Menu
Tax Notes logo

New York ‘Donor’ vs. Midwest ‘Moochers’: Beyond the Rhetoric

Posted on Oct. 26, 2020
Robert Tannenwald
Robert Tannenwald

Robert Tannenwald spent nine years as the adjunct lecturer in public policy and budgeting at the Heller School of Brandeis University, from which he retired in June 2020. He spent 28 years as an economist with the Federal Reserve Bank of Boston, from which he retired as vice president in 2010. He was the founding director of the bank’s New England Public Policy Center. He has also served as a senior fellow at the Center on Budget and Policy Priorities and as an analyst in taxation at the Congressional Research Service.

In this installment of Economist’s Roost, Tannenwald analyzes the debate over donor states versus moocher states, arguing that a recent analysis in The Wall Street Journal is flawed and suggesting that other methods of determining states’ COVID-19 relief needs are preferable.

Is New York a donor state or a moocher state? That is, are the federal taxes paid by its residents, businesses, and workers greater than or less than the federal governmental benefits they receive? The issue has been politically significant over the past several months, with Gov. Andrew Cuomo (D) citing State University of New York (SUNY) data bolstering the donor state theory1 to support calls for generous coronavirus relief packages.2

In a 2019 New York Times column, Nobel laureate Paul Krugman showcased the SUNY studies on states’ balance of payments with the federal government, titling one, “The Moochers of Middle America.” Krugman’s point was that many blue states, like New York, have been “lining federal coffers,” while many red states, like those in the Midwest, have been moochers enjoying much larger federal payments and benefits than they have contributed in federal tax payments.3 All the more reason, then, for a generous virus relief package — even if East Coast blue donor states might benefit disproportionately.

Not so, according to U.S. Senate Majority Leader Mitch McConnell (R) of red Kentucky and U.S. Sen. Rick Scott, R-Florida. Jeering at Krugman’s Midwest moochers moniker, McConnell referred to the Democrats’ coronavirus relief package proposed in April as a “blue-state bailout.” He suggested that rather than propping up inefficient states, laws should be changed to make it easier for them to file for bankruptcy.4 In June Scott said that “we shouldn’t bail out states with ridiculous budgets like California, Illinois and New York.”5 In September, McConnell supported a so-called skinny coronavirus relief package with no direct additional aid to states and localities other than school districts.6

Let’s put those political battles aside and focus on SUNY’s analysis. What methods did it use to estimate each state’s “balance of payments”? Which states did it conclude are the biggest donors? What factors drive the results for New York? What criticisms have been leveled against SUNY’s estimates, in particular those made by Matthew Schoenfeld in The Wall Street Journal this summer?7 Are they valid?

SUNY’s Methods and Findings

Studies exploring the donor state issue were first commissioned by then-Sen. Daniel Patrick Moynihan of New York in 1976. He sponsored those studies through his retirement from the Senate in 2000; after his death in 2001, studies continued at Harvard’s Kennedy School of Government through 2009. The tradition continued at the Public Policy Institute of New York City’s Business Council, New York’s Fiscal Policy Institute, and the Office of the New York State Comptroller. Thus, SUNY’s work, analyzing data from 2015 through 2018, continues a long research tradition.

SUNY used two steps. One, it broke down federal receipts and spending into subcategories. Examples of receipts categories include the personal income tax, payroll taxes, and the corporate income tax. Examples of expenditure categories include Social Security, federal contracts (such as defense), and wages of federal employees. Two, it allocated each subcategory of spending by state. Whenever possible, it used official federal data sources to determine the allocation. For example, to allocate federal income tax collections, it used IRS data. When a federal budgetary source was not available, it used what it considered to be the next best available source. For example, to allocate federal alcohol excise tax receipts, it used state-by-state data on alcohol consumption, as reported by the National Institute for Alcohol Abuse and Alcoholism. To allocate housing assistance, it used Center on Budget and Policy Priorities data.

In some cases, SUNY allocated subcategories by state using arbitrary assumptions. For example, it assumed that the federal corporate income tax receipts were allocated by state according to a factor that weighted capital income by 75 percent and wages by 25 percent. SUNY did not report estimates of how alternative weighting might have influenced its results.8

In each year from 2015 through 2018, New York had the fourth biggest “deficit” — that is, in per capita terms it had the fourth smallest difference between outlays received from the federal government and receipts paid to it. In fiscal 2018 that difference was -$1,125. That year, only New Jersey (-$1,293), Massachusetts (-$1,315), and Connecticut (-$2,254) had larger deficits. New York’s deficit was so large in part because of its high federal personal income tax payments (as well as high corporate income and estate and gift tax payments). It also ranked low in federal wages paid by federal military and civilian facilities.9 Those factors were more than sufficient to outweigh relatively large per capita direct federal grants paid to New Yorkers (nationwide, over half of those grants were Medicaid programs).10

Schoenfeld’s Critique of SUNY’s Analysis

Schoenfeld criticized SUNY’s methods on three principal grounds.

Criticism 1

Some federal spending generates goods and services benefiting the nation as a whole. Economists call them public goods and services. A prime example, in Schoenfeld’s view, is defense spending. He notes that in 2018 Texas got $51 billion more than New York ($65 billion to $14 billion) for the privilege of defending every American. Had all defense-related spending been allocated on a per capita basis, according to Schoenfeld, with this one correction New York would have “gotten” $2 billion more per capita than it gave in the aggregate. Consequently, it should not be characterized as a donor state. By contrast, according to Schoenfeld, Supplemental Nutritional Assistance Program (SNAP) benefits “stick where they hit”; that is, their beneficiaries are limited to those who receive them.11

There are several problems with Schoenfeld’s argument. First, defense-related spending is not a purely public good. States like Texas — with large concentrations of that spending — would suffer more poverty and thus receive more means-tested federal “gifts” if defense outlays were more dispersed geographically. Suppose total actual defense spending was allocated between Texas and New York according to population. How would spending on SNAP benefits in each state change? The two spending flows are interrelated. However, according to Schoenfeld’s accounting, they are not. The majority of defense workers earn too much to qualify for SNAP benefits, although tens of thousands of active duty service men and women avail themselves of them every year.12 (Does that mean some of SNAP is devoted to our national defense?)

Second, I dispute Schoenfeld’s characterization of programs such as SNAP benefits as gifts or, in economics lingo, purely “private goods.” Many federally provided means-tested benefits, including SNAP benefits, are designed to respond to the material needs of states facing natural disasters because we are a humane nation.

In October 2017 the percentage of Texas’s population receiving SNAP benefits skyrocketed from 12 percent to 21 percent because of Hurricane Ike.13 Did more than one in five Texans use SNAP benefits in that year, as opposed to approximately one in seven in New York, because Texas was more poorly managed? Perhaps Texas, as well as other red states along the Gulf of Mexico, should anticipate the damage from hurricanes, inducing or even compelling residents to move away from the Gulf Coast? How does this planning failure, frequently characterized as state governmental mismanagement, differ from alleged mismanagement of the COVID-19 crisis by blue states like New York? Should the federal government have refrained from giving aid to hungry Texans cut off from food supplies by Hurricane Ike in the fall of 2017 because the state allegedly should have planned better? Or are hurricanes and pandemics hard to anticipate and plan for?

Third, is hunger only the fault of the hungry or the consequence of natural disasters? Or are perhaps societal factors involved — for example racism, the erosion of protections for labor, intensifying greed, and concomitant inequality?

Criticism 2

Schoenfeld asserts that any estimate of New York’s balance of payments with the federal government should consider the disproportionately large benefit that New Yorkers enjoy from the exclusion of interest on general-purpose state and local municipal bonds. He notes that in 2018, New York state issued $34 billion in such debt. Citing a Congressional Budget Office estimate that each dollar of issuance generates 26 cents in “federal subsidization,” he concludes that SUNY’s estimated federal subsidization of New York long-term debt is $9 billion too low.

The 26-cent figure to which Schoenfeld refers is an estimate of the present value of the cost to the federal government over a 20-year period of a dollar of bond issuance in 2023. The 20-year cost flows consist of tax revenue the federal government would forego from 2023 through 2042 as a result of the exclusion of the interest earned by the bondholders.14 However, for each state, SUNY estimates the payments to and benefits received from the federal government on an annual — not present-value — basis, because only annual figures are available for almost all federal budgetary items.

The exclusion of interest on municipal bonds reduces the federal income tax payments of all owners of those bonds wherever they may be located. If purchasers of New York munis are filing federal returns from Florida, the exclusion reduces the receipts paid by that state into the federal treasury, worsening that state’s balance of payments. If purchasers of New York munis are filing federal returns from New York, New York’s balance of payments is adversely affected. If munis are owned by investment funds, the balance of payments of many states are affected.

If the exclusion of interest on munis were an efficient subsidy, substantially lowering the borrowing costs of states issuing them, Schoenfeld might have a point. However, the most credible analysts conclude that most of the exclusion simply lowers the bondholder’s federal tax liability, having little impact on the those borrowing costs.15 However, suppose that the exclusion does subsidize the borrowing costs of issuing state and local governments, disproportionately benefiting states with high per capita debt issuance. What if New York issues a lot of municipal debt because its state and local public investment needs are high through no fault of its own? The New York City Metropolitan Statistical Area is the nation’s largest in terms of population by a significant margin.16 New York City has unusually large infrastructure needs given its population density and its multi-island geographic configuration. Moreover, since the city is the nation’s largest financial hub, much of its municipal investment generates extensive benefits to people throughout the nation (just as military installations supposedly do, according to Schoenfeld).

Finally, why single out the exclusion of interest on municipal debt as the only tax expenditure erroneously excluded from SUNY’s analysis? What about the exclusion of allowances and selected other benefits paid to military personnel? Tax benefits for the oil and gas industries? I suspect that attempts to allocate those tax expenditures by state would be met with much controversy because data sources do not exist.

Criticism 3

Schoenfeld paints a “life cycle” view of state-by-state balance of payments. Consider a state like Florida that, thanks to its sun, sand, and surf, attracts senior migrants who spent most of their life up north in, say, a state like New York. In their working years, young and middle-aged adults pay federal income and payroll taxes into the federal treasury, improving the northern state’s balance of payments. In their senior years, living in Florida, they receive federal benefits such as Medicare and Social Security, making the Sunshine State seem like more of a moocher than it really is if a taxpayer’s whole life were taken into account.

While that argument perhaps could have been made in 1980, 1990, or even 2000, available statistics on the migration of the elderly have not supported it since. According to Karen Smith Conway and Jonathan Rork, leading authorities on interstate elderly migration flows, the largest such flows in 1980, 1990, and 2000 were, in fact, from northern states into Florida. Flows from New York to Florida led the pack by far: 74,000 and 60,000 in 1990 and 2000, respectively. But over the entire five-year period of 2006 to 2010, only 44,000 elderly left the Empire State for the Sunshine State. The fourth largest interstate elderly flow was from Florida to Georgia. Other sizeable migration flows (top 30 in volume) in which Floridians left their state to live in another included the destinations of New York, North Carolina, Tennessee, Michigan, Texas, and Ohio. With the shrinkage of inflows and expansion of outflows, Florida’s net migration of the elderly shrunk from 11 percent of its elderly population in 1990 to 2 percent in 2010.17

The long-run effect of the pandemic on migration of the elderly is uncertain. Moreover, as the percentage of the elderly over 75 grows, more of them might prefer to stay close to their children in case they need care. Given those and other factors, Schoenfeld’s life cycle thesis just doesn’t fly.

Conclusion: How Much Does ‘Donor vs. Moocher’ Matter?

While it has drawn quite a bit of attention, the donor-versus-moocher debate is badly muddled. More importantly, it is irrelevant to the size and interstate distribution of federal COVID-19 relief. The pandemic has directly and indirectly undermined the physical, economic, and psychological health of most of the American people. In every state, it has damaged a panoply of vital institutions: businesses, state and local governments, universities, and healthcare providers, just to name a few.

Going forward, the volume of additional federal aid should reflect the magnitude of Americans’ urgent needs across the board. That aid should be distributed among the states according to need. While those fiscal decisions will inevitably be subjective and political to some degree, conjecture on which states are donors and which are moochers is not relevant. Even less pertinent are baseless judgments concerning the quality of the governance of blue states and their localities.

FOOTNOTES

1 Laura Schultz and Michelle Cummings, “Giving or Getting? New York’s Balance of Payments with the Federal Government,” Rockefeller Institute of Government, State University of New York at Albany (Jan. 2020).

2  See, for example, Cuomo’s COVID-19 Briefing of Apr. 23, 2020.

3 Paul Krugman, “The Moochers of Middle America,” The New York Times, July 2, 2019.

6 Doug Sword, “‘Skinny’ Coronavirus Relief Bill Blocked in Senate,” Roll Call, Sept. 10, 2020.

7 Matthew Schoenfeld, “New York Is No ‘Donor State,’” The Wall Street Journal, July 21, 2020.

8 Schultz and Cummings, supra note 1, at 25-41.

9 Excluding the U.S. Postal Service.

10 Id. at 10, Table 2.

11 Schoenfeld, supra note 7.

13 USDA Food and Nutrition Service, SNAP Data Tables.

14 Congressional Budget Office, “Federal Support for Financing State and Local Transportation and Water Infrastructure,” Publication 5459, at 3 (Oct. 2018).

15  See, for example, CBO and U.S. Joint Committee on Taxation, “Subsidizing Infrastructure Investment with Tax Exempt Bonds” (Oct. 2009).

16 U.S. Census Bureau.

17 Karen Smith Conway and Jonathan C. Rork, “How Has Elderly Migration Changed in the 21st Century? What the Data Can — and Cannot — Tell Us,” 53 Demography 1011-1025 (Aug. 2016).

END FOOTNOTES

Copy RID