"Which Way Forward for a Basic Income in the United States—Expand the EITC or Social Security?"

"Paper presented at the Fifteenth Basic Income Earth Network Congress,
McGill School of Law, Montreal, June 27-29, 2014. Almaz Zelleke, Visiting Scholar,
NYU Robert F. Wagner Graduate School of Public Service, az22@nyu.edu. Rev. 7/2/14.

Note: Not for citation. This is very much a work in progress, and I welcome feedback to
improve the analysis and proposal.

Abstract: Recent attention to basic income in the U.S. prompted by the successful
petition for a basic income referendum in Switzerland suggests the time is ripe for
concrete basic income proposals. 

In this paper, I consider ways forward based on the expansion of existing income support programs—the Earned Income Tax Credit for lowwage workers and Social Security for senior citizens, among others.

The EITC is a household-based negative income tax paid annually, while Social Security benefits are
partially-taxable individual grants paid on a monthly basis. I discuss the pros and cons of
these approaches for a subsistence-level basic income in the U.S.


1. Introduction

Interest in a basic income in the U.S. is higher than at any time since the 1960s, largely due to the media attention around the petition for a basic income referendum in Switzerland. As is often the case, attention is focused on the amount of a potential basic income (with the amazing figure of $2,800 a month having been introduced by the conversion of the suggested Swiss figure of 2,500 Swiss francs a month into U.S. dollars) and its unconditionality. 

Little attention in these discussions has been paid to how a basic income might be implemented. In particular, a basic income is often conflated with a negative income tax (NIT). Are there other forms a basic income might take? Are there important differences? Does a specific goal, such as poverty elimination, or reduction of inequality, lead to a preference for a specific form? If so, are there grounds for preferring a different form to an NIT? In this paper, I explore the ways a basic income might be implemented and consider the pros and cons of each.

2. Four Models for Income Supplementation

When designing income supplements, there are several important questions
related to eligibility, size, and funding of the benefit. In terms of the form the benefit will
take, there are two essential questions: whether the benefits are phased in, and whether
they are the phased out, in relation to other income or assets. This provides four basic
models for designing income supplements (table 1).

Table 1: Four Models of Income Supplementation

SSI=Supplemental Security Income; FAP=Family Assistance Plan; SS=Social Security;
EITC=Earned Income Tax Credit; Alaska PFD=Alaska Permanent Fund Dividend
2a. No phase in, phase out (figure 1) Under the first model, the maximum benefit
is available to those with no other income, but phases out as income rises.

Supplemental Security Income (SSI) in the U.S. is an income supplement in this form with no phase in, but a phase out as earned income rises.

President Nixon’s proposed Family Assistance Plan (FAP) also had this structure. Social Security benefits follow a version of this structure, with no phase in but with phase out of 50-85% of benefits as other income rises. Since this structure guarantees a certain amount of income before benefits phase out, let’s call it the Guaranteed Minimum Income, or GMI model.

In this model, the benefit builds slowly along with earned income, plateaus once it reaches its maximum, and then phases out as earned income continues to rise. The Earned Income Tax Credit (EITC) in the U.S. is an income supplement in the form of an NIT that phases in and phases out. Let’s call this the NIT model.

In this model, the benefit builds slowly along with earned income, plateaus once it reaches its maximum, and then phases out as earned income continues to rise. The Earned Income Tax Credit (EITC) in the U.S. is an income supplement in the form of an NIT that phases in and phases out. Let’s call this the NIT model.

Finally, income supplements can vary according to whether they are scaled to family or household size, as with the EITC and SSI, or whether the benefits are individually paid and the amounts are uniform, as with the PFD. Basic income is generally defined as an individual benefit, but given the family and household basis of the U.S. income tax code, it’s worth considering whether this is an essential feature of basic income implementation, or whether it could be set aside for transition purposes, if not on a permanent basis.

This question is beyond the scope of this paper. Now that I’ve outlined the basic distinctions, let’s take a closer look at the various forms a basic income might take.

While a basic income in the form of a flat grant like the PFD might be the preference of basic income advocates, without a sovereign wealth fund or other large and well-defined asset on hand to fund it, any significant basic income in the U.S. is likely to involve raising or reallocating tax revenue, and will lead to a desire to keep total costs down in order to keep taxes as low as possible.

That will likely push the debate in the direction of a GMI or an NIT—an income supplement with a phase out, as incomes rise, and with or without a phase in. So let’s take a closer look at the GMI and NIT models first.

4. A Guaranteed Minimum Income (GMI) or Negative Income Tax (NIT) GMIs and NITs, as they were first discussed in the U.S., were designed to address the problems created by the structure of social assistance programs at the time.

Social assistance benefits, because they were tied to a standard of need, were often more generous than what could be earned by wages by the lowest waged workers.

And yet, because social assistance was meant to be a substitute for wages, rather than a supplement to them, social assistance beneficiaries faced dollar-for-dollar benefit reductions, or effective 100% marginal tax rates, when they began to earn income.

The problem was how to provide benefits in a way that encouraged, rather than discouraged, 8
employment, and the proposed solution was to phase out benefits at a lower marginal tax
rate. This approach was termed a negative income tax (NIT), which at the time referred to
both variants—what I’m calling the GMI model, with no phase in (figure 1), and what
I’m calling the NIT model (figure 2), with a phase in.

President Nixon proposed a GMI in the 1960s that would have guaranteed a family with children a minimum income of $1,600, about half the poverty threshold for a family of four at the time. The phase out rate would have been about 40%, meaning a reduction of 40¢ for every dollar earned—a significant improvement over 100% tax rates,

but still a much higher marginal tax rate than faced by low-income earners who did not receive the grant. FAP generated considerable opposition and did not pass Congress, but after its defeat Congress did enact the Earned Income Tax Credit (EITC), an NIT for lowincome workers with children, which has been increased and expanded repeatedly in the decades following.

The inherent dilemma of the NIT is that the more generous the benefit and the lower the rate at which the benefit phases out, the higher the cost of the program and the more taxpayers fall under the program. On the other hand, targeting the benefit to only the neediest keeps costs lower, but requires steeper phase-outs, back in the direction of the 100% marginal tax rates. This dilemma is illustrated by the EITC (figure 9), which phases in slowly, increasing as earned income increases, and then phases out slowly, as income surpasses the poverty line adjusted for family size. The irony of its NIT design is that some benefits continue to be paid until income approaches, or even (for families with
3 children) surpasses, $50,000, but the amount of the benefit at that point is trivially small.

This tradeoff is well known to basic income advocates, but it’s worth parsing out
the underlying assumptions of the GMI and NIT approach. Both models are based on a
particular theory of economics, which holds that

• Employment alone provides fair or adequate income to most workers and their
families

• Income subsidies can be limited to a relatively small group of workers for whom the
market mechanism does not work (for reasons of racial discrimination, geographic
location, preference for childrearing over work, etc.)

• The poor need to be incentivized to work for more than a subsistence level benefit

• This cash benefit is superior to other forms of subsidy (such as a higher minimum
wage, a jobs program, etc.) because it is either cheaper or less distorting to the
market.

This theory holds today for the EITC, which is widely heralded as having encouraged
single mothers—a population that is thought to need incentives to go to work rather than
remain on social assistance benefits—to enter the workforce and leave what is left of
welfare in the U.S. SSI, on the other hand, acknowledges the lower likelihood of its
eligible population—seniors and the disabled of any age—to earn enough to live on, but
its meager benefit level (75-80% of the poverty threshold) and high phase out rates of
100% on unearned income, and 50% on earned income (above modest exclusion
amounts) serves as a deterrent to feigning disability, failing to save for retirement, or
shifting from low-wage work to SSI benefits.

5. The Flat Grant Models (RFG and TFG)

In contrast, the Flat Grant models are based on very different theories of
economics. The RFG is based on the theory that

• All members of a political community are entitled to share in the resources of that
community, regardless of their work effort or other income

• Work disincentive effects are either small, shared by all recipients regardless of
their work effort or other income, or irrelevant, given that funding is independent
of employment and the tax base.

The size of the benefit is based on the size of the asset, along with the time horizon over
which the community wishes the asset to hold value and pay a benefit.

The TFG, on the other hand, is based on a theory that is more directly in conflict
with the theory underlying the GMI and NIT models. This theory holds that

• Employment alone does not provide a fair or adequate income to most workers
and their families

• The market mechanism fails many workers for reasons unrelated to their own
efforts or desert

• The poor need no greater incentives to work for more than a subsistence level (or
even higher) benefit than those with higher incomes need to remain employed in
high-paying jobs

• This cash benefit is superior to other forms of subsidy (a higher minimum wage, a
jobs program, etc.) because it maximizes the freedom of the individual recipients
to choose how to live.

None of this comes as news to basic income advocates. However, it does suggest that it
makes strategic sense to explicitly distinguish basic income from the NIT and GMI
models. While it is in theory possible to construct a GMI with a much lower phase out
rate and a much longer “tail” of progressively increasing tax rates and decreasing benefit
amounts, which might appear to make it indistinguishable from a progressive TFG, the
rationales underlying the two models, at least as they have developed in the U.S., suggest
that it would be strategically sound for U.S. basic income advocates to choose a model
that makes its underlying rationale clear.

As Brian Steensland argues in his exhaustive The RFG can also be a way of protecting an asset like clean air, by taxing those who pollute it in order to limit and, at the same time, share the profits of the negative externalities of polluting activities. In the U.S., a carbon tax and dividend seems like the
most likely form of RFG that could be enacted, given the state laws that at least currently
govern the exploitation of other natural resources such as oil and water. Contemporary
Georgists have also suggested the use of land taxes to fund an RFG.

History of the U.S. guaranteed income debates of the 1960s, and Karl Widerquist echoes
in his review of the literature on the guaranteed minimum income experiments of the
1960s and 70s,4 it is easy for the general idea of an income supplement to mean different
things to different people, and then to generate opposition from across the political spectrum.

For this reason, it might be best for basic income advocates to work as hard to
distinguish basic income from the NIT and GMI models as advocates of those programs
worked to distinguish them from social assistance, or the much maligned “welfare.”
If I’m right about this, then part of our strategy has to be continuing to highlight
the structural and systemic factors that lead to poverty and economic insecurity, as the
left has long done.

We now have more evidence than ever about this, thanks in large part to the kind of long-term analysis of income trends undertaken by Thomas Piketty5 and his colleagues. Another part of the strategy has to be confronting the rhetoric about work incentives embedded deep in the NIT and GMI models. Charles Karelis provides a theoretical argument against the way marginal utility theory has been used against the  3 Brian Steensland,

The Failed Welfare Revolution: America’s Struggle over Guaranteed Income Policy (Princeton University Press, 2008). See also Felicia Korhnbluh, The Battle
for Welfare Rights: Politics and Poverty in Modern America (University of Pennsylvania
Press, 2007).

4 Karl Widerquist, “A Failure to Communicate: What (If Anything) Can We Learn From
the Negative Income Tax Experiments,” Journal of Socio-Economics, 34: 49-81.
5 Thomas Piketty, Capital in the Twenty-First Century (Harvard University Press, 2014).13
poor,6 and Sendhil Mullainathan and Eldar Shafir recently published psychological
research to back up Karelis’ intuitions.

On the other hand, after the instinctive resistance to the idea of unconditionality
and the concern about work incentives, the next line of resistance to a basic income is
inevitably the high cost of a universal grant, so the question of financing has to be
addressed as well. Assuming that the TFG model is the way to go, what would a basic
income look like?

6. The Tax-Financed Flat Grant (TFG)

The TFG form of basic income designed to eliminate poverty and maximize work
incentives should look more like Social Security, and less like either the EITC or SSI.
While the Social Security benefit amount varies from recipient to recipient, based partly
on their earnings history (or that of their spouse or parent), once it is set it does not
change according to income or assets, until other income reaches a threshold.8 This
reflects Social Security’s structure as a contributory program whose benefits feel
“earned” by those who receive them, and preserves incentives for covered workers to
save additional funds for retirement and for those seniors who are able to keep working
and earning income.

Once other income reaches $25,000 for single tax filers, or $32,000, 50-85% of Social
Security benefits are taxable at the same progressive tax rates that apply to income for all
taxpayers.

A basic income designed on the TFG model would be a uniform, tax-exempt flat
grant financed not by past contributions but by progressive taxes on all other income,
earned or unearned (figure 10). Current tax rates on the first dollar of earnings are limited
to the flat Social Security and Medicare payroll taxes of 7.65%, but with the income floor
of an untaxed subsistence level basic income, progressive income taxes could begin on
the first dollar of earnings as well; the basic income could replace the personal
exemptions and standard deduction income disregards. This has two advantages. It would
raise funds, but it would also make all workers contributors to the financing of the basic
income as well as recipients of its benefits—the contributory model that is one of the
elements of Social Security’s political durability.


The basic income grant forms an income floor exempt from taxation; all other income is
subject to progressive tax rates.

The advantages of this structure are

• The benefit remains in place as income fluctuates, maintaining the economic
security of the benefit for precarious workers

• It removes the disincentive effects of the high marginal tax rates the GMI and NIT
forms impose on low-income earners

• It restores progressivity to the tax and benefit system, under which marginal tax
rates rise as income increases, rather than the other way around, as is currently the
case in the U.S.

Following the Social Security approach, the TFG could be administratively
separated from the tax system, and integrated only at the point of the annual tax filing,
when the receipt of the grant is reported and used to determine the first dollar of taxable
income. (High earners who know that their taxes due will be greater than their basic
income grants will likely prefer to forgo the monthly payment in favor of claiming it as a
tax exemption on their tax returns.) But there are reasons to prefer a full integration with
the tax code.

One advantage of the tax-integrated approach is that it invites a reform of the tax
code in the direction of simplification. If all citizens receive what is essentially a
refundable personal exemption and standard deduction tied to an adequate minimum,
what is the rationale for the myriad tax exemptions and deductions that complicate the
U.S. tax code and that redistribute benefits up the income ladder? Some of the
exemptions, like the lower rate on capital gains, are clearly designed to privilege the
income streams particular to the politically powerful, but others, including the home
mortgage interest deduction, the child tax credit, and of course, the EITC, are designed—
but poorly so—to subsidize socially-approved investments on the part of low- and
middle-income taxpayers.

These tax exemptions are now estimated to cost the U.S. Treasury $1 trillion in foregone revenues. It is true that in the current political culture in the U.S. any reduction in exemptions is viewed as a tax increase by anti-tax activists like Grover Norquist, author of the “Taxpayer Protection Pledge”9 signed by the majority of Congressional Republicans. But the strategy of linking economic security for all Americans with tax simplification for all taxpayers might have the chance of breaking
through the current political stagnation in Washington by uniting progressive Democrats
on the left and classically liberal Republicans on the right.

7. How High Should the Basic Income Be?

For a basic income that eliminates poverty—which should be the first priority in
the U.S.—the poverty thresholds present themselves as a target amount for a basic
income. While there is widespread agreement that they are outdated and inadequate, it
seems unrealistic that in the current climate a basic income could be set at a higher level,
and evidence from the NIT experiments in the U.S. and Canada in the 1960s and ‘70s,
along with more recent evidence from a study of Native Americans who received
dividends from casino profits, suggests that even less than poverty threshold-level income
supplements can have significant effects on well-being and opportunity

The TFG model suggests a salient amount as a starting point for a basic income in
the U.S. The current value of the personal exemption is $3,900, and the standard
deduction, which applies not to individuals but to tax units, is $12,200 for married
couples filing jointly.

Dividing the standard deduction by 4, to get at a rough per person amount for a typical family, yields a little over $3,000 per person. A $7,000 per person basic income that goes to adults and children would not provide an income sufficient to meet the poverty threshold for a single person living alone, but it would exceed it for couples or joint householders and all families with children. The experience of Social Security suggests that popular universal benefits that start out at a low level will increase
over time.

Of course, the basic income could be targeted to the poverty threshold for
individuals who live alone, or about $12,000 per person, but it seems unlikely that a basic
income at that level would also go to children, which would leave our most vulnerable
families—single-parent families—at disproportionate risk for poverty.

To achieve a higher level basic income probably requires a rationale in addition
the one based on poverty elimination, and a source of funding. Growing inequality in the
U.S. is a problem that can also be alleviated by a basic income. Thomas Piketty provides
compelling evidence that inequality cannot be reduced by the redistribution of income
from earnings only, but requires a tax on capital as well. This could be addressed in one
(or both) of two ways.

The first may appeal to some of the same potential allies on the right for whom income tax simplification is a goal. The second won’t appeal to many on the right, but it’s worth trying anyway.

The first option is a tax on the imputed income of capital. There is widespread
agreement among tax specialists that the U.S. system of taxation of capital gains, even at
a lower rate than earned income, distorts the behavior of owners of capital—particularly
in a political regime like ours in which tax rates are changed fairly frequently.

Taxation of capital gains—the amount earned when assets are sold—but not of the value or
earnings of the capital if they are unrealized—induces owners of capital to hold on to
assets that might be better (in the sense of economic efficiency) sold. (It also taxes
owners on gains that might be illusory—based on inflation, but not real gain.) Owners of
capital are more likely to sell when capital gains taxes are decreasing, and to hold when
they are increasing, encouraging them to lobby for reductions, even if they are for short
windows only.

A way around this problem that encourages the circulation of capital and its allocation to the highest yielding investments, is to tax on an annual basis an imputed earning of, say, 4% or 5% on the value of capital assets.13 The other way is to forget about taxing capital gains and to impose a tax on the value of the capital itself, as Piketty has proposed.

Higher, progressive rates of taxation on capital could fund a higher basic income aimed at reducing economic inequality—something a subsistence-level basic income would have little effect on.

The second option would be to restore meaningful taxation of estates—or better
yet, of inheritances. Ideally, the full amount of revenue from estate or inheritance taxation
should be devoted to the redistribution of wealth through a basic income. Under the
current U.S. tax code, estates valued at under $5.25 million face no estate tax at all.

With the median inheritance for the baby boom generation estimated to be only $64,000,15 it’s
clear that the high threshold for estate taxation benefits a small minority of inheritors. To
have the strongest effect on inequality, taxation of estates should shift to taxation of
inheritances, with lifetime exclusion amounts high enough to exclude most inheritances
from taxation and to encourage bequests that distribute accumulated wealth more
broadly.

8. Conclusion

A TFG financed by progressive taxes on income, capital, and inheritances would
eliminate poverty and, depending on how progressive the tax rates are, have the potential
to stop the trend of rising inequality in the U.S. and even reverse it. If the TFG model
seems like a plausible approach to the implementation of a basic income in the U.S., the
next step would be to estimate the tax rates required to finance a basic income at this
level, but that is beyond the scope of this paper."

Source and Full Article with Graphs
http://www.usbig.net/papers/Zelleke,%20A%20Basic%20Income%20in%20the%20U.S..pdf