: Three-Year Results of the National Guard Youth ChalleNGe Evaluation. MDRC. http://www.mdrc.org/sites/default/files/full_510.pdf

National Institute of Justice. 2012. "Swift and Certain" Sanctions in Probation Are Highly Effective: Evaluation of the Hope Program. http://www.nij.gov/topics/corrections/community/drug-offenders/hawaii-hope.htm

National Research Council and Institute of Medicine. 2009. Preventing Mental, Emotional, and Behavioral Disorders Among Young People: Progress and Possibilities. Committee on Prevention of MentalDisorders and Substance Abuse Among Children, Youth and Young Adults: Research Advances and Promising Interventions. Mary Ellen O'Connell, Thomas Boat, and Kenneth E. Warner, Editors. Board on Children, Youth, and Families, Division of Behavioral and Social Sciences and Education. Washington, DC: The National Academies Press. Online at http://www.nap.edu/catalog.php?record_id=12480.

Office of Management and Budget. 2009a. Analytical Perspectives: Budget of the U.S. Government, February. At http://www.gpo.gov/fdsys/pkg/BUDGET-2010-PER/pdf/BUDGET-2010-PER.pdf

_____. 2009b. Building Rigorous Evidence to Drive Policy. OMB Blog-post, http://www.whitehouse.gov/omb/blog/09/06/08/BuildingRigorousEvidencetoDrivePolicy

_____. 2009c. Increased Emphasis on Program Evaluations. Memorandum M-10-01, October 7, 2009.

_____. 2010. Analytical Perspectives: Budget of the U.S. Government, at http://www.gpo.gov/fdsys/pkg/BUDGET-2011-PER/pdf/BUDGET-2011-PER.pdf

_____. 2011. Analytical Perspectives: Budget of the U.S. Government, February 14, 2011 at http://www.gpo.gov/fdsys/pkg/BUDGET-2012-PER/pdf/BUDGET-2012-PER.pdf

_____. 2012. Analytical Perspectives: Budget of the U.S. Government, February 13, 2012 at http://www.gpo.gov/fdsys/pkg/BUDGET-2013-PER/pdf/BUDGET-2013-PER.pdf

_____. 2013a. Creating a 21^st Century Government, Budget of the United States Government, Fiscal Year 2014 http://www.whitehouse.gov/sites/default/files/omb/budget/fy2014/assets/21st_century.pdf

_____. 2013b. Next Steps in the Evidence and Innovation Agenda. Memorandum M-13-17, July 26, 2013.

_____. 2014. Guidance for Providing and Using Administrative Data for Statistical Purposes. Memorandum M-14-06, February 14, 2014.

Paulsell, Diane, S. Avellar, E. Sama Martin, P. Del Grosso. 2010. "Home visiting Evidence of Effectiveness Review: Executive Summary." Washington, DC:Mathematica Policy Research.

Perez-Arce, Francisco, et al. 2012. "A Cost-Benefit Analysis of the National Guard Youth ChalleNGe Program." RAND Corporation. http:// www.rand.org/pubs/technical_reports/TR1193.html#key-findings

Poe-Yamagata, E., Benus, J., Bill, N., Carrington, H., Michaelides, M., & Shen, T. 2011. "Impact of the Reemployment and Eligibility Assessment (REA) Initiative." Columbia, MD: IMPAQ International LLC.

Rivkin, Steven G., Eric A. Hanushek and John F. Kain. 2005. "Teachers, Schools, and Academic Achievement." Econometrica, vol. 73 (2) (March): 417-458.

Rothstein, Jesse. 2011. "Unemployment Insurance and Job Search in the Great Recession." Brookings Papers on Economic Activity Fall 2011: 143-213.

Shadish, William R., Thomas D. Cook, and Donald Thomas Campbell. 2002. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Boston, MA: Houghton Mifflin.

Schlafer, Rebecca J., Poehlmann, Julie, Coffino, Brianna, Henneman, Ashley (2009). "Mentoring Children with Incarcerated Parents: Implications for Research, Practice, and Policy." Family Relations (58), 507-519.

Schochet, Peter Z., John Burghardt, and Sheena McConnell. 2008. "Does Job Corps Work? Impact Findings from the National Job Corps Study." American Economic Review 98:5, 1864-1886. http://www.jstor.org/stable/29730155?seq=2

Shadish, William R., Thomas D. Cook, and Donald Thomas Campbell. 2002. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Boston, MA: Houghton Mifflin.

Shrank, W. 2013. "The Center for Medicare and Medicaid Innovation's Blueprint for Rapid-Cycle Evaluation of New Care and Payment Models." Health Affairs 32(4):807-812.

Social Security Administration. 2013. Solicitation: PROMISE Evaluation Statement of Work. SSA-RFP-13-0018, July 16, 2013. https://www.fedconnect.net/FedConnect/PublicPages/PublicSearch/Public_Opportunities.aspx

Stock, James H., and Mark Watson. 2010. Introduction to Econometrics. 3^rd ed. Boston, MA: Addison-Wesley.

St.Pierre, R., A. Ricciuti, F. Tao, C. Creps, J. Swartz, W. Lee, A. Parsad and T. Rimdzius. 2003. "Third National Even Start Evaluation: Program Impacts and Implications for Improvement." Abt Associates, Inc. and U.S. Department of Education.

United States Interagency Council on Homelessness. 2013a. "US Labor Department awards nearly $24 million in Pay for Success grants." September 26, 2013, available at http://usich.gov/media_center/news/us_labor_department_awards_nearly_24_million_in_pay_for_success_grants (accessed November 24, 2013).

United States Interagency Council on Homelessness, 2013b. "Opening Doors," available at http://usich.gov/opening_doors/ (accessed November 24, 2013).

Walker, Robert, Hoggart, Lesley, and Hamilton Gayle. 2006. "Making Random Assignment Happen: Evidence from the UK Employment Retention and Advancement (ERA) Demonstration." Policy Studies Institute (PSI), the Office for National Statistics (ONS), the Institute for Fiscal Studies (IFS) and MDRC for the UK Department for Work and Pensions.

Wansink, Brian, David Just, and Laura Smith. 2011. "Move the Fruit: Putting Fruit in New Bowls and New Places Doubles Lunchroom Sales." Journal of Nutrition Education and Behavior 43:4.

The Council of Economic Advisers submits this report on its activities during calendar year 2013 in accordance with the requirements of the Congress, as set forth in section 10(d) of the Employment Act of 1946 as amended by the Full Employment and Balanced Growth Act of 1978.

Council Members and Their Dates of Service

 

 

 

 

 

 

Report to the President

 

on the Activities of the

 

Council of Economic Advisers

 

During 2013

 

From 2007 to 2008 Furman was a Senior Fellow in Economic Studies and Director of the Hamilton Project at the Brookings Institute. Previously, he served as a Staff Economist at the Council of Economic Advisers, a Special Assistant to the President for Economic Policy at the National Economic Council under President Clinton and Senior Adviser to the Chief Economist and Senior Vice President of the World Bank. Furman was the Economic Policy Director for Obama for America. Furman has also served as Visiting Scholar at NYU's Wagner Graduate School of Public Service, a visiting lecturer at Yale and Columbia Universities, and a Senior Fellow at the Center on Budget and Policy Priorities.

Alan B. Krueger resigned as Chairman on August 2, 2013 to return to Princeton University, where he is the Bendheim Professor of Economics and Public Affairs.

James H. Stock was appointed by the President on February 7, 2013. He served as Chief Economist of the Council of Economic Advisers from September 12, 2012 until then. Dr. Stock is on leave from Harvard University, where he is the Harold Hitchings Burbank Professor of Political Economy. Dr. Stock served as the Chair of the Harvard University Department of Economics from 2006 to 2009.

Katharine G. Abraham resigned as Member of the Council on April 19, 2013 to return to the University of Maryland, where she is a professor in the Joint Program in Survey Methodology and faculty associate in the Maryland Population Research Center.

The Council works closely with various government agencies, including the National Economic Council, the Office of Management and Budget, White House senior staff, and other officials and engages in discussions on numerous policy matters. In the area of international economic policy, the Council coordinates with other units of the White House, the Treasury Department, the State Department, the Commerce Department, and the Federal Reserve on matters related to the global financial system.

Among the specific economic policy areas that received attention in 2013 were: housing policies; health care cost growth and the Affordable Care Act; individual and corporate taxation; college affordability and ratings; regional development; the economic cost of carbon pollution; renewable fuel standards; energy policy; intellectual property and innovation; infrastructure investment; regulatory measures; trade policies; poverty and income inequality; unemployment insurance and the minimum wage; labor force participation; job training; and foreign direct investment. The Council also worked on several issues related to the quality of the data available for assessing economic conditions.

The Council prepares for the President, the Vice President, and the White House senior staff a daily economic briefing memo analyzing current economic developments and almost-daily memos on key economic data releases. Chairman Furman also presents a monthly briefing on the state of the economy and the Council's energy analysis to senior White House officials.

The Council, the Department of Treasury, and the Office of Management and Budget -- the Administration's economic "troika" -- are responsible for producing the economic forecasts that underlie the Administration's budget proposals. The Council initiates the forecasting process twice each year, consulting with a wide variety of outside sources, including leading private sector forecasters and other government agencies.

The Council was an active participant in the trade policy process, participating in the Trade Policy Staff Committee and the Trade Policy Review Group. The Council provided analysis and opinions on a range of trade-related issues involving the enforcement of existing trade agreements, reviews of current U.S. trade policies, and consideration of future policies. The Council also participated on the Trade Promotion Coordinating Committee, helping to examine the ways in which exports may support economic growth in the years to come. In the area of investment and security, the Council participated on the Committee on Foreign Investment in the United States (CFIUS), reviewing individual cases before the committee.

The Council is a leading participant in the Organisation for Economic Cooperation and Development (OECD), an important forum for economic cooperation among high-income industrial economies. The Council coordinated and oversaw the OECD's review of the U.S. economy. Chairman Furman is chairman of the OECD's Economic Policy Committee, and Council Members and staff participate actively in working-party meetings on macroeconomic policy and coordination and contribute to the OECD's research agenda.

The Council issued a wide range of reports in 2013 and early 2014. In June, the Council released a report assessing the economic benefits of reforming US's "broken immigration system." In October, the Council released a report analyzing the negative impact that the government shutdown and debt limit brinksmanship had on the economy. Also in October, the Council released a report describing why the US provides such an attractive investment climate for firms across the globe and how this is beneficial to the US. In November, the Council released a report analyzing the recent health care cost trends in addition to the contributions the Affordable Care Act has had on reducing health care cost growth. In December, the Council worked with the Department of Labor to study the benefits of extending unemployment insurance. In January 2014, the Council released a progress report on the "War on Poverty" that Lyndon B. Johnson declared 50 years prior. In February 2014, the Council transmitted the final report on the fifth Anniversary of the Recovery Act to Congress. All of the aforementioned reports can be found on the Council's website, and some of them are incorporated into this annual report as well. (http://www.whitehouse.gov/ administration/eop/cea/factsheets-reports.)

The Council continued its efforts to improve the public's understanding of economic developments and of the Administration's economic policies through briefings with the economic and financial press, speeches, discussions with outside economists, and regular updates on major data releases and postings of CEA's Reports on the White House and CEA blogs. The Chairman and Members also regularly met to exchange views on the economy with the Chairman and Members of the Board of Governors of the Federal Reserve System.

The Council frequently prepared reports and blog posts in 2013, and the Chairman and Members gave numerous public speeches. The reports, posts and texts of speeches are available at the Council's website, www. whitehouse.gov/cea. Finally, the Council published the monthly Economic Indicators, which is available online at www.gpo.gov/economicindicators.

Brian A. Amorosi

Doris T. Searles

The senior economists who resigned in 2013 (with the institutions to which they returned after leaving the Council in parentheses) were: Bevin Ashenmiller (Occidental University), Benjamin H. Harris (Brookings Institution), Susan Helper (Case Western University), Justin Joffrion (U.S. Air Force Academy), Chinhui Juhn (University of Houston), Paul Lengermann (Federal Reserve Board), Emily Y. Lin (U.S. Department of the Treasury), Rodney D. Ludema (Georgetown University), James M. Williamson (U.S. Department of Agriculture), and Wesley Yin (UCLA).

The economists who departed in 2013 were David Cho (Princeton University) and Judd N.L. Cramer (Princeton University). David served the CEA for more than two years and was a recipient of the Robert M. Solow Award for Distinguished Service.

The staff economists who departed in 2013 were Nicholas Li, Ben Meiselman, Nicholas Tilipman, Lee Tucker, and Jeffrey Y. Zhang.

The research economists who departed in 2013 were Carys Golesworthy, Dina Grossman, and Spencer Smith.

Petra S. Starke resigned from her position as General Counsel. Michael Bourgeois resigned from his position as Special Assistant to the Chairman. Emily C. Berret resigned from her position as Special Assistant to the Members. Sarah A. Murray resigned from her position as Economic Statistician. Thomas F. Hunt resigned from his position as Staff Assistant.

Unless otherwise noted, all dollar figures are in current dollars.

Because of the formula used for calculating real gross domestic product (GDP), the chained (2009) dollar estimates for the detailed components do not add to the chained-dollar value of GDP or to any intermediate aggregate. The Department of Commerce (Bureau of Economic Analysis) no longer publishes chained-dollar estimates prior to 1999, except for selected series.

Symbols used:

Excel versions of these tables are available at www.gpo.gov/erp.

Data presented in the past remain available to the public through their source agencies. For each table in last year's statistical appendix, the Sources section starting on page 397 lists source agency (or agencies), website, and data program for the data featured, along with selected contact information.

                    2013-TO-2014 TABLE NUMBER MATCH

Statistical agencies and data aggregators also offer tools allowing users to download, graph, map, and program data themselves. The Federal Reserve Bank of St. Louis, a notable aggregator of economic data, features an online database at its Federal Reserve Economic Data (FRED) site comprising more than 154,000 economic time series from 59 national, international, public, and private data sources. In addition to mobile apps and other data tools, FRED provides application programming interfaces (APIs) to allow developers to create applications or programs that directly utilize its website content. For more information, see www.research.stlouisfed.org/fred2.

 

GDP, Income, Prices, and Selected Indicators

 

 

Table B-1. Percent changes in real gross domestic product, 1965-2013

 

[Percent change from preceding period;

 

quarterly data at seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 366.]

 

 

Table B-2. Gross domestic product, 1999-2013

 

[Quarterly data at seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 368.]

 

 

Table B-3. Quantity and price indexes for gross domestic product,

 

and percent changes, 1965-2013

 

[Quarterly data are seasonally adjusted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 370.]

 

 

Table B-4. Growth rates in real gross domestic product

 

by area and country, 1995-2014

 

[Percent change]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 371.]

 

 

Table B-5. Real exports and imports of goods and services, 1999-2013

 

[Billions of chained (2009) dollars;

 

quarterly data at seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 372.]

 

 

Table B-6. Corporate profits by industry, 1965-2013

 

[Billions of dollars; quarterly data at

 

seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 373.]

 

 

Table B-7. Real farm income, 1950-2014

 

[Billions of chained (2009) dollars]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 374.]

 

 

Table B-8. New private housing units started, authorized, and

 

completed and houses sold, 1970-2014

 

[Thousands; monthly data at seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 375.]

 

 

Table B-9. Median money income (in 2012 dollars) and

 

poverty status of families and people, by race, 2003-2012

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 376.]

 

 

Table B-10. Changes in consumer price indexes, 1945-2013

 

[For all urban consumers; percent change]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 377.]

 

 

Labor Market Indicators

 

 

Table B-11. Civilian population and labor force, 1929-2014

 

[Monthly data seasonally adjusted, except as noted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 378.]

 

 

Table B-12. Civilian unemployment rate, 1970-2014

 

[Percent¹; monthly data seasonally adjusted,

 

except as noted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 380.]

 

 

Table B-13. Unemployment by duration and reason, 1970-2014

 

[Thousands of persons, except as noted;

 

monthly data seasonally adjusted¹ ]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 381.]

 

 

Table B-14. Employees on nonagricultural payrolls,

 

by major industry, 1970-2014

 

[Thousands of jobs; monthly data seasonally adjusted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 382.]

 

 

Table B-15. Hours and earnings in

 

private nonagricultural industries, 1970-2014¹

 

[Monthly data seasonally adjusted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 383.]

 

 

Table B-16. Productivity and related data, business

 

and nonfarm business sectors, 1965-2013

 

[Index numbers, 2009=100; quarterly data seasonally adjusted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 385.]

 

 

Interest Rates, Money Stock, and Government Finance

 

 

Table B-17. Bond yields and interest rates, 1942-2014

 

[Percent per annum]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 386.]

 

 

Table B-18. Money stock and debt measures, 1974-2014

 

[Averages of daily Figures, except debt end-of-period basis;

 

billions of dollars, seasonally adjusted]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 388.]

 

 

Table B-19. Federal receipts, outlays, surplus or deficit,

 

and debt, fiscal years, 1947-2015

 

[Billions of dollars; fiscal years]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 389.]

 

 

Table B-20. Federal receipts, outlays, surplus or deficit,

 

and debt, as percent of gross domestic product,

 

fiscal years 1942-2015

 

[Percent; fiscal years]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 390.]

 

 

Table B-21. Federal receipts and outlays, by major category,

 

and surplus or deficit, fiscal years 1947-2015

 

[Billions of dollars; fiscal years]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 391.]

 

 

Table B-22. Federal receipts, outlays, surplus or deficit,

 

and debt, fiscal years 2010-2015

 

[Millions of dollars; fiscal years]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 392.]

 

 

Table B-23. Federal and State and local government

 

current receipts and expenditures, national income and

 

product accounts (NIPA), 1965-2013

 

[Billions of dollars; quarterly data at

 

seasonally adjusted annual rates]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 393.]

 

 

Table B-24. State and local government revenues and

 

expenditures, selected fiscal years, 1954-2011

 

[Millions of dollars]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 394.]

 

 

Table B-25. U.S. Treasury securities outstanding

 

by kind of obligation, 1976-2014

 

[Billions of dollars]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 395.]

 

 

Table B-26. Estimated ownership of

 

U.S. Treasury securities, 2000-2013

 

[Billions of dollars]

 

 

[ Editor's Note: To view the table,

 

see Doc 2014-5547 , page 396.]

 

 

SOURCES

 

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-2. Real gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-3. Quantity and price indexes for gross domestic product, and percent changes

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-4. Percent changes in real gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-5. Contributions to percent change in real gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-6. Chain-type quantity indexes for gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-7. Chain-type price indexes for gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-8. Gross domestic product by major type of product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-9. Real gross domestic product by major type of product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-10. Gross value added by sector

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-11. Real gross value added by sector

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-12. Gross domestic product by industry, value added, in current dollars and as a percentage of GDP

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP-by-industry, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-13. Real gross domestic product by industry, value added, and percent changes

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP-by-industry, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-14. Gross value added of nonfinancial corporate business

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-15. Gross value added and price, costs, and profits of nonfinancial corporate business

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-16. Personal consumption expenditures

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-17. Real personal consumption expenditures

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-18. Private fixed investment by type

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-19. Real private fixed investment by type

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-20. Government consumption expenditures and gross investment by type

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-21. Real government consumption expenditures and gross investment by type

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-22. Private inventories and domestic final sales by industry

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-23. Real private inventories and domestic final sales by industry

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-24. Foreign transactions in the national income and product accounts

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-25. Real exports and imports of goods and services

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-26. Relation of gross domestic product, gross national product, net national product, and national income

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-27. Relation of national income and personal income

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-28. National income by type of income

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-29. Sources of personal income

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-30. Disposition of personal income

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-31. Total and per capita disposable personal income and personal consumption expenditures, and per capita gross domestic product

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-32. Gross saving and investment

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-33. Median money income and poverty status of families and people, by race

Department of Commerce (Bureau of Census), www.census.gov

Income, www.census.gov/hhes/www/income/data/historical

Poverty, www.census.gov/hhes/www/poverty/data/historical

ask.census.gov/newrequest.php, (301) 763-3243 (income), (301) 763-3213 (poverty)

Department of Commerce (Bureau of Census), www.census.gov

Population Estimates, www.census.gov/popest/data/index.html

ask.census.gov/newrequest.php, (301) 763-2422

B-35. Civilian population and labor force

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-36. Civilian employment and unemployment by sex and age

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-37. Civilian employment by demographic characteristic

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-38. Unemployment by demographic characteristic

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-39. Civilian labor force participation rate and employment/population ratio

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-40. Civilian labor force participation rate by demographic characteristic

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-41. Civilian employment/population ratio by demographic characteristic

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-42. Civilian unemployment rate

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-43. Civilian unemployment rate by demographic characteristic

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-44. Unemployment by duration and reason

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

cpsinfo@bls.gov, (202) 691-6378

B-45. Unemployment insurance programs, selected data

Department of Labor (Employment and Training Administration), www.doleta.gov

Weekly Claims and Program Statistics, www.ows.doleta.gov/unemploy/

ui-reports@uis.doleta.gov, (202) 693-3029

B-46. Employees on nonagricultural payrolls, by major industry

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Current Employment Statistics (National), www.bls.gov/ces cesinfo@bls.gov,

(202) 691-6555

B-47. Hours and earnings in private nonagricultural industries

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Current Employment Statistics (National), www.bls.gov/ces

cesinfo@bls.gov, (202) 691-6555

B-48. Employment cost index, private industry

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Employment Cost Index, www.bls.gov/ect

ncsinfo@bls.gov, (202) 691-6199

B-49. Productivity and related data, business and nonfarm business sectors

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Productivity and Costs, www.bls.gov/lpc

dprweb@bls.gov, (202) 691-5606

B-50. Changes in productivity and related data, business and nonfarm business sectors

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Productivity and Costs, www.bls.gov/lpc

dprweb@bls.gov, (202) 691-5606

Board of Governors of the Federal Reserve System, www.federalreserve.gov

G.17 -- Industrial Production and Capacity Utilization, www.federalreserve.gov/releases/g17

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-52. Industrial production indexes, market groupings

Board of Governors of the Federal Reserve System, www.federalreserve.gov

G.17 -- Industrial Production and Capacity Utilization, www.federalreserve.gov/releases/g17

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-53. Industrial production indexes, selected manufacturing industries

Board of Governors of the Federal Reserve System, www.federalreserve.gov

G.17 -- Industrial Production and Capacity Utilization, www.federalreserve.gov/releases/g17

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-54. Capacity utilization rates

Board of Governors of the Federal Reserve System, www.federalreserve.gov

G.17 -- Industrial Production and Capacity Utilization, www.federalreserve.gov/releases/g17

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-55. New construction activity

Department of Commerce (Bureau of Census), www.census.gov

Construction Spending, www.census.gov/constructionspending

ask.census.gov/newrequest.php, (301) 763-1605

B-56. New private housing units started, authorized, and completed and houses sold

Department of Commerce (Bureau of Census), www.census.gov

New Residential Construction, www.census.gov/construction/nrc

New Residential Sales, www.census.gov/construction/nrs

mcd.rcb.customer.service@census.gov, (301) 763-5160

B-57. Manufacturing and trade sales and inventories

Department of Commerce (Bureau of Census), www.census.gov

Manufacturing and Trade Inventories and Sales, www.census.gov/mtis

retail.trade@census.gov, (301) 763-2713

Retail Indicators, www.census.gov/retail

retail.trade@census.gov (monthly), (301) 763-2713

sssd.annual.retail.survey@census.gov (annual)

Manufacturers' Shipments, Inventories, and Orders, www.census.gov/manufacturing/m3

ask.census.gov/newrequest.php, (301) 763-4832

Annual and Monthly Wholesale, www.census.gov/wholesale

sssd.wholesale.trade@census.gov, (301) 763-2703

B-58. Manufacturers' shipments and inventories

Department of Commerce (Bureau of Census), www.census.gov

Manufacturers' Shipments, Inventories, and Orders, www.census.gov/manufacturing/m3

ask.census.gov/newrequest.php, (301) 763-4832

B-59. Manufacturers' new and unfilled orders

Department of Commerce (Bureau of Census), www.census.gov

Manufacturers' Shipments, Inventories, and Orders, www.census.gov/manufacturing/m3

ask.census.gov/newrequest.php, (301) 763-4832

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index, www.bls.gov/cpi

cpi_info@bls.gov, (202) 691-7000

B-61. Consumer price indexes for selected expenditure classes

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index, www.bls.gov/cpi

cpi_info@bls.gov, (202) 691-7000

B-62. Consumer price indexes for commodities, services, and special groups

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index, www.bls.gov/cpi

cpi_info@bls.gov, (202) 691-7000

B-63. Changes in special consumer price indexes

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index, www.bls.gov/cpi

cpi_info@bls.gov, (202) 691-7000

B-64. Changes in consumer price indexes for commodities and services

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index, www.bls.gov/cpi

cpi_info@bls.gov, (202) 691-7000

B-65. Producer price indexes by stage of processing

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Producer Price Indexes, www.bls.gov/ppi

ppi-info@bls.gov, (202) 691-7705

B-66. Producer price indexes by stage of processing, special groups

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Producer Price Indexes, www.bls.gov/ppi

ppi-info@bls.gov, (202) 691-7705

B-67. Producer price indexes for major commodity groups

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Producer Price Indexes, www.bls.gov/ppi

ppi-info@bls.gov, (202) 691-7705

B-68. Changes in producer price indexes for finished goods

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Producer Price Indexes, www.bls.gov/ppi

ppi-info@bls.gov, (202) 691-7705

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.6 -- Money Stock, www.federalreserve.gov/releases/h6

Z.1 -- Financial Accounts of the U.S., www.federalreserve.gov/releases/z1

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-70. Components of money stock measures

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.6 -- Money Stock, www.federalreserve.gov/releases/h6

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-71. Aggregate reserves of depository institutions and the monetary base

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.3 -- Aggregate Reserves/Monetary Base, www.federalreserve.gov/releases/h3

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-72. Bank credit at all commercial banks

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.8 -- Assets and Liabilities, www.federalreserve.gov/releases/h8

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-73. Bond yields and interest rates

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.15 -- Selected Interest Rates, www.federalreserve.gov/releases/h15

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-74. Credit market borrowing

Board of Governors of the Federal Reserve System, www.federalreserve.gov

Z.1 -- Financial Accounts of the U.S., www.federalreserve.gov/releases/z1

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-75. Mortgage debt outstanding by type of property and of financing

Board of Governors of the Federal Reserve System, www.federalreserve.gov

Mortgage Debt Outstanding, www.federalreserve.gov/econresdata/releases/mortoutstand

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-76. Mortgage debt outstanding by holder

Board of Governors of the Federal Reserve System, www.federalreserve.gov

Mortgage Debt Outstanding, www.federalreserve.gov/econresdata/releases/mortoutstand

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-77. Consumer credit outstanding

Board of Governors of the Federal Reserve System, www.federalreserve.gov

G.19 -- Consumer Credit, www.federalreserve.gov/releases/g19

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

Office of Management and Budget, www.whitehouse.gov/omb

The Budget, www.whitehouse.gov/omb/budget/historicals

(202) 395-3080

Department of the Treasury, www.treasury.gov

Monthly Treasury Statement, www.fms.treas.gov/mts

budget.reports@fms.treas.gov, (202) 622-2970

B-79. Federal receipts, outlays, surplus or deficit, and debt, as percent of gross domestic product

Office of Management and Budget, www.whitehouse.gov/omb

The Budget, www.whitehouse.gov/omb/budget/historicals

(202) 395-3080

Department of the Treasury, www.treasury.gov

Monthly Treasury Statement, www.fms.treas.gov/mts

budget.reports@fms.treas.gov, (202) 622-2970

B-80. Federal receipts and outlays, by major category, and surplus or deficit

Office of Management and Budget, www.whitehouse.gov/omb

The Budget, www.whitehouse.gov/omb/budget/historicals

(202) 395-3080

Department of the Treasury, www.treasury.gov

Monthly Treasury Statement, www.fms.treas.gov/mts

budget.reports@fms.treas.gov, (202) 622-2970

B-81. Federal receipts, outlays, surplus or deficit, and debt

Office of Management and Budget, www.whitehouse.gov/omb

The Budget, www.whitehouse.gov/omb/budget/historicals

(202) 395-3080

Department of the Treasury, www.treasury.gov

Monthly Treasury Statement, www.fms.treas.gov/mts

budget.reports@fms.treas.gov, (202) 622-2970

B-82. Federal and State and local government current receipts and expenditures, national income and product accounts (NIPA)

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-83. Federal and State and local government current receipts and expenditures, NIPA, by major type

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-84. Federal Government current receipts and expenditures, NIPA

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-85. State and local government current receipts and expenditures, NIPA

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-86. State and local government revenues and expenditures

Department of Commerce (Bureau of Census), www.census.gov

State and Local Government Finances, www.census.gov/govs/local

govs.finstaff@census.gov, (301) 763-5153

B-87. U.S. Treasury securities outstanding by kind of obligation

Department of the Treasury, www.treasury.gov

Monthly Statement of Public Debt, www.treasurydirect.gov/govt/reports/pd/mspd/mspd.htm

treasury.bulletin@fms.treas.gov, (202) 622-2000

B-88. Maturity distribution and average length of marketable interest-bearing public debt securities held by private investors

Department of the Treasury, www.treasury.gov

Treasury Bulletin, www.fms.treas.gov/bulletin

treasury.bulletin@fms.treas.gov, (202) 622-2000

B-89. Estimated ownership of U.S. Treasury securities

Department of the Treasury, www.treasury.gov

Treasury Bulletin, www.fms.treas.gov/bulletin

treasury.bulletin@fms.treas.gov, (202) 622-2000

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-91. Corporate profits by industry

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-92. Corporate profits of manufacturing industries

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

GDP and Personal Income, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-93. Sales, profits, and stockholders' equity, all manufacturing corporations

Department of Commerce (Bureau of Census), www.census.gov

Quarterly Financial Report, www.census.gov/econ/qfr

csd.qfr@census.gov, (301)763-3359

B-94. Relation of profits after taxes to stockholders' equity and to sales, all manufacturing corporations

Department of Commerce (Bureau of Census), www.census.gov

Quarterly Financial Report, www.census.gov/econ/qfr

csd.qfr@census.gov, (301)763-3359

B-95. Historical stock prices and yields

New York Stock Exchange, www.nyse.com

Dow Jones & Co., www.dowjones.com

Standard & Poor's Financial Services, www.standardandpoors.com

Nasdaq Stock Market, www.nasdaq.com

B-96. Common stock prices and yields

New York Stock Exchange, www.nyse.com

Dow Jones & Co., www.dowjones.com

Standard & Poor's Financial Services, www.standardandpoors.com

Nasdaq Stock Market, www.nasdaq.com

Department of Agriculture (Economic Research Service), www.ers.usda.gov

Farm Income and Wealth, www.ers.usda.gov/data-products/farm-income-and-wealth-statistics

webadmin@ers.usda.gov, (202) 694-5000

B-98. Farm business balance sheet

Department of Agriculture (Economic Research Service), www.ers.usda.gov

Farm Income and Wealth, www.ers.usda.gov/data-products/farm-income-and-wealth-statistics

webadmin@ers.usda.gov, (202) 694-5000

B-99. Farm output and productivity indexes

Department of Agriculture (Economic Research Service), www.ers.usda.gov

Agricultural Productivity, www.ers.usda.gov/data-products/agricultural-productivity-in-the-us

webadmin@ers.usda.gov, (202) 694-5000

B-100. Farm input use, selected inputs

Department of Agriculture (Economic Research Service), www.ers.usda.gov

Agricultural Productivity, www.ers.usda.gov/data-products/agricultural-productivity-in-the-us.aspx

Major Land Uses, www.ers.usda.gov/data-products/major-land-uses

webadmin@ers.usda.gov, (202) 694-5000

B-101. Agricultural price indexes and farm real estate value

Department of Agriculture (National Agricultural Statistics Service), www.nass.usda.gov

Agricultural Prices, usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1002

nass@nass.usda.gov, (800) 727-9540

B-102. U.S. exports and imports of agricultural commodities

Department of Agriculture (Economic Research Service), www.ers.usda.gov

U.S. Agricultural Trade, www.ers.usda.gov/topics/international-markets-trade/us-agricultural-trade

webadmin@ers.usda.gov, (202) 694-5000

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

International Transactions, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-104. U.S. international trade in goods by principal end-use category

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

International Transactions, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-105. U.S. international trade in goods by area

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

International Transactions, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-106. U.S. international trade in goods on balance of payments (BOP) and Census basis, and trade in services on BOP basis

Department of Commerce (Bureau of Economic Analysis and Bureau of Census), www.bea.gov and www.census.gov

Foreign Trade, www.census.gov/ft900

ftd.data.dissemination@census.gov, (301) 763-2311 (goods)

International Transactions, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900 (services)

B-107. International investment position of the United States at year-end

Department of Commerce (Bureau of Economic Analysis), www.bea.gov

International Investment, www.bea.gov/itable

customerservice@bea.gov, (202) 606-9900

B-108. Industrial production and consumer prices, major industrial countries

Board of Governors of the Federal Reserve System, www.federalreserve.gov

Industrial Production (IP), www.federalreserve.gov/releases/g17

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Consumer Price Index (CPI), www.bls.gov/cpi

French National Institute of Statistics and Economic Studies, www.insee.fr/en

IP, www.insee.fr/en/themes/indicateur.asp?id=10

CPI, www.insee.fr/en/themes/indicateur.asp?id=29

portail-statistique-publique-dg@insee.fr

German Federal Statistics Office, www.destatis.de/EN

IP, www.destatis.de/EN/FactsFigures/EconomicSectors/IndustryManufacturing/IndustryManufacturing

CPI, www.destatis.de/EN/FactsFigures/NationalEconomyEnvironment/Prices/ConsumerPriceIndices/ConsumerPriceIndices

www.destatis.de/EN/Service/Contact/Contact

Italian National Institute of Statistics, www.istat.it/en

www.istat.it/en/contact-us

Japanese Ministry of Economy, Trade, and Industry, www.meti.go.jp/english

IP, www.meti.go.jp/english/statistics/tyo/iip

CPI, www.stat.go.jp/english/data/cpi

www.meti.go.jp/honsho/comment_form/comments_send.htm

Statistics Canada, www.statcan.gc.ca

Economic and Financial Data, www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/dsbbcan-eng.htm

infostats@statcan.gc.ca

United Kingdom Office for National Statistics, www.ons.gov.uk

Economy, www.ons.gov.uk/ons/taxonomy/index.html?nscl=economy

info@ons.gsi.gov.uk

B-109. Civilian unemployment rate, and hourly compensation, major industrial countries

Department of Labor (Bureau of Labor Statistics), www.bls.gov

Labor Force Statistics from the Current Population Survey, www.bls.gov/cps

International Labor Comparisons, www.bls.gov/fls/

pressoffice@bls.gov, (202) 691-5902

The Conference Board, www.conference-board.org

International Labor Comparisons, www.conference-board.org/ilcprogram

B-110. Foreign exchange rates

Board of Governors of the Federal Reserve System, www.federalreserve.gov

H.10 -- Foreign Exchange Rates, www.federalreserve.gov/releases/h10

www.federalreserve.gov/apps/contactus/feedback.aspx, (202) 452-3204

B-111. International reserves

International Monetary Fund, www.imf.org

International Financial Statistics, www.imf.org/external/data.htm

International Reserves and Foreign Currency Liquidity, www.imf.org/external/np/sta/ir/irprocessweb/colist.aspx

statisticsquery@imf.org, (202) 623-7764

B-112. Growth rates in real gross domestic product

International Monetary Fund, www.imf.org

World Economic Outlook, www.imf.org/external/ns/cs.aspx?id=29

statisticsquery@imf.org, (202) 623-7764