Wednesday, November 30, 2022

Final NYU Tax Policy Colloquium for 2022: Ariel Jurow Kleiman's Fiscal Impoverishment by Taxation

Yesterday (November 29) we had our last session of the year. Ariel Jurow Kleiman presented a forthcoming law review article, Impoverishment by Taxation, along with a related (but broader) book proposal.

While I don't discuss the sessions as such here because they're off the record, just as a quick word I have high aspirations for them that can be challenging to fulfill entirely in practice. Basically I want to combine a stimulating class that will deeply engage the students with a cutting-edge academic seminar that will enlighten and inform all of the participants (myself and the author included). I felt that this year went well, for reasons that had as much to do with the students, and with other participants in the public sessions, as with anything that I or even the authors brought to the table. 

The sessions have evolved over the years (and indeed decades), reflecting the thing's evolution as an institution but also I suppose my changing goals. I used to be a lot more combative, but the more senior you are the less appropriate this is. (Plus I have grown more agnostic about a lot of things.) I think of myself as trying now to be late-career Chris Paul rather than any-stage Russell Westbrook. This requires a deep, strong, effective, and multi-talented team if it is to succeed, but luckily that is what we've had.

Doing it solo the last couple of years is the other big change. While I've enjoyed and benefited from working with all of my co-teachers, doing it this way does leave more air in the room for others to breathe (if I may mangle the metaphor a bit), along with more time for reflection as a byproduct of our having a new paper every 2 weeks rather than weekly. It also allows students to take the class without making as big a sacrifice of other course needs.

Anyway, on to the paper. In addition to being itself interesting, it also helped round out the semester a bit. Generally the papers all concerned inequality. We had looked at high-end economic inequality, and at gender and racial inequality, so it was very appropriate to shift the focus this time to low-end economic inequality.

This week's paper discusses fiscal impoverishment (FI), which arises when people are made poor (or poorer) by paying federal and/or state taxes in excess of the antipoverty public benefits that they receive. FI arises, despite the US fiscal system's overall progressivity, for 3 main reasons:

1) Our fiscal system is not very generous.

2) It imposes a number of first-dollar taxes (i.e., without a zero bracket or its effective equivalent), such as payroll taxes and retail sales taxes.

3) Its coverage on the transfer side is spotty and gap-filled, for reasons that appear to combine the deliberate with the inadvertent. Hence, within a heterogeneous population of low-income individuals and households, some inevitably miss out, even on such benefits as there are.

Fiscal Impoverishment (FI) as a Measure

Proposed fiscal measures can serve one or both of two purposes: the purely analytical or descriptive, and the rhetorical purpose of helping to motivate policy change. This reflects the "scientist versus moralist" choice in one's scholarship that I discussed here with regard to Stanley Surrey and Boris Bittker.

Thus, consider tax expenditure analysis. Although the distinction it addresses actually lies between distributional and allocative objectives, rather than between "taxes" and "spending" as such, it partly aims (and serves) to provide what is simply a more accurate accounting of what (and in a sense "how much") the government is doing. But it also served, at least in Stanley Surrey's hands, as effectively a hit list of provisions that should be eliminated from the income tax code even if they were permitted to re-emerge somewhere else.

FI likewise combines science with enlightened advocacy. The science part is well-illustrated by the old joke about the statistician who says: "My head is in a block of ice and my feet are on fire. On average, I'm very comfortable."

It's a joke because we know that the statistician is wrong. The distribution of outside temperatures for various body parts matters, not just the average.

By analogy, suppose that the fiscal system's average effects, considered for all income levels together, are progressive, inequality-reducing, and poverty-reducing. Taking that as given, it's worse if we have the accountant's problem than if we don't. For example, suppose that people whose market income lies right at the poverty line gain on average 5% after considering both taxes and transfers. It would be better if each of them gained exactly 5% than if (a) half of them lost 50% of their market income, while (b) the other half of them gained 60%. Declining marginal utility alone would tend to support this normative conclusion, and there may be other grounds, too.

FI is not, however, aimed at simply measuring dispersion, even at the bottom of the income scale in particular. It adds two further premises, each subject to challenge from the standpoint of pure "science," albeit each defensible from the standpoint of enlightened advocacy. They are as follows:

        1) The poverty line is (or should be treated as) discontinuously important. Say that the fiscal system's net effect is to reduce your market income by $1. FI counts this as a violation if the loss pushes you below (or further below) the poverty line, but not if it pushes you from $1 above the poverty line right to it. However, while this claimed discontinuity is open to challenge from a "scientific" standpoint, it allows the FI measure to have a dramatic and salient takeaway that might help to motivate desirable policy changes.

        2) Market income is normatively meaningful as a baseline. FI addresses, not where one stands relative to the poverty line after considering taxes and transfers, but how that placement compares to where one stood considering only market income. This seemingly invites the usual Murphy-Nagel critique of market income as lacking normative relevance given (for example) markets' dependence on government's existence and role. Plus FI looks separately just at the fiscal system, and combines the effects of what one might consider separate sovereigns given our federal form of government. But these "scientific" objections do not prevent FI from offering useful and salient information about how fiscal instruments that we can conceptualize separately and change as we like are affecting things.

The paper's discussion of FI also raises a host of interesting measurement issues. It relies on the federal poverty line, based not on a conviction that it is canonically correct but rather as a placeholder that one could separately adjust or replace if one liked. I do think that the measure ought to be regionally adjusted for price-level differences, whether or not actual fiscal rules should be so adjusted, given that it does indeed cost more (say) in NYC than Biloxi to pay for food, rent, heating, etcetera. And there are also measurement issues relating, e.g., to Medicaid benefits and to the accrual during one's working years of expected Social Security retirement benefits. Without seeking to cover all that here, the paper offers cogent arguments for including, say, SNAP benefits that help one to afford the needed calories but not, say, the value to a given individual of national defense protection or a nice park across the street.

As an extension, the paper's arguments might also support measuring what I call regulatory impoverishment. Suppose, for example, that a minimum wage increase leads to an increase in low-wage workers' aggregate income, but that this comes (as it might not) at the price of net job loss. Or suppose that, while there is no net job loss, the increase causes some of the jobs to shift from people in low-income households to (say) high school and college students in high-income households. This is an example of information that would be highly relevant to a policy assessment, and that aggregated information misses.


        1) How could fiscal impoverishment be reduced? - The most obvious and straightforward approach would be to make the fiscal system more generous towards those at the bottom. However, stating it that way would miss the main takeaway of FI analysis, which pertains to how the fiscal system treats some low-income individuals and households much less favorably than others, leading to adverse effects on some in the face of heterogeneity. 

The main groups that the paper identifies as being disadvantaged on the transfer side today, and thus facing most of the FI, are as follows: the undocumented, childless households, the unemployed, and those whose use of extended kinship relationships (e.g., with cousins or neighbors supplying unpaid care) prevents them from qualifying under child-focused rules. For each of those groups, there are indeed arguments out there for disfavoring them. These arguments are worth noting, in terms of the underlying substantive debate that FI encourages, even though I happen to disagree with each of them.

The undocumented: Here we obviously face issues of immigration policy, which, even in the absence of white supremacist racism, would require analyzing how broadly (or not) we want to limit the fiscal system's generosity, along with access to US wage levels, to those who arrive or stay without being authorized.

Childless households - While I have thought for a long time that US fiscal rules (e.g., for the EITC) are often insufficiently generous to childless households, there are clearly reasons for providing additional benefits to households that have minor children, whose welfare is therefore at stake.

The unemployed - Here we get into such longstanding debates as those concerning TANF work requirements, the UBI vs. EITC approaches, and so forth. I tend not to favor conditioning benefits on having a job (it's like providing reverse insurance for the winners of the job search lottery), but there are arguments for doing so, pertaining e.g. to claims about both externalities and internalities.

Extended kinship relationships - It seems clearly anomalous, and perhaps "unintended," for various childcare benefits to be denied to people because their use of extended networks to provide childcare (in lieu of just the parents plus paid providers) causes them to fall outside the eligibility rules for particular benefits. But there is an underlying administrative issue of needing to direct benefits to the right people (i.e., those who would be most likely to use them on the kids' behalf). So there may be design challenges to think through here even if current law unnecessarily falls short.

    2) Relationship to support for universal basic income (UBI)

From the standpoint of avoiding FI that results from households' heterogeneity across various benefit design dimensions, the obvious answer would seem to be providing a generally available and unconditional UBI. So if FI troubles you that's a point in favor of UBI, and if you favor UBI that may help to make FI seem a salient and relevant measure.

But even if one supports UBI (which is my own inclination), I wouldn't be overly driven by FI as such in determining what the benefit level should be. That might instead be better driven by the standard policy tradeoffs regarding the distributional and incentive effects that it would have if it were set (and funded) at various alternative levels.

                                            *            *            *            *            *    

This wraps up my blogposts on the NYU Tax Policy Colloquium for fall 2022. But we'll be back, with blogposts from me after each public session, at the same time and place next year.

Wednesday, November 16, 2022

Tax Policy Colloquium for November 15: Goldburn Maynard's Biden's Gambit and Wage Enslavement papers

Yesterday at the colloquium, Goldburn Maynard presented two thematically linked papers: Biden's Gambit: Advancing Racial Equity While Relying on a Race-Neutral Tax Code, and Wage Enslavement: How the Tax System Holds Back Historically Disadvantaged Groups of Americans. (The latter was co-authored by David Gamage.) Here are some of my thoughts about the issues discussed therein:

1) Racial equity

Biden's Gambit (which discusses the American Rescue Plan Act) notes a distinction that has been drawn in the literature between "racial equity" (RE) and "racial equality." As it notes: "While racial equality has come to connote equal treatment and race blindness," RE is about fairness and anti-subordination, and is "firmly race-conscious."

I would describe this distinction as involving, not "equity" versus "equality" as such, but the choice between substantive and purely formal definitions of racial equality.

To illustrate the purely formal view of racial equality, consider Chief Justice Roberts' infamous statement (or at least it should be): "The way to stop discrimination on the basis of race is to stop discriminating on the basis of race."

This silly and disingenuous bromide suggests that there would be nothing to worry about if the elimination of affirmative action in college admissions led to elite institutions' being 100 percent white, even if this result was due to historical  and continuing private sector racism. It purports to perceive no difference between using race-conscious categories to dismantle and to enforce racial subordination.

While using racial categories raises important issues, I can't imagine a coherent normative view that was based on caring only about formal, not substantive, racial equality. Using race-conscious categories to advance substantive racial equality clearly raises a bunch of serious issues (e.g., pertaining to perceived fairness, political support, and potential misuse), but cannot reasonably be ruled out in principle.

2) "Wage enslavement"

The Wage Enslavement paper uses this term, a bit apologetically, because it was picked as a framing device by the organizers of the symposium in which it appears.

The term "wage slavery" has a bit of a mixed history, as it was used pre-Civil War by slavery proponents to argue that Northern workers were no better-off than enslaved persons in the South. But it has come to denote being trapped in a bad job (e.g., one that is not only low-paying but also unpleasant, exploitative, and physically dangerous) because one otherwise can't stay afloat.

Its metaphorical relevance to the paper arises from the concern that (a) people who don't accumulate significant savings must keep working to support themselves, and (b) the US tax system contributes to this lack of savings from one's work by imposing higher tax rates on labor income than on capital income. 

Among the standard tax policy issues raised by this view are (a) the incidence of taxes on capital income, e.g., do they affect wages over time, (b) the income tax's system's discouragement of saving even if capital income tax rates are nominally lower, and (c) whether one should view the accrual of Social Security retirement benefits as reducing the net marginal tax rate on the underlying wages.

I also would argue that a big part of the issue here is not so much the relative tax rates on "wages" and "capital income" as the fact that much of what is economically labor income may be mislabeled by the tax system as capital income. E.g., suppose I am a founder who creates a successful company, and that I "underpay" myself in salary terms because I am profiting from the stock appreciation (but I hold the shares until death). Then my labor income, in economic terms, faces only the 21% corporate rate, and not even that insofar as I can tax-plan effectively at the company level.

The paper mentions the well-known "Buy-Borrow-Die" (BBD) formulation by my good friend Ed McCaffery, but I have long thought that a better formulation would be "Earn-Borrow-Die." After all, the point isn't to buy stocks on margin - thus profiting only to the extent that they earn a rate of return exceeding the interest rate on the loan, and not even creating a tax arbitrage if the interest is caught by the investment interest limitation. This is not necessarily a well-advised plan. Rather, the idea is to do BBD with the proceeds of untaxed "sweat equity" - thus making it, in effect, EBD.

3) The racial wealth gap

As the paper notes, a Brookings analyst, Vanessa Williamson (who also kindly commented on my new book at an October 6 NYU event) estimates the "racial wealth gap" at $10.14 trillion. This is the estimated amount by which Black households' actual share of US national wealth ($2.54 trillion) falls short of what it would be ($12.68 trillion) if held proportionately to population.

There are several reasons, including at least the following, why the racial wealth gap matters:

--Even if it had no concerning empirical consequences as such, it would be of interest as a diagnostic. The gap "shouldn't" be so great (and it's much greater than the related racial income gap), unless other, underlying things are wrong.

--This is all the more true given that various conventional modes of explanation appear to fall short of accounting for it. For example, it persists at a high level even when one adjusts for income or education differences. It appears not to reflect different savings preferences. Some of it may reflect different types of asset holding (although that itself might call for further explanation), but there is also a large unexplained residual that those working in the field of stratification economics have sought to explain.

--Despite wealth's flaws as a measure, it offers more of a lifetime perspective than income, since it includes resources that are neither earned nor consumed in the current year, but that are relevant to one's level of material wellbeing. Put differently, savings matter a lot, as does the intergenerational transmission of material wealth.

--It contributes to substantive racial inequality and racial subordination, which reduce the wellbeing not only of its victims but actually of all Americans, who live our lives in its malign shadow. Consider the evidence (e.g., from Wilkinson & Pickett's The Spirit Level) that economic inequality reduces subjective wellbeing (and increases social gradient ills) for all groups. This is surely true as well for racial inequality. I suspect that even the millions of insecure Trumpian white supremacists who seek solace from subjugating non-whites would be better-off if our country's pervasive racial consciousness and anxiety could be shunted off centerstage by the achievement of secure racial equality.

--It's easy to think of the racial wealth gap in zero-sum terms. E.g., if national wealth were the same but there were no racial wealth gap, then Black households' having $10.14 trillion more in savings would arithmetically require other households' having $10.14 trillion less. But in a positive-sum world one group's having more than previously, because their opportunities have been increased, does not imply others having less.

4) Tax policy takeaways

As I discuss in this article, which recently appeared in this volume, the tax policy literature is still at an early stage of incorporating racial inequality concerns into an analysis that has traditionally been more focused on general economic inequality. But herewith a few quick points:

--Bad rules are even worse than we thought: Work by Dorothy Brown and others suggests that a number of income tax rules that already were widely viewed in the biz as defective, also tend to aggravate racial inequality (even when one adjusts their impact to income). Examples include the home mortgage interest deduction, income tax preferences for retirement saving that have been found to do little to increase retirement saving by those with the greatest need for it, and marriage bonuses for one-earner couples. The new information about these provisions' adverse racial impact is good to have, but it doesn't change bottom-line conclusions insofar as we already knew that these were bad rules.

--But what about what are otherwise good rules?: Things get more complicated if one views a rule as being good tax policy except for its having an adverse effect on substantive racial inequality. Then such questions arise as: How should one evaluate the tradeoffs? Should one now oppose this rule despite its otherwise beneficial net effects? Can one compensate / adjust for its adverse racial effects, thereby continuing to secure its other benefits?

An example might be switching to the income tax to a progressive consumption tax, if one otherwise favors this this switch (as I did at one time) but it were to turn out to increase the racial wealth gap relative to not making the change. A second example might be adopting a VAT without, rather than with, an exemption for "necessities," where the VAT would be used to fund good things. A third might be repealing (or declining to adopt) hypothetical income tax preferences that, while otherwise inefficient, were opposite to the home mortgage interest deduction in that they happened to benefit Black households relative to White ones at the same income level due to differences in asset choice, etc.

A natural way to think about this is in terms of the standard argument that one should think only about efficiency in designing the tax base, since once can separately adjust for progressivity. Thus, suppose again that one is choosing between an income tax and a progressive consumption tax, in light of one's believing that the latter is more efficient but (all else equal) less progressive. In principle, one can get the best of both worlds by adopting the progressive consumption tax but making it more progressive than the income tax to which it is being compared in its marginal rate structure, allowance for "basic income" cash grants, etcetera.

With respect to racial inequality, progressivity adjustments get you at least part of the way there, given the racial income gap. But one could view this as not doing enough, be it due to the larger racial wealth gap or the adverse racial effects that particular rules have even once one has adjusted for income. This in turn might call for (1) addressing wealth inequality more aggressively than one would otherwise consider optimal, given all the other tradeoffs, and also (2) considering the use of race-conscious rules (the tax equivalent of affirmative action) to make up the difference and/or move broadly in the equalizing direction.