Unpacking the Complex Challenges of Tax Expenditures: Transcript

Cara Griffith: Welcome everyone. I'm Cara Griffith, the president and CEO of Tax Analysts. I'm so pleased that you've joined us for what I know will be a very interesting discussion on the need for, and the challenges of reforming tax expenditures. And we have put together a wonderful panel that is ready to tackle the issues. Today's event is the eighth in Tax Analysts’ series of public discussions we call "Taxing Issues." We launched the series as part of our 50th anniversary celebration, and through it we're bringing the tax community together with leading policymakers and experts for bipartisan discussions on the future of tax policy. While I'm increasingly hopeful that we will be able to hold in-person events in the not-too-distant future, with COVID-19 still threatening the health of Americans from coast to coast we will continue to host these discussions in a virtual format, and we welcome your feedback on how we can make them more interactive. We also welcome your suggestions on future webinar topics. You can send any feedback or suggestions to events@taxanalysts.org.

And now on to the topic at hand. Tax expenditures are the deductions, credits, and other preferences that are scattered throughout the tax code. Governments all around the world use them generously to promote public policy objectives like homeownership or savings or investments. But from a budget perspective, tax expenditures are very expensive in terms of lost government revenue. In fact, our federal government forgoes well over $1 trillion in revenue every year because of tax expenditures. Tax expenditure reform has been hotly debated but has remained elusive over the years despite their costs, the fact that they often don't achieve their desired policy goals, and that sometimes those goals can be achieved by other means. Tax expenditures also tend to benefit higher-income taxpayers, largely because those taxpayers are in higher tax brackets. While tax expenditures can't be put into a single bucket and declared bad, there's certainly room for improvement.

So why has reform been so hard to achieve? The idea of broadening the base, which means eliminating tax expenditures or at least scaling them back, and lowering rates has long been viewed as good tax policy. A discussion of tax expenditure reform is likely to gather steam in the near future as federal policymakers seek to raise revenue to pay for lots of additional spending. President Biden’s executive order on distributional fairness of programs and an increased focus on income, wealth, and racial inequalities will further drive that conversation. So to help us sort all of this out, we have an outstanding panel.

First we have Tom Neubig, who was instrumental in conceiving of this panel, and I want to thank him and the Council on Economic Policies, where Tom is a senior associate, for helping to put it together. Tom is also a founding member of TaxSageNetwork.com and a member of Tax Analysts’ board of directors. Previously, he was with the OECD, and before that, EY.

Chye-Ching Huang is the executive director of the Tax Law Center at NYU law. She was previously with the Center on Budget and Policy Priorities, where she focused on federal tax, fiscal, and economic policy proposals. Chye-Ching is a frequent author on tax and economic policy issues.

Gene Steuerle is an Institute fellow and the Richard B. Fisher chair at the Urban Institute. He's also a co-founder of the Urban-Brookings Tax Policy Center, the Urban Institute center on Nonprofits and Philanthropy, and several continuing Urban Institute programs. Last, but certainly not least, Marty Sullivan is chief economist and a contributing editor with Tax Analysts. Previously, he served at both the Treasury Department and the Joint Committee on Taxation. Marty has the admirable ability to explain technical tax topics in a way that makes sense. And I know that our readers and myself greatly appreciate it.

Marty is going to set the stage for us and begin the discussion among the panelists. The slides he will refer to are available for download directly below the video window on your screen. At the outset I noted that we welcome your questions. Thank you to those who emailed your questions in advance. Please also use the chat feature to submit questions during today's event. The panelists will have a discussion of the issues, and then I will pose your questions to them, and we will get to as many of them as time permits. And with that, Marty, I turn it over to you.

Martin A. Sullivan: Cara, thank you very, very much. Before we get to our panel of experts, I just want to get some background out of the way for our listeners, our audience who may not be familiar with tax expenditures. So Laura and Michelle, if we get up the first slide, please. What is a tax expenditure? Well, there's all different definitions, but here's the one from the 1974 Budget Act. They are "revenue provisions attributable to provisions of the Federal laws which allow a special exclusion, exemption, or deduction from gross income or provide a special credit, a preferential rate of tax, or a deferral of liability." Well, what you learned from this is that tax expenditures can take many different forms — credits, deductions, exclusions, et cetera — but you also learn from this that the word "special" is in there, and what exactly does that mean?

And as with a lot of legislation, it's pretty vague, and that's a little bit of what we'll be talking about today: the criteria that go into defining and identifying tax expenditures. Next slide, please. Tax expenditures are big, as Cara was telling us. She said over $1 trillion. She was correct. $1.44 trillion is the total estimate from the U.S. Treasury Department. There's 167 items on the latest tax expenditure budget. Well, how big is $1.44 trillion? Well, that's 7 percent of GDP. It's comparable in size to major chunks of our revenue and our expenditures. So tax expenditures are not a detail . . . an obscure provision . . . obscure idea. It's very central to what goes on in tax policy and overall government policy. Next slide, please. This slide's a little hard to read, but let me just give you the executive summary. This is the — these are regular expenditures in the blue and tax expenditures in the orange, and we're looking at different categories of a policy.

And you can see, for example, in the third and fourth column, starting from the left in the area of Commerce and Transportation, for example, that tax expenditures are the major way that we spend in these areas of policy. So it's very important, even if you're not a tax geek, to understand what's going on with the tax expenditure budget. Next slide, please. OK. These dollar figures are in billions. So we can see this is the list of the top 10 tax expenditures. No. 1 on the list is the deduction for employer-provided medical insurance, and that number would be even larger if we included the tax benefits of the exemption from the payroll tax. No. 2 on the list are all the retirement provisions, your IRAs, your pension benefits. No. 3, capital gains. No. 4, housing benefits in the code, and so on down the line.

The top 10 provisions count for 78 percent of the total. The remaining 130 or so are the remaining 32 percent. Next slide, please. This is a little chart that was harder to put together than you might imagine, but it is tax expenditures measured as a percentage of GDP. Now, tax expenditures go up when you create more tax expenditures and they go down when you repeal them, but they also go up when you increase tax rates and they go down when you reduce tax rates. And the most dramatic change is all the way on the left-hand side, where the Tax Reform Act of 1986 caused a tremendous decline in tax expenditures because of the reduction in the number of tax expenditures and the reduction of the rate. So tax reform reduces tax expenditures. And the next slide, please. This is something we're going to talk a lot about today.

As Cara mentioned, many or most tax expenditures are tilted in favor of higher-income taxpayers, and this is just a chart from the Tax Policy Center, and over 50-something percent of the benefits go to the top fifth. So we're going to talk about how, why this is so and what might be done to remedy it. And finally, last little bit of introduction is the next slide. And this is a picture on your left-hand side of the Ways and Means Committee. Down on the right-hand side are the Senate Finance Committee. These are the tax writing committees in Congress. Now these committees are pretty powerful. They have jurisdiction over tax, trade, the public debt, Medicare, and Social Security. Well, that's enough to keep you busy for a few days, but on top of that, because they have the ability to in effect do spending through the tax code, they can dabble in any area of fiscal policy, or any type of policy, just by engaging in tax expenditures.

So we're going to talk a little bit about the process of how tax expenditures get through Congress. OK, well, that's my little introduction and now I want to get to our panel of experts. Let me say, you really couldn't have found a better panel. All three of these folks have written thoughtfully and extensively on the subject of tax expenditures. So I highly recommend it, but today we're going to give a little preview. So let me just start with some simple questions to get the conversation going. So, let's see here . . . loopholes, shelters, subsidies. That's what people sometimes call the tax expenditure budget. Those sound like some pretty bad things that we don't want in the tax law. If I'm trying to raise revenue like Bowles-Simpson in 2010, well, I might go to the tax expenditure budget. Or if I'm a tax reformer like Dave Camp in 2013, I might go to the tax expenditure budget and just start picking off items to raise revenue. So my question for the group is: What's wrong with that? Are all tax expenditures bad? Should we discriminate between them or just start cutting away? Gene, you want to opine?

C. Eugene Steuerle: Sure, I can jump in here. I think the best way to think about the problem we're going to address broadly today is that you have a direct spending budget, and you have a tax-expending budget, and they operate often in separate compartments, and we're going to discuss a lot of reasons why that creates a problem. However, just as a lot of direct spending items aren't necessarily bad, neither are all tax expenditures bad. So there's an issue of whether they're fair, there's an issue of whether they're efficient, there's an issue of whether they're being administered by the right agency — by the way, that could apply in reverse, you could apply some of the same criteria to direct spending items. So that's my simple way of answering your question, Marty. They're neither good, nor bad, nor indifferent just by the fact that they're there, with one major exception. I think the budget accounting for tax expenditures as a tax cut is almost always wrong because it leads to the misleading impression that any tax expenditure is a reduction in taxes rather than similar to something that is a direct spending item. But beyond that I think the question of efficiency, fairness, effectiveness is open to each tax expenditure.

Sullivan: Chye-Ching, what do you think about the tax expenditure budget? All bad, all good, indifferent? 

Chye-Ching Huang : Oh, you've got everything from one end of the spectrum to another, which is why tax is so fun and the tax expenditure budget. I mean, I can find loopholes and subsidies and shelters and all of the things that deserve the bad connotations that you were talking about. But I can also look to the child tax credit [CTC] or the earned income tax credit, two of the most effective anti-poverty programs that we have across the entire budget. And arguably [they] should become even more effective if recent improvements to the CTC and EITC are made permanent. So, again, you do really have to look at each one separately while also sort of recognizing — as the chart that you showed, really sort of underscored — but at the moment, overall, despite being able to sort of point to the EITC and CTC, overall these things are large, expensive, and skewed to the top of the income distribution.

Sullivan: Yeah. At the top of that chart was the employer-provided medical care, capital gains, and pensions. And I don't need to do a statistical study to know that that's probably heavily tilted in favor of our high-income taxpayers. Tom, do you have any tax expenditures you like more than others or hate more than others?

Thomas Neubig: As Gene said there are some that are I think well designed, but others that really would benefit from increased transparency and evaluation. And that's where the Council on Economic Policies and [inaudible] think tank really appreciates Tax Analysts hosting this webinar. They're trying to increase the amount of information policy voters have around the world with a new global tax expenditure database. And although the United States could be viewed as having a best practice in terms of providing information about tax expenditures, I think we've got a long way to go in terms of what information is available about the tax expenditures. I guess I have written about how tax expenditures, to paraphrase Rodney Dangerfield, don't get any respect, that after 60 years they seem to not be taken in much importance. There was a 2010 piece of legislation by Sander Levin, the [then-]ranking minority member of the Ways and Means Committee, to say the JCT should evaluate all the tax expenditures, but he said that they should publish the information for the smallest tax expenditures first and only later do it for the largest tax expenditures. So that would take, you know, 17 years if they did 10 a year.

Sullivan: Spoken like a true former staff person. I want to get back to the issue of transparency and visibility, but let's talk a little bit about the process of the way tax expenditures get enacted. Let me put it this way to you. Supposing you were a lobbyist or an advocate for a particular industry and you wanted to get a billion-dollar subsidy for that industry — I'm sure it's a worthy cause if you were endorsing it and putting that aside — what do you think would be the pros and cons of trying to get a direct expenditure versus a tax expenditure? Do you want to put on a lobbyist hat?

Steuerle: So I am going to subdivide that issue. Although part of it is to fund the lobbyist rather than the firm. I might like it if it comes up annually because I’d make more money lobbying, but let's suppose I really do represent the industry. What I'd want is the most permanent program possible, and so I'd want a mandatory direct expenditure, something like Social Security or something. I'd want something that grows automatically. Like many tax expenditures, the larger the economy, the more it grows automatically. I don't have to get a new appropriation. And most tax expenditures are equivalent to mandatory direct expenditures in the sense that they're more permanent. So the main thing I'd want is something, if I'm representing the industry, that's permanent and growing.

Sullivan: Okay. So you would want a tax expenditure because if it was not sunsetting, if it was not sunset. . . .

Steuerle: I’d want a tax expenditure more than an appropriation, but I might take a mandatory, direct spending as well.

Sullivan: OK. What about the public perception? I think you mentioned this originally, Gene. How does the public perceive you when you go in and say, "I want a billion-dollar subsidy"? Or more likely a billion-dollar tax cut?

Steuerle: I was not speaking to the politics of getting it through the process. I may be, as you’re clearly indicating, I may be much more likely to be able to sell it as a tax cut than as a direct spending item. In fact, I remember debates — I was not participating in them. My colleagues were in Treasury with [Lawrence] Summers and he was often arguing, "Yes, you're right. You know, probably a direct spending item might be better than a tax cut, but you know what, I can get this subsidy through a tax cut and I can't get it any other ways." And maybe Chye-Ching can jump in with respect to the child credit, [which] is one great example there.

Huang: Well, before I move on there on sort of the business side, in particular, I think one sort of overlooked potential benefit of going the tax expenditure route is that if I'm a business established enough to be spending a lot of money on lobbyists, I'm probably fine to receive something as a fairly complicated deduction or exemption or some other exclusion from profits. I can access the tax lawyers and accountants that are needed to sort of be able to make my way through those rules, and I probably have enough profit and cash flow that I don't need to worry about having something be made refundable or paid out in advance. So especially for the sort of large established businesses, that sort of tax route actually potentially has competitive advantages in the sense that smaller competitors, maybe people that are trying to enter the industry and undercut your existing advantage, may not be able to access the expenditure to the same extent. 

Sullivan: Let me talk a little bit more about process. Digging through the wonderful Tax Notes archives, I found this article from 1984 talking about this fellow from Delaware, a Democrat, Biden his name. And he said, "There is no system for tax expenditures, no equivalent to the authorization process for the appropriation of funds. The Congress is not considered in light of recommendations from the committees with specialized information. They do not consider the programs which are to be financed by tax expenditures, nor does it consider the integration of tax expenditures with direct spending." And so he proposed the Tax Expenditure Control Act of 1984, which would remove tax expenditures from the purview of the tax writing committees and require them to be authorized by the relevant, substantive committees. He would have them sunset every 10 years, and with a review every 10 years. So 37 years later. What do you think, folks? Is it time for something like this? Maybe President Biden might remember what he proposed 37 years ago, or you think that's a step in the right direction?

Neubig: Marty, I think it is a step in the right direction in terms of trying to integrate tax expenditures in the housing area, in the commerce area, in international security, with the direct spending and other regulations, and so having a comprehensive overview of that, as opposed to doing it piecemeal with tax done by Finance and Ways and Means, and housing done by the Housing Committee. So I think there's a real advantage. Whether or not you would want to take it away completely from the tax, I'm sure that that is not going to happen. And so I think what we need is a more comprehensive look at how government is intervening in various sectors, be it direct spending, tax expenditures, or loan programs or regulations.

Sullivan: What I'm getting from the panel so far is we need to . . . we can't just mechanically say all tax expenditures are good or bad. We need to review them individually, but what would help that review is increased visibility. Right now, you know, the tax expenditure budget is in the back of the budget, which is like — if you even get the printed-out copy, it's thicker than the Manhattan telephone book. And then the Joint Committee puts it out and it's pretty obscure as well. And I think while we would not want to say, "oh, everything on here is bad or should all be scrutinized," it is useful information that should be more widely available. And what I'm also hearing is that there should be more review of these to put them on par with direct spending. But let me just ask this question about the — let's say we did have an annual or every five-year review of the tax expenditures. And I think Cara mentioned a little bit of this as well, but what criteria would you want to have them evaluated on? What's a good tax expenditure and what's one that's not so good?

Huang: I'm going to answer a different question and actually sort of go back to the process point while I sort of in the back of my head touch on your other question about how to assess these things. But I think the Biden proposal that you talked about sort of had a couple of different elements. And I would think of one element as potentially useful, and the other element is probably not so useful. I think process reforms, generally, when you sort of think about changing a process around tax expenditures or budgets or various other things, the elements of that that increase transparency and awareness of what's going on tend in my view to be more valuable and more important and more productive over time than changes that are designed to force lawmakers to do something that they might not want to do anyway.

And I think if we look at sort of the history of forcing mechanisms, whether it's 10-year sunsets, or super committees, or fiscal cliffs, or Gramm-Rudman-Hollings, but the track record isn't very good. And the sense of the sorts of policies that come out of them, not necessarily being a fit for purpose. So I would tend in the realm of process reforms to be more attracted to the ones that, as Tom was talking about, create a bit of information, and, maybe as you were talking about, means that more people read the analytical perspectives on the back of the budget and the Joint Committee estimates. The other pieces . . . there is a question of competence about administration of these things where you probably, I do think, still want tax committee involvement.

If you do have the IRS ultimately administering some of these programs, and there are good reasons for the IRS to administer some of these tax programs, these programs as tax provisions, you do want oversight, and you do want drafting that is responsive to the way that that tax administration works. To actually try to answer your other question, I do think that the main issue is that these provisions should be evaluated against the goals that they're trying to achieve, and a lot of these things are trying to achieve different things.

Sullivan: I think also, and we'll get into this in a minute, is what is the distributional impact of these provisions? And looking through the Tax Notes archives, I did see several proposals, of course none of them enacted, to require the Joint Committee to publish instant distributional analysis. Now the Joint Committee does produce distribution [analyses] and Chye-Ching, you know where they are, they're in back of the back of the back of the tax expenditure budget. It's actually one of the most useful pieces of information you can find, but we had regular review of the distributional impact. That would be another bit of information.

Steuerle: I don't want to completely drop this jurisdictional issue because I do think that process reform and jurisdictional issues are crucial to reform, just as Chye-Ching raised. And I think there's a history here, long history, which I can try to give a 15 second summary . . . if you look back early in the history of taxes and spending in this country, most spending was public spending in the broad sense that they weren't individually directed transfers or benefits. There are things like public goods. Basically most spending was for war roads, or highways, or items like this that were mutually shared. In that sense the Ways and Means Committee really was about tax-raising to pay for those public goods. But once our system, as gradually happened in the 20th century, it became more and more a system of transfers. Basically the tax writing committees decided, "you know, we want to be on the gravy side and not just the takeaway side of the budget," or what I call the giveaway side of the budget.

And so now you've got these two jurisdictions that now complete the tax writing committees and the other committees in terms of who gets to give money away and make the public feel good about things. So that has all sorts of implications. It goes well beyond tax expenditures. Even some of the things that drive the deficit is nobody wants to be on the takeaway side anymore. And then I still remember a time when I testified — you raised the issue of housing — and I was making the quick point that housing vouchers might be superior to some tax subsidies for low-income housing construction. What can they still —

Sullivan: What committee were you testifying . . . ?

Steuerle: Ways and Means. The representative was Charles Rangel. Maybe he had been chair then. And he leans over the dais and he looks over. He says, "Mr. Steuerle, when we have jurisdiction over housing vouchers, we could have that discussion. But right now let's talk about what the Ways and Means Committee can deal with."

Sullivan: You can't underestimate how the gloves come off when they're fighting for turf. It's really amazing. Just to set up another item I'd like to have us discuss. . . . Let me ask Tom. Tom, can you tell us why, or if the mortgage interest deduction is unfair or an upside-down subsidy? If it is, why is it?

Neubig: Well, I think the goal of the mortgage interest deduction was to try to encourage home ownership, but people who've looked at whether or not it's actually meeting that objective I think question whether it is successful in terms of increasing homeownership. In fact, it may be just encouraging higher home prices. They crowd out first-time homebuyers and [it] also encourages the use of debt, but many of these tax expenditures, like the mortgage interest deduction and other exemptions and exclusions, are valued based upon the taxpayer's marginal tax rate. With a progressive tax system, that means a $1,000 deduction for charity or for an exclusion of unemployment insurance or for the mortgage interest deduction is worth $370 for the top income taxpayers and perhaps zero or only $120 for lower-income taxpayers. And so what we oftentimes refer to is upside-down subsidies, where they are regressive and more of the benefit is going to high-income taxpayers. And also with the changes in the 2017 tax act increasing the standard deduction, I think about one-third of taxpayers were taking the mortgage interest deduction in 2017. That fell into the low teens afterwards. And so is it really achieving its objectives? I think it's even less likely to now than before.

Sullivan: Does anybody have any ideas on how it might be improved?

Steuerle: I've written an article with Ben Harris, who's now assistant secretary for economic policy, on why a first-time homebuyer’s credit might make a huge amount of difference. This of course is another example where you want to think about spending and tax expenditures almost in the same boat. It's not clear that the first-time homebuyer’s credit shouldn't even be in the direct spending budget rather than the tax expenditure budget. And there’s a question of who could even administer it best, depending on whether you're using large companies and other intermediaries involved, but a first-time homebuyer’s credit would really help new homeowners. And we know that the wealth distribution has not only got dramatically worse, it's got dramatically worse for young people who would be more likely to be the [first-time] homebuyers.

On top of that, in terms of thinking about the wealth distribution, which is much worse than the income distribution, the two major items in the tax code, better the major subsidies for most low- and middle-income people, have to do with homes and retirement accounts. That's where most of us actually accrue most of our savings unless we happen to own businesses or directly own a lot of corporate stock. And so there's huge headway possible in getting at the wealth distribution issues if you tackle those two major areas of tax subsidies, homeownership and retirement.

Sullivan: Let me move on to probably our final topic, which has to do with the tax code and race. Chye-Ching, you've written in 2019 with a coauthor, the bulk of tax expenditures likely increase racial disparities and should be overhauled. And Tom, you just wrote a few weeks ago in Tax Notes, "Differences in income and wealth inequality can cause facially race-blind tax rules to have this disparate effect across racial groups." I was wondering if you two could talk a little bit about your work on the tax code and race. Chye-Ching, you want to lead off?

Huang: Sure. And that piece that you mentioned, I can't claim to be coauthoring with a current assistant secretary for tax policy, but Robert Taylor, you want to watch out for him. He might be a future one. Terrific coauthor on that paper. One of the examples that we looked at is the one that we were just talking about, the deduction for home mortgage interest, and this is something that Dorothy Brown and a number of other academics have written about previously, and what they pointed to is that because of a history and a prison of barriers to full economic opportunity for Black people and people of color, you have this income distribution and this wealth distribution where white people are disproportionately and overwhelmingly represented at the top, and people of color are more overrepresented towards the bottom and the middle of the income distribution.

And so that means that income tilt that we were just talking about means that the deduction generally accrues to filers who are higher income and likely disproportionately white. But this research has also pointed out that that's not the end of the story. Where the inequities might lay are wrong, because even among households with similar income, there are barriers that are influenced by racism. That means that Black people and other people of color might be less likely to be able to buy a home, or to access lending, or to purchase in an area where the initial investment grows at the same rate as the white filer who has purchased the home with the benefit of the tax break. Gene also mentioned employer-provided savings vehicles. There's data showing that Black people are less likely to have access to those sorts of vehicles even when you do control for income and types of employers or employment in a certain industry. So that's an additional way in which it seems that there's a lot of suggestive evidence that you have tax expenditures that are doubly inequitable in this way. Although we don't yet know, we don't yet have the hard data that shows that to us because a lot of the way that we do analyses by race really uses income as a proxy for race, so you're missing some of that variation that might be picked up if you were able to do it more directly.

Sullivan: No, it's very interesting. Of course nobody's claiming that the tax code, taxwriters are overtly trying to be a racist in any way, but they may inadvertently be causing racial disparities, and I was fascinated to learn about the marriage penalty and how if you have couples with equal incomes, you’re — how does it work? — you’re more likely to have a marriage penalty than if you have couples with unequal incomes. And it just turns out that in white America, there's more of the traditional male being the primary bread earner and the female being either a lesser earner or a non-earner, while the Black households might have more equal earnings and that is a very pronounced demographic trend and that's a very pronounced tax impact. And so it’s interesting to learn about these things. Tom, you wrote a great article, but what do you want to share with us on that?

Neubig: Well, I think there's been increasing concern about the significant income and wealth disparities in the U.S., and in the last three years, I think there now is much more heightened appreciation about the racial disparities that go even beyond the income and wealth disparities. But I think that is an additional reason why government spending through the tax code — i.e., tax expenditures — really should be reviewed much more systematically, and I think with a greater focus on equity considerations. Oftentimes the government spending is or programs are evaluated based upon efficiency. And oftentimes equity gets a short shrift. And I think we now know that even though the tax code is, you know, facially blind in terms of explicit discrimination, there really are significant differences in the amount of government spending through the tax code that go to your white versus Black versus Hispanic families.

In the case of, you know, there are 13 percent of households are Black, and yet using some aggregate numbers from the JCT and the Census Bureau, I estimate that only 5 percent of the dividend of lower tax rates goes to Black households. And you can look at the other large tax expenditures, and many of them were based upon your wealth and, you know, high incomes. So, really, you would like to have more of a systematic review with more emphasis in terms of equity considerations.

Steuerle: To be able to do that type of analysis, I should mention that [Leonard] Burman and some of my colleagues at the Urban Institute, as well as the Statistics of Income Division, IRS, are busy trying to create synthetic data files that would allow the type of analysis without threatening any privacy concern whatsoever.

Sullivan: Well, when you say threatening privacy concerns, because I know that some people are saying maybe we should collect race information on tax returns because then we would have the data to a very, you know, not directly assess the distribution across race and ethnicity, but I don't think that's ever going to happen where you have a 1040 we're on the top you check off what your racial profile is.

Steuerle: The synthetic data set would help us avoid that problem.

Sullivan: Right. Exactly, exactly.

Neubig: But I do think race and ethnicity is collected for many government programs. So if you're evaluating, you know, Medicare, Medicaid, Social Security, housing programs, that information is collected. And so I think, as part of President Biden's executive orders, they are going to have a data task force that includes the assistant secretary for tax policy to look at whether or not more information should be available to address issues of racial disparities. We have a lot of safeguards in terms of privacy in the income tax code. And so, whether or not, you know, there's privacy violations by putting down your race and ethnicity versus putting down that you're making $10 million of income, I think hopefully everybody's protected in that sense. But I would say that even if we were able to collect that information on income tax returns in 2021, the efforts by Gene’s colleagues at the Tax Policy Center for imputations, and the JCT and Treasury doing imputations for non-filers and for things that aren't currently on the income tax return should be pursued, so that we can start getting some of this better analysis of racial disparities sooner than, you know, 2024, 2025.

Sullivan: Yeah, the Statistics of Income Division of the IRS produces an amazing amount of data and they produce data by ZIP code. And many years ago I wrote an article in Tax Notes comparing some of the poorest ZIP codes with the wealthiest ZIP codes, and I came up with a headline grabbing number of . . . there's a hundred times more per capita benefit housing in the wealthy ZIP code than in the poorer ZIP codes. And that's not too hard to believe because as you point out, Tom, I'm getting $370 for every $1,000 of deduction. And I live in — 

Steuerle: Marty, did you get in trouble because one ZIP code was a building in New York, and that was too much identification?

Sullivan: Hey, they published it. All I did was transmit it. Oh, it was the Helmsley building. Leona? Remember her? At any rate, it's interesting to look at the ZIP code data, but yeah, the very wealthy ZIP codes are very small.

Steuerle: So by the way, a quicker effort, even in synthetic data files, these might not be available to the public, is just to allow a lot more income matching across departments. And that's something people would be complaining about for 30 years. And that's mainly just a lack of a moderate amount of resources in government to allow those matches to take place, as well as in some cases the Congress allowing those matches to be done, even with the fact that any data set has some minute privacy issue involved. So income matching or matching of data sets could go a long way towards dealing with a lot of this as well.

Huang: I think given the possibility to do that relatively quickly with a moderate amount of resources, Gene said, I think that's the direction I would probably hit on less for the privacy concern and more for the chilling concern. I think there are data and information that does suggest that particularly Hispanic filers may be among those that are least likely to claim the EITC and CTC, but those data are a little bit old. I'm concerned that given the four years that we've just seen with the sorts of fear in the communities that have a high immigrant populations in them, that people would be even less willing to file a tax return and claim tax benefits if they thought that information that they disclosed around their race or ethnicity might be used in ways that that could not be used for — let's be clear that IRS is very, very careful about the use of its data — but just given that general climate, I do think we need to sort of be aware that there may be some downsides and asking people for additional information that they may consider sensitive and not want to give government.

Sullivan: Well, it sounds like a lot of interesting research is going to take place very soon. Oh, I do want to mention before we leave the topic, the federal government isn't the only one that does tax expenditure analysis. Tom, you've mentioned other countries doing tax expenditure analyses. And then also the states have been very active in doing tax expenditure analyses. And apparently, just glancing, I'm not — Cara, actually you can chime in here — they're all over the place in terms of how far they've developed these tax expenditure budgets. They're very hard to do. You have to do the estimates. You have to make judgments about what is and is not included. And of course it's a political minefield. So people do not like to be identified in the tax expenditure budget, but I think we all can agree as imperfect as any, even the best tax expenditure budget will be, it conveys a great deal of useful information that should be more widely available and subject to review. And I think every college economics student should, you know . . . you have 160 students in your class, give each one a tax expenditure and have them do an evaluation of it. I think that would be a step in the right direction.

Neubig: Marty, we've principally talked about income tax expenditures, mentioned that in the case of the exclusion of employer-provided health insurance, the total tax expenditure is much, much larger because of the payroll tax exemption. I think it's real important to perhaps expand the tax expenditure analysis to non-income taxes. And that clearly is very important in non-U.S. countries where value-added tax exemptions and payroll tax exemptions are very important. I know the U.S. Treasury has once or twice done a special analysis of a state tax as well as payroll tax expenditures. So I think that is definitely worth pursuing.

Steuerle: Well, you were already involved in getting out that state tax expenditure budget at the very end of the Reagan administration. It soon got dropped a few years later.

Sullivan: And with so much all — for example, in the [Coronavirus Aid, Relief, and Economic Security Act], you had the ERC, the employee retention credit, and that was a credit against the payroll tax. So I think it's a whole other issue, whether we should be eating into the payroll taxes and into the trust funds with tax expenditures, but as that increases that's going to be important. And excise taxes. There've been some real doozy loopholes in there in the alcohol fuels credits and so forth. So, yeah, it really needs to be . . . again, it's easy for us to sit here and go, "Go ahead and do it." It's extremely difficult for a staff that's already overworked. 

Steuerle: Payroll tax expenditure could come up in a big way. It is really hidden for the most part in Social Security reform, which we know is going to have to take place within a decade or so, and maybe the biggest reform that we all face in the next 10 years or so. And the payroll tax expenditures are very important to be considered in that light.

Neubig: And as you're talking, Marty, about the limited resources that the JCT and Treasury have to do all of the work plus potentially analyze tax expenditures, I think this is another reason why we should have more integration between tax expenditures and other government spending. That with 167 tax expenditures, $1.4 trillion annual outlays, there really does need to be more analysis of the objectives. Who are the beneficiaries? Is it successful? The cost-benefit analysis. And instead of doing all this work every year for just a single number on a tax expenditure, I would almost trade off having them put out more information every three or five years, rather than having them spend time just updating 167 just pure budget numbers.

Sullivan: Well, they’re required by statute, I believe, right? But nevertheless that's a good point. We must mention, speaking of good staff work, the Congressional Research Service every two years puts out a wonderful volume reviewing tax expenditures. And they just put one out, I believe, late last year. It's over 1,000 pages long, but it is a nonpartisan pros and cons description of all of the tax expenditures and it's online and freely available, and it's just an incredible reference for anybody who wants to delve a little bit more into this.

Neubig: Now the Biden administration's infrastructure proposal includes repealing a number of fossil fuel tax expenditures, but also to increase and expand a number of other renewable fuel tax expenditures. It would be very helpful to have more analysis of both the ones they are proposing to repeal, as well as those they're planning on expanding and starting new ones.

Sullivan: Well, he'll be proposing cuts to capital gains, number three on our list, and to the 20 percent deduction for qualified business expenses. And then he's going to increase the corporate rate, which would increase the total. So yeah, lot’s going to happen. And the tax —

Huang: Although the increasing the minimum tax rate on offshore profits decreases what's treated as a tax expenditure for the lower rates on offshore profits, which is another big piece of the puzzle. Exactly, exactly. I think, just to grab onto one of the things that you sort of [mentioned] in your chat earlier, that the total value of these tax expenditures goes up and down with the rate. That also means that they have the feature that overall, because they are overall regressive, they are also overall pro-cyclical in a way that seems somewhat relevant now as we come out of a recession where maybe being sort of pro-cyclical may not be the sort of thing that you want to be thinking about. So that's another sort of rationale for looking at reforming tax expenditures.

Sullivan: Absolutely. I think that that gets overlooked. You want to increase tax, you want to increase stimulus to break the recession — Economics 101 — and tax expenditures just work in exactly the opposite direction.

Huang: Just to sort of say something that might be controversial with Gene and Tom, although I do think it is a very good thing, and the ideal thing to sort of go one by one, because we've said there's such a great spread of tax expenditures and their goals and how they're performing against these goals. As a second-best option, I would not at all look askance at something like a limit on the dollar value of tax expenditures at the top or some other across-the-board way of making some of the sort of more skewed ones less regressive and potentially less wasteful at the top, even though ideally you would deal with them one by one.

Steuerle: Although I’ve said I think each one needs to be examined one-by-one, I do think — this gets tricky — but I think we have to separate the issues of efficiency, equity, and administrability. And actually we haven't even got to the administrability, except Chye-Ching you touched on it briefly but then we moved on. But let me get to the first two, the efficiency and the progressivity. I think in terms of efficiency, we need to examine each one directly to see how well it's doing, but also this always gets complicated relative to other alternatives, which could be direct or indirect. But on progressivity, I think the issue is really confusing. And let me give a direct spending analogy because it's a simple way. The education budget, no matter how you design it, is going to be regressive. The main beneficiaries of people that get an education are the people that get the most education; they'll be the people [that] get the most gains from it.

They’ll make the most income. To me, you solve that problem by creating the progressive rate schedule that says when you earn income, you pay back to society what you get. In fact, I'd even apply that to student loans, and in many cases, I don’t think you have to make higher ed progressive. You can have alternative ways of getting the money back. So I think the same thing happens to tax expenditures. We'll take the increase in the standard deduction. That was a progressive movement in a fairly regressive bill, the Tax Cut and Jobs Act of 2017. But yet it made certain tax expenditures look more regressive because now the home mortgage interest deduction, the charitable deduction was more concentrated at the top. But if you said, what would be better, to extend those deductions down the income scale or to increase the standard deduction? Well, increasing the standard deduction was more progressive. Or I can go to those of you who were involved in the tax reform effort in ’84. People come and say, "Well, we really got a problem with regressivity and getting this insufficiency."

I said, well, no, it's not an issue because, when we were designing that, we were going to adjust the rate schedule to hit our ultimate desired progressivity. So any attack on the regressivity of one program got adjusted in the right schedule, and it was offset. So progressivity should, in an ideal sense, be thought of in a more composite way. The bigger the package, the more you can deal with the progressivity issue, by designing the right progressive provisions or the adjusting the rate schedule, and not debate whether you need to have some inefficient program or you can get rid of it or expand it because it's a regressive or progressive.

Sullivan: I just mentioned Bill Gale's words of wisdom, which are: "Every provision of the code doesn't have to be progressive, just the entire code should be progressive." Cara, do we have any questions in the inbox?

Griffith: We have actually quite a lot of questions, although you did a very good job of reading them with your minds and you've answered some as we've gone along. So I'm going to just kind of start from the bottom and we'll work our way up and see how many we can get to. First, Gene, you kind of touched on this just a moment ago with outlining how to evaluate tax expenditures. Do you all have any advice for the Biden administration? The infrastructure plan proposes expanding or introducing nearly a dozen tax expenditures. How would you rate it, and do you have any advice for the administration and how they can be effective in implementing this?

Sullivan: That's for Gene.

Steuerle: Well, I just talked, so I'm thinking I was going to defer to my colleagues here.

Neubig: Well, I think they really should be doing an assessment of what the objectives are, and are they going to be done in a cost-effective way, and are they the best approach to achieving them? And so in the case of some of the renewable fuels credits, there is a push to have, I think it's called the direct payment option, so that some of the smaller companies will get an immediate refundability of those credits, similar to what is now with the EITC and the CTC, in refundable credit to a company that's experiencing cash flow issues is very important in terms of having an effective investment incentive.

Griffith: Tom, this question may actually be for you as well, though anyone is welcome to jump into it. Someone asked, thinking about the comparison of global tax expenditures, say, for example, France subsidizes higher education with tax dollars, America gives only a tax credit. How do you compare those tax expenditures between different countries?

Neubig: Well, you want to try to compare apples with apples. And one of the things in terms of doing international comparisons, the U.S. is oftentimes viewed as a low tax, low spending country. But as you can see with Marty’s slides, where tax expenditures account for 41 percent of total tax revenue? That would boost the U.S. tax-to-GDP ratio from 25 percent up to the OECD average of 33 to 34 percent. So tax expenditures can really mislead people who look at some of the budget numbers.

Sullivan: Yeah, I think our current spending is $3.6 trillion. So if you throw tax expenditures on there and let's just say it's another $1 trillion, well, that's $4.6 trillion. That's a big difference in the way you look at our government.

Neubig: This also goes to how people are characterized as taxpayers. I hear oftentimes that over half the people don't pay any income tax. Well, a lot of that is because of refundable and even nonrefundable tax credits. But if those people earn an additional $100 of income, their refundable credit is going to decrease by 12 percent. So they really are still in the income tax system. They are still income taxpayers. We just happened to be running some of the child allowances and work subsidies through the tax code.

Griffith: I have to say it’s all incredibly fascinating. There's so many good questions here, and we have a minute and 30 seconds left. One more. We'll go a little bit over time if that's OK with you guys. Is dynamic scoring used to price tax expenditures, and do you think the costs are accurate?

Neubig: The quick answer is no. Dynamic scoring is not used in the estimates of the foregone revenue. Tax expenditure estimates are budget estimates, and they are not revenue estimates of proposed legislation. And so I think we are accurate for the purpose, but they should not be treated as revenue estimates or including expected behavioral effects as a result of a repeal or a modification of them.

Sullivan: All estimates are that: estimates. They're not perfectly accurate. And these numbers are even . . . require much more interpretation because they're not like revenue estimates. There is no behavioral response incorporated into them. But they still convey a hell of a lot of excellent information. So we don't want to . . . we do want to critically evaluate the numbers, but we do not want to say, "Oh, throw the whole thing out because it's not accurate." No, that would be wrong.

Steuerle: Very few accounting systems, whether for direct spending or tax expenditures, can take into account behavior except where it's so direct you can sort of easily account for it.

Griffith: Well, I can't thank you guys enough. This was an incredibly interesting conversation today and, judging by the number of questions that we had, I think everyone else thought so as well. And we could probably go on for another hour, but we will conclude here today. And I thank everyone for your time and have a wonderful rest of your day.

Sullivan: Thank you.

Neubig: Thanks.