Tax Title of
Testimony
of Leonard E. Burman, The Urban Institute Tax Policy Center, Georgetown
Public Policy Institute:, Before the House Ways and Means Committee, The
Impact of Tax Reform on Low- and Middle-Income Households
June 9, 2005
109th Congress
The Impact of Tax
Reform on Low- and Middle-Income Households
Testimony submitted
to the House Committee on Ways and Means
June 8, 2005
Leonard E. Burman
The Urban Institute
Tax Policy Center
Georgetown Public Policy Institute
Chairman Thomas, Ranking Member Rangel, and distinguished members of the
Committee. Thank you for inviting me to testify on the principles that
should guide efforts to reform the tax system.
I applaud the committee on taking on this crucially important subject. I
came to
Washington
20 years ago to work for the Treasury Department on what became the Tax
Reform Act of 1986. Although far from perfect, that reform was guided
from the start by the bedrock tax policy principles of fairness,
simplicity, and economic efficiency. Although some parts of the final
bill were simple and some weren't, it clearly made the tax system fairer
and more efficient. I would be delighted if we could repeat the trick
again today, while also making the tax system simpler.
Although I think people exaggerate when they claim that the 1986 Tax
Reform has been fully undone in the intervening two decades, the tax
code is once again in need of reform. It is needlessly complex. It is
riddled with loopholes. It imposes vastly different tax burdens on
people with similar abilities to pay. And it does not raise enough
revenue to finance current government operations, much less the growing
costs of the retirement of the baby boom generation.
In my testimony, I will focus on how the income tax system affects low-
and middle-income taxpayers and the potential effects of tax reform on
those populations. I have six main conclusions:
Ÿ First, despite its flaws and some recent erosion, the income tax is
highly progressive. In other words, low- and middle-income families bear
much smaller proportional tax burdens than those with high incomes. This
mitigates the effects of other regressive taxes, such as federal payroll
and excise taxes and state and local sales taxes.
Ÿ Second, the income tax code is an important source of income support
for low-income households.
Ÿ Third, tax reform could help low- and middle-income households by
reducing their tax burdens further --both by lowering their rates and by
simplifying and consolidating tax benefits to which they are entitled.
Ÿ Fourth, some so-called fundamental tax reform proposals could shift
the tax burden away from those most able to pay to those least able.
Ÿ Fifth, the claimed economic gains from such proposals are speculative
at best, based solely on theoretical models that have little
relationship to economic reality.
Ÿ And, last, systemic tax reform presents the ideal opportunity to
bring our fiscal system back into balance. If it closed loopholes under
the income tax and used the revenues to reduce the budget deficit, such
reform would spur economic growth by making the tax system more neutral,
increasing national savings, and lightening tax burdens on future
generations.
I. Current Situation
The President's executive order establishing the Advisory Panel on Tax
Reform called for revenue-neutral tax reform that would advance these
objectives: "(a) simplify Federal tax laws..., (b) share the
burdens and benefits of the Federal tax structure in an appropriately
progressive manner..., and (c) promote long-run economic growth."
Although I think revenue neutrality is a misplaced priority given our
current fiscal situation, the President's objectives stand on the
bedrock principles of public finance --simplicity, fairness, and
economic efficiency.
Let's first consider the President's all-important desire to share the
burden progressively and look at how the current federal tax code
affects low- and middle-income Americans. Its glaring flaws
notwithstanding, the current income tax does have many strengths. To
start, it is highly progressive. In 2005, the
Tax
Policy
Center
estimates that 87 percent of the individual income tax will be paid by
the highest-income 20 percent of households ranked in terms of cash
income. (Table 1.) Almost 61 percent will be paid by the top 5 percent.
By comparison, the bottom 40 percent of households receives more in
refundable tax credits than they pay in taxes on average. Collectively,
the bottom fifth receives net tax credits worth 5.5 percent of income;
the top 1 percent pays taxes averaging 20.1 percent of income.
Although the estate tax and the corporate income tax are also quite
progressive, federal payroll taxes are regressive, consuming a much
larger share of income for low- and middle-income households than for
those at the top. 1 And here's the rub: since payroll taxes are the second
largest share of revenue after the individual income tax, and much
larger than the other federal taxes, the overall tax system is less
progressive than the income tax. Including state and local taxes --which
rely much more heavily on regressive sales taxes --some analysts
conclude that the overall tax system is not progressive at all. 2
Recent federal tax changes have provided important benefits to
lower-income households. The Economic Growth and Taxpayer Relief Act of
2001 (EGTRRA) increased the child tax credit (CTC) and made it partially
refundable, expanded the earned income tax credit (EITC), increased the
standard deduction for married couples, and created a new 10-percent tax
bracket. 3
Legislation enacted in 2003 and 2004 sped up the effective date for some
of these provisions. Nonetheless, by cutting top individual income tax
rates, phasing out the estate tax, cutting the corporate income tax, and
expanding opportunities for tax-free saving, the 2001-2004 tax cuts on
balance made the tax system less progressive. Measured as a share of
income, the top tenth of one percent of taxpayers --that's one in one
thousand --got tax cuts 18 times as large as the bottom fifth got.
(Table 2.)
Table 2 also shows that households in every income class benefited from
the tax cuts, but that view is misleading. Since none of the tax cuts
were offset by tax increases or spending cuts elsewhere, it is
impossible to say who the winners and losers are. If the resulting
budget deficits lead to cuts in programs mostly benefiting middle- and
lower-income households, then they and their children will be the big
losers. If burgeoning debt starves businesses of capital, tomorrow's
families may bear the brunt. If instead middle-class benefits are
politically too popular to curtail and Congress can't or won't cut
spending, then high-income people may end up worse off than they would
have been without the tax cuts.
The bottom line is that it is impossible to assess the winners and
losers from tax changes that are not revenue neutral: we cannot gauge
the effects of the 2001 to 2004 tax cuts until we see how Congress
ultimately finances them. 4
A. How the income tax affects low- and middle-income households
The tax system is a mixed bag for low- and middle-income households. On
the one hand, it is overly complex. Tax filers must fill out numerous
worksheets and forms to claim tax credits for working, children, child
care, education, and many other activities. On the other hand, these
programs provide significant income support for households that are
struggling to meet essential needs. A better tax system would not make
families jump through so many hoops to get this support, but tax reform
that just swept all of these subsidies away to help broaden the tax base
would eviscerate income support for low- and middle-income households.
1. Refundable tax credits for low-income families
Low-income families rely particularly heavily on the income tax system.
Although they do not benefit from traditional deductions and credits
because most do not owe income tax, they do benefit from refundable tax
credits, which are available even if a tax filer does not owe income
tax.
In fact, the refundable EITC is the largest source of cash assistance
for low-income families --bigger in the aggregate than temporary
assistance for needy families (TANF) or food stamps. EGTRRA also
substantially increased the refundable child tax credit in 2001. In
2005, families could claim a refundable child tax credit up to 15
percent of earnings over $10,800. 5
Both of these credits encourage work and help families with children
meet basic needs. Since the EITC and CTC phase in with earnings, they
encourage labor force participation among lowincome single parents. The
phase-out of the EITC can discourage a spouse from working, but since
most EITC recipients are single heads of household this isn't a major
concern. 6
Research suggests that, on balance, the EITC encourages work among
recipient households. 7
These two refundable tax credits now represent a very large portion of
income for low-income households with children. The typical household
with one eligible child and income between $10,000 and $15,000 receives
tax credits worth $2,523, or 22.9 percent of income, in 2005. (Table 3.)
A household with two children and the same income receives $3,764, or
34.5 percent of income, in refundable child tax credits and EITC. For
the average household with three or more children, the credits are worth
almost $4,000, or 36 percent of income. Families with incomes between
$15,000 and $20,000 receive even larger tax benefits, though they amount
to a smaller share of income. Even at incomes of $25,000 to $30,000, the
EITC and CTC boost income by more than 15 percent for families with two
or more children. 8
A very large percentage of households with children receive these
benefits. Almost 74 percent of one-child households and 83 percent or
more of households with two or more children benefit from the CTC or the
EITC or both. Participation is lower for very low-income households
because more of them do not have earnings, and for higher income
households because more of them have incomes above the phase-out
thresholds for the credits. But, among eligible households,
participation is very high. 9
The great value of these credits also poses a risk for tax reform. Any
tax reform that eliminated or reduced these credits would devastate
low-income households, unless new spending programs were created to
provide cash assistance. In fact, although many tax incentives are
probably less effective than comparable spending programs, the EITC and
CTC have a lot to recommend them. Despite being overly complex, the EITC
is a very efficient way to provide cash support for lowincome
households. 10
Most recipients of the tax credits would be filing returns anyway to get
refunds of withheld income taxes, and much of the information about
income eligibility is already reported on tax returns. The refundable
credits also avoid the stigma associated with traditional welfare
programs. And, despite the complexity, filing a tax return is often
easier for low-income working families than waiting in line at a welfare
office during working hours. 11
2. Tax subsidies for middle-income families
Middle-income families benefit from an ever-growing panoply of social
programs that have been injected into the tax code. Among them are
credits for childcare expenses, credits and deductions for education, a
tax credit for adoption expenses, and itemized deductions for mortgage
interest, charitable contributions, state and local income, sales, and
property taxes, and exclusions from income for such employer-provided
fringe benefits as pensions and health insurance. The nonrefundable tax
credits are often of limited value to lower-middle-income taxpayers
because they have limited tax liability, and the deductions and
exclusions are worth the most to those with the highest incomes. The
value of a deduction is equal to the deduction amount multiplied by the
marginal tax rate for those who itemize deductions. Since higher income
households tend to have more and larger deductions and also the highest
marginal tax rates, they get the largest benefits from deductions and
exclusions.
The consequences of this hodge-podge of targeted tax benefits are
complexity and inequity. Households with similar ability to pay tax can
end up owing much different amounts, depending on how many hoops they
jump through to qualify for credits and deductions. Table 4 shows that
there can be considerable variation in average tax rates for similar
families with comparable incomes. The variation arises from differences
in use of credits and deductions and whether households are eligible for
benefits (for example, based on the age of children). A homeowner in a
high-tax state can pay much less tax than a renter in a low-tax state,
for example. Variations among lower-income families with children can be
enormous, depending on whether they qualify for the EITC and CTC.
Table 5 shows that there is even more variation in effective marginal
tax rates --that is, the amount of additional tax paid on a dollar of
additional income. 12
The negative tax rates for lower-income families and individuals arise
from the phase-in of eligibility for the EITC and CTC. The positive tax
rates arise from the statutory tax brackets, the phase-out of
eligibility for benefits, and the individual alternative minimum tax,
which raises effective marginal tax rates for most taxpayers who must
pay it. 13
A major source of variation arises from the notion that every tax
incentive must be progressive: the EITC, CTC, education tax incentives,
and many other provisions phase out at certain income levels. A major
reason why ever more taxpayers must pay the AMT is the phase-out of the
exemption allowed to calculate taxable income for AMT purposes, which
raises effective marginal tax rates by 25 percent. Although phase-outs
reduce the revenue losses from each provision, they also add complexity
and make it hard for some families to know in advance whether they will
be eligible for a subsidy and. if so, how much. As noted, these
phase-outs create hidden tax surcharges that are tantamount to higher
statutory tax rates.
To return to the AMT for a moment, a special problem is that it will
affect more and more middle-income households in coming years. By 2010,
almost all married taxpayers with incomes between $75,000 and $100,000
and with two or more children will be subject to this pointlessly
complicated tax. Its effect, like that of the phase-outs, is to raise
marginal tax rates on most families subject to the tax.
The best thing that tax reform could do for low- and middle-income
families would be to consolidate income-support programs and simplify
eligibility. To the extent possible, the hidden taxes created by
phase-outs and the AMT should be replaced by explicit adjustments to the
tax rate schedules or financed by closing loopholes. For example, the
phase-out of the CTC at incomes over $110,000 adds 5 percentage points
to marginal tax rates in that income range. A better and equally
progressive option would be to eliminate the phase-out and raise
statutory tax rates slightly starting at the same income level.
II. Effects of tax reform
The consequence of moving so much economic support into the tax system
is that "tax reform" could lead to a massive cut in income
support for low- and middle-income families. Base broadening is
equivalent to slashing cash transfers.
Base broadening is a good idea, but policymakers would need to adjust
refundable credits and tax rates to hold low- and middle-income
households harmless, on average. Even then, there would be many winners
and losers. 14
Arguably, it might make sense to consolidate cash assistance programs in
the tax code into a couple of refundable credits. For example, a
20-percent work tax credit for the first $10,000 of wages for each
nondependent, non-student, adult worker, and a $1,500 per child fully
refundable child tax credit would provide about the same amount of
assistance to a single mother with two children and $20,000 of earnings
as current law. If eligibility for the work credit was based solely on
work (and not the presence of children) and all children were eligible
for the child tax credit, then administration and compliance would be
vastly simplified. All workers would be eligible for the work credit,
whether or not they had children, and all households with children would
be eligible for the child benefit, regardless of income. 15
To make that work, tax rates would have to be adjusted to raise the same
amount of revenue (effectively "taking back" the credits from
higher income households).
But barring such an offset, even fundamental income tax reform could end
up hurting the most vulnerable members of society.
1. Consumption taxes
Many tax-reform proposals would shift away from a tax based on income to
a tax based on consumption. Such proposals include the value added tax;
the flat tax, which is effectively a subtraction-method VAT in which the
wage portion of the tax is collected from workers rather than firms and
which is somewhat progressive since it exempts some portion of wages; a
national retail sales tax, which is collected entirely at the retail
stage; and a consumed income tax, which is a progressive variant on the
consumption tax.
Although these proposals are often motivated by concerns about
complexity and efficiency, they would almost inevitably shift tax
burdens onto middle- and/or lower-income groups. Consumption is a much
larger share of income for lower- and middle-income households than for
those with high incomes. Data from the Consumer Expenditure Survey
suggest that families earning less than $30,000 (in 2003 dollars) spend
virtually all of their income while those with incomes exceeding
$200,000 spend less than 40 percent. 16
(Table 6.) This pattern is most pronounced for necessities, such as
food, housing, and clothing. Families earning $10,000 to $20,000 spend
three-quarters of their incomes on those items, compared with one-sixth
of income for those earning more than $200,000.
Proposals for consumption taxes often include measures to reduce their
regressivity, such as demogrants --cash transfers to offset the tax due
on a basic level of consumption --for lowincome households, tax
exemptions for some necessities, or even progressive rates. All of those
options raise issues, but most salient is that effective consumption tax
rates for high-income households would have to be very large to be as
progressive as the current tax system. Compared with a comprehensive
income tax, a consumption tax would exclude two-thirds of income from
the tax base for the highest-income households. Thus, a consumption tax
rate would have to be three times as large as an income tax rate to keep
the same tax burden on high-income households. 17
Otherwise, the tax burden would inevitably shift onto at least some
lower- and middle-income households.
Another way to look at a consumption tax is as an income tax with an
unlimited exemption for capital income and no deduction for interest. 18
In other words, the tax base would be wages rather than income. Wages,
like consumption, decline as a share of income as income increases.
(Table 7.) Wages and salaries make up 28 percent of income for
households with incomes over $1 million in 2005, compared with 68
percent for households with incomes between $75,000 and $100,000. Among
households headed by someone under age 65, almost 80 percent of income
is wages for those with incomes between $30,000 and $75,000, compared
with 32 percent for those with incomes over $1 million. Under a wage
tax, more than two-thirds of income of the highestincome households
would be exempt. In other words, they would either face very high tax
rates or end up paying less tax than under an income tax.
In principle, it is possible to design a progressive "consumed
income" tax that would maintain the same distribution as current
law (on average). But the Treasury Department, after examining such
proposals, concluded that they would be much more complex than current
law and basically unworkable. 19
The implication is that a real-world consumption tax would inevitably
shift the tax burden away from those with the highest incomes to those
with more modest incomes. Although some proposals would protect the poor
through a demogrant that would simply squeeze middle-class households
even more.
Despite concerns about equity, a consumption tax might still be
worthwhile if there were huge economic benefits. But there aren't likely
to be. Most of the claimed benefits of switching to a consumption tax
come from base broadening and the large tax imposed on existing capital
during the transition to the new tax. 20
Base broadening --that is, eliminating all credits and deductions --is
probably no more politically feasible under a consumption tax than under
an income tax. In fact, in his executive order establishing to the
Advisory Panel on Federal Tax Reform, the President insisted that
incentives be maintained for homeownership and charitable contributions.
Most likely, these two tax breaks are simply the tip of the iceberg.
As for the transition, switching from an income to a consumption tax
would effectively devalue all existing capital. For example, if the
income tax were replaced with a VAT or a national retail sales tax, the
prices of all taxed goods and services would immediately rise by the
amount of the VAT. The Federal Reserve Board could tighten the money
supply to prevent this price increase, but the resultant increase in
interest rates would reduce the value of existing capital. In either
case, old people would find that their savings could buy much less than
they did the day before the new tax regime was announced. Although such
a lump-sum tax is doubtless efficient --effectively, the government is
raising revenue by confiscating a portion of outstanding wealth --it is
unlikely to be politically feasible.
But if the government provided transition relief (for example, by
continuing to allow companies to take depreciation deductions on old
capital), tax rates would have to be much higher to make up the lost
revenue. Old people would come out ahead, since their capital assets are
worth the same amount as before and all future income from those assets
is tax-free. But most other groups end up worse off because the higher
taxes more than offset gains from a more efficient tax base. 21
In fact, it is not a given that a consumption tax would raise economic
efficiency more than a similarly comprehensive income tax would.
Exempting capital income from tax would eliminate the tax penalty on
saving, but raise the burden on labor. The reason is simple: if the
return to saving is exempted from the tax base, then taxes have to
increase on what is left, which is wages and salaries. If labor supply
is very sensitive to taxation, and saving is not sensitive, then a
consumption tax could harm the economy in the long run (and in the short
run too if there is transition relief). 22
In fact, the economic evidence seems to suggest that both labor supply
and saving are relatively insensitive to taxes, so any efficiency gains
are likely to be modest.
The models discussed so far are largely based on empirical evidence
about responsiveness of savings and labor supply to taxation, but there
is another line of argument that relies almost entirely on theory to
argue that taxing capital would never be optimal. The relatively simple
version of this theory was advanced by Peter Diamond and James Mirrlees.
23
They showed that if there are no restrictions on commodity taxes and if
economic profits either do not exist or can be taxed away, then it would
never be optimal to tax capital or other inputs to the production
process. But, as, Joel Slemrod points out, that the underlying
assumptions behind this oft-cited economic result are extreme. 24
Tax authorities cannot measure economic profits (that is, those profits
over and above the "normal" or required return to capital)
and, even if they could, it would be politically problematic to apply a
100-percent tax to them.
Similarly, there are many constraints on commodity taxes. For starters,
it would be virtually impossible to tax household production (e.g.,
caring for children, cooking, house cleaning, home repairs, gardening,
etc.) --a requirement for production efficiency in the Diamond-Mirrlees
set up. Policymakers might also blanch at the notion of assessing high
taxes on necessities, such as insulin, even though such taxes are highly
efficient since people's demand for life-saving drugs is quite
insensitive to price.
A more recent line of argument has been advanced in separate papers by
Christopher Chamley and Kenneth Judd. 25
Although mathematically elegant, these models rest on even less
realistic assumptions about policy than the Diamond-Mirrlees model. In
these models, individuals live forever and have perfect foresight.
Exempting capital income from tax in the long run is economically
efficient, but only after the government has levied the maximum feasible
tax on capital long enough to endow the government with a huge surplus,
from which it can finance all future government operations without
taxing capital or labor! If people do not live forever or have unlimited
ability to borrow, capital owners might strongly object to that
transition path. And I would bet that there would not be many votes in
Congress for establishing the government endowment fund, much less any
feasible mechanism for preventing government from tapping into principal
to pay for increased cash transfers or more spending. (Consider the
Social Security trust fund as a less ambitious experiment on the
feasibility of financing future operations with government endowments.)
Moreover, these models ignore human capital --that is, investments
people make in themselves to build skills that will pay future returns
through higher wages. Larry Jones, Rodolfo Manuelli, and Peter Rossi
showed that if it is optimal to exempt the returns on physical capital,
then it is also optimal to exempt the returns on human capital. 26
Indeed, the logical extension of the Judd-Chamley models is that wages
should also be exempt from tax. So in this economic utopia, nothing
would be taxed!! But, if Congress cannot build a huge endowment, this
model provides no practical guide to public policy.
Incredibly, a follow-up paper by Judd argued that in general the optimal
tax rate on capital should be negative. In short, not only should
capital not be taxed, but tax incentives for investment are warranted.
To derive that result, the paper resurrected the heroic assumptions of
the Diamond-Mirrlees model. 27
Utopian models aside, there are other concerns about consumption taxes.
If capital is exempt from tax, high-income people will seek out tax
shelters to make wages look like capital (as they already do with
capital gains). Self-employed people and small businesses will have an
incentive to incorporate, pay the owner a low wage, and accumulate large
untaxed profits. Some entrepreneurs already do this to avoid the payroll
tax.
Not every theoretical argument favors consumption taxation. Shinichi
Nishiyama and Kent Smetters argue that a progressive income tax is
equivalent to a kind of insurance that is not available in the
marketplace; it basically smooths after-tax income. 28
As income varies over time, taxpayers pay lower taxes (as a share of
income) in bad years than they do in good. The progressive income tax
could be viewed as a flat-rate income tax bundled with an insurance
policy that pays off when income falls, offsetting part of the income
tax burden. For risk-averse taxpayers, this can be quite valuable.
Perhaps most surprising, in the real world shifting from an income to a
consumption tax would not necessarily increase saving, at least not for
middle-income families. Currently, middleincome families save mainly by
contributing to pensions and 401(k)-type plans. Employees have an
incentive to participate because they avoid income tax on contributions.
Nondiscrimination rules give employers an incentive to induce
lower-income workers to participate. But, under a consumption tax, all
saving is exempt from tax so there is nothing special about pensions.
Without the inducement of a subsidy, many workers would choose to keep
all of their savings in less restrictive accounts. But behavioral
economics (the study of how real people, rather than homo economicus,
behave) suggests that without the restrictions that apply to pension
plans people would be much less likely to contribute without the
inducements offered by employers and tax savings and more likely to
withdraw balances before retirement.
There are also some connections between income taxes and other programs
that help low-income families. Many tax and expenditures programs for
low-income people (e.g., food stamps and EITC) phase out as income
rises. But it does not make sense to phase them out based on consumption
or wages only. Would we want to preserve an income tax only for
low-income families?
Finally, a federal switch to a consumption tax would undermine state
governments' ability to raise revenue. States rely much more on
consumption taxes (mostly retail sales taxes) than the federal
government does. But if the federal government imposed its own retail
sales tax, the combined federal and state rates could be quite high. 29
In consequence, compliance with state sales taxes would fall sharply.
But at the same time, if the federal government is no longer collecting
income taxes, it would be very hard for the states to maintain their own
income tax systems. Further, many of the claimed benefits of simplifying
the consumption tax would be lost if states continued to collect income
tax. As a result, states would likely have to sharply curtail services,
which could further harm low- and middle-income households.
III. Conclusion
Tax reform would be a singular accomplishment if it made the tax system
simpler, fairer, and more conducive to economic growth. Good starting
points would be fixing the income tax to reduce incentives for
inefficient tax sheltering, eliminating or retargeting the individual
alternative minimum tax, consolidating income- support programs for low-
and middle-income taxpayers, and eliminating complicated eligibility
rules and phase-out provisions, adjusting tax rates to raise the desired
level of revenue. 30
But tax reform poses risks for vulnerable populations that have come to
rely on the tax system for substantial income support. Broadening the
base by eliminating refundable tax credits, for example, would devastate
low-income families. Switching the base of the tax system from income to
consumption would shift the tax burden away from those most able to pay
onto those who are less able. Meanwhile, the claimed economic benefits
from such a radical shift reflect questionable unproven assumptions.
Rather than radical tax reform, a surer path to economic growth is to
reduce the deficit, which would increase national savings directly (by
reducing public dissaving). Tax reform would be an ideal opportunity to
address the deficit. Even revenue- neutral tax reform spells tax
increases on many Americans. The losers from tax reform may be more
willing to shoulder the greater burden if they knew that their children
would pay lower taxes and enjoy a healthier economy as a result. And the
best way to reduce the deficit would be to close the loopholes that
allow businesses and high-income individuals to avoid their fair share
of tax.
Table
1. Current-Law Distribution of Federal Taxes By Cash Income Percentiles,
2005 1
_______________________________________________________________________________________________________________________________________________________
Share of Total Average Effective Tax Rate
____________________________________________________________________________________________________________________________________
Cash Income Class Cash Individual Payroll Corporate Estate All Individual Payroll Corporate Estate All
2 Income Income Tax 4 Income Tax Federal Income Tax Income Tax Federal
Tax 3 Tax Tax 5 Tax Tax Tax
___________________ ________________________________________________________________________________________________________________________
Lowest Quintile 2.4 -1.4 2.2 1.1 0.2 0.4 -5.5 7.5 1.2 0.0 3.2
Second Quintile 6.3 -1.9 6.9 2.7 0.4 2.2 -3.0 9.1 1.1 0.0 7.2
Middle Quintile 11.4 3.1 14.6 4.1 1.6 7.8 2.6 10.6 0.9 0.0 14.2
Fourth Quintile 19.7 13.2 25.5 8.9 0.9 17.5 6.5 10.8 1.1 0.0 18.4
Top Quintile 60.5 87.0 50.7 82.3 93.6 72.0 14.0 7.0 3.4 0.4 24.7
All 100.0 100.0 100.0 100.0 100.0 100.0 9.7 8.3 2.5 0.2 20.7
Top 10 Percent 44.9 73.2 30.5 74.4 89.5 56.5 15.9 5.6 4.1 0.5 26.1
Top 5 Percent 33.7 60.8 17.1 67.2 83.8 44.4 17.6 4.2 4.9 0.6 27.3
Top 1 Percent 18.6 38.3 4.6 50.8 58.0 26.6 20.1 2.0 6.8 0.7 29.6
Top 0.5 Percent 14.7 30.8 2.8 44.5 49.1 21.5 20.5 1.6 7.5 0.8 30.4
Top 0.1 Percent 8.5 18.0 1.0 32.0 28.8 13.0 20.7 1.0 9.3 0.8 31.9
_______________________________________________________________________________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version 0305-1), Table T05-0700,
http://www.taxpolicycenter.org/TaxModel/tmdb/Content/Excel/T05-0070.xls.
(2) Tax units with negative cash income are excluded from the lowest quintile but are included in the totals. Includes both filing and non-filing
units. Tax units that are dependents of other taxpayers are excluded from the analysis. For a description of cash income, see
http://www.taxpolicycenter.org/TaxModel/income.cfm
(3) After tax credits (including refundable portion of earned income and child tax credits).
(4) Includes both the employee and employer portion of Social Security and Medicare tax.
(5) Excludes customs duties and excise taxes.
Table
2. Effect of 2001-2004 Tax Cuts on Effective Federal Tax Rates, by
Cash Income Class, 2005 1,2
____________________________________________________________________________________
Effective Federal Tax Rates (in Percent) 3
_______________________________________________________________
Cash Income Class Law Current Law Change
____________________________________________________________________________________
Lowest Quintile 3.5 3.2 -0.2
Second Quintile 9.1 7.2 -1.9
Middle Quintile 16.4 14.2 -2.2
Fourth Quintile 20.3 18.4 -1.9
Top Quintile 27.3 24.7 -2.6
Top 10 Percent 28.8 26.1 -2.7
Top 5 Percent 30.1 27.3 -2.7
Top 1 Percent 32.7 29.6 -3.1
Top 0.5 Percent 33.7 30.4 -3.3
Top 0.1 Percent 35.5 31.9 -3.6
____________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version 0304-3),
Tables T04-0113 and Table T04-0096.
(1) Baseline is pre-EGTRRA law. Includes provisions in EGTRRA, JGTRRA, and WFTRA
affecting the following: marginal tax rates; the 10-percent bracket; the child tax
credit; the child and dependent care credit; the AMT; the standard deduction,
15-percent bracket, and EITC for married couples; tax rates on long-term capital
gains and dividends; pension and IRA provisions; expansion of student loan interest
deduction (excludes other education provisions); and estate tax exemption, rates,
and state death tax credit.
(2) Tax units with negative cash income are excluded from the lowest quintile but
are included in the totals. For a description of cash income, see
http://www.taxpolicycenter.org/TaxModel/income.cfm
(3) Average federal tax (includes individual and corporate income tax, payroll
taxes for Social Security and Medicare, and the estate tax) as a percentage of
Table
3. Participation and Benefits from Earned Income Tax Credit (EITC) and
Child Tax Credit (CTC)
by Number of Children and Cash Income Class, 2005 1
_____________________________________________________________________________________________________________________________________________________________
No Children One Child Two Children Three of More Children
____________________________________________________________________________________________________________________________________
Cash Income Class Tax Units With Tax Units With Tax Units With Tax Units With
Percent Benefit Percent Benefit Percent Benefit Percent Benefit
______________________ ______________________ ______________________ ______________________
dollars) 2 Who Average Percent Who Average Percent Who Average Percent Who Average Percent
Benefit Credit of Income Benefit Credit of Income Benefit Credit of Income Benefit Credit of Income
_____________________________________________________________________________________________________________________________________________________________
Less than 5 14.7 193 7.8 69.1 956 34.2 60.7 1,104 39.8 26.9 1,158 45.2
5-10 14.7 323 5.0 84.6 2,036 30.6 81.9 2,587 37.1 68.6 2,495 37.0
10-15 8.1 116 1.2 88.5 2,523 22.9 93.0 3,764 34.5 84.5 3,975 36.0
15-20 0.7 143 0.9 93.6 2,664 17.6 95.5 4,348 28.6 91.1 4,379 29.0
20-25 0.3 152 0.7 96.0 2,403 12.6 97.4 4,248 21.9 98.4 4,305 22.2
25-30 0.2 100 0.4 96.7 1,815 7.8 98.7 3,800 16.2 97.7 4,228 17.8
30-40 0.1 214 0.7 87.8 1,188 4.1 98.6 2,787 9.5 99.0 3,611 12.1
40-50 0.0 314 0.7 73.9 1,025 2.8 93.0 1,933 5.2 96.6 2,925 7.7
More than 50 0.0 177 0.2 55.0 957 1.5 73.6 1,694 2.4 82.8 2,666 3.6
All 3.4 229 3.4 73.9 1,603 4.8 83.3 2,585 6.0 85.9 3,156 6.4
_____________________________________________________________________________________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version 0305-3A).
(1) Calendar year. Baseline is current law without the EITC and CTC.
(2) Tax units with negative cash income are excluded from the lowest income class but are included in the totals. For a description of cash income, see
http://www.taxpolicycenter.org/TaxModel/income.cfm
(3) Includes both filing and non-filing units. Tax units that are dependents of other taxpayers are excluded from the analysis.
Table
4. Variation in Average Tax Rates by Marital Status and Number of
Children, 2005
__________________________________________________________________________________
Cash Income Class Tax Units Mean Deviation 5th 95th
__________________________________________________________________________________
Married Filing Jointly 2,368,415 -4.03 9.14 -27.93 0.00
Singles and HOH 8,502,452 -3.35 10.82 -27.04 5.90
MFJ 0 Children 1,867,464 0.03 0.41 0.00 0.00
MFJ 1 Child 210,763 -13.79 7.11 -19.99 0.00
MFJ 2+ Children 290,188 -23.02 10.41 -31.63 0.00
Children 6,291,330 2.59 2.59 0.00 6.60
Singles and HOH 1 Child 1,212,148 -15.59 3.97 -19.96 -8.32
Children 998,974 -25.97 4.44 -31.54 -19.47
Married Filing Jointly 1,630,697 3.29 3.25 -2.06 7.93
Singles and HOH 2,120,734 8.55 3.98 1.15 13.20
MFJ 0 Children 851,275 4.91 2.69 0.00 8.06
MFJ 1 Child 336,897 3.61 1.74 0.39 5.74
MFJ 2+ Children 442,525 -0.07 2.49 0.39 5.74
Children 1,566,384 10.03 3.00 4.40 13.30
Singles and HOH 1 Child 375,533 5.87 2.24 2.55 8.98
Children 178,817 1.26 3.35 -6.75 6.11
Married Filing Jointly 825,856 7.92 2.89 3.62 12.29
Singles and HOH 219,668 13.24 4.20 2.40 18.40
MFJ 0 Children 383,282 9.58 2.55 4.97 12.65
MFJ 1 Child 169,704 7.85 2.19 4.76 11.47
MFJ 2+ Children 272,870 5.62 1.98 2.12 8.96
Children 182,038 13.73 4.00 5.75 18.42
Singles and HOH 1 Child 28,181 10.91 4.49 2.40 15.37
Children 9,449 10.69 3.88 1.28 15.73
__________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version
Notes: MFJ refers to married filing joint returns; HOH refers to returns filing
Table
5. Variation in Marginal Tax Rates by Marital Status and Number of
Children,
2005
__________________________________________________________________________________
Cash Income Class Tax Units Mean Deviation 5th 95th
__________________________________________________________________________________
Married Filing Jointly 2,368,415 -0.89 7.21 -14.95 7.61
Singles and HOH 8,502,452 7.72 15.97 -14.93 25.86
MFJ 0 Children 1,867,464 0.33 2.38 0.00 0.00
MFJ 1 Child 210,763 -2.66 12.52 -33.83 15.95
MF J 2+ Children 290,188 -7.43 14.80 -40.00 6.05
Children 6,291,330 9.16 15.33 0.00 17.39
Singles and HOH 1 Child 1,212,148 7.53 16.12 -14.99 25.96
Children 998,974 -1.12 16.81 -40.00 21.02
Married Filing Jointly 1,630,697 15.59 9.31 0.00 24.98
Singles and HOH 2,120,734 21.79 7.31 14.97 30.95
MFJ 0 Children 851,275 16.30 11.52 0.00 27.72
MFJ 1 Child 336,897 15.39 3.35 14.79 22.25
MFJ 2+ Children 442,525 14.38 7.26 14.79 22.25
Children 1,566,384 23.87 6.91 14.97 33.82
Singles and HOH 1 Child 375,533 16.16 4.22 14.97 27.21
Children 178,817 15.45 5.74 6.05 30.72
Married Filing Jointly 825,856 22.24 5.69 14.98 28.19
Singles and HOH 219,668 25.78 6.50 24.97 30.58
MFJ 0 Children 383,282 23.28 5.39 14.99 26.40
MFJ 1 Child 169,704 22.66 5.69 14.99 32.63
MFJ 2+ Children 272,870 20.51 5.70 14.98 27.00
Children 182,038 25.96 4.99 24.98 29.75
Singles and HOH 1 Child 28,181 24.00 12.50 14.98 33.52
Children 9,449 27.61 4.85 14.98 30.95
__________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version
Notes: MFJ refers to married filing joint returns; HOH refers to returns filing
Table
6. Consumption as a Percentage of Income, by Type
________________________________________________________________________________
Income Expenditures Food Housing Clothing Necessities
________________________________________________________________________________
$1 - 10K 254 52 81 10 143
________________________________________________________________________________
Source:
Tax
Policy
Center
calculations based on Consumer Expenditure Surveys
Note: Income is composed of earned and unearned income, and government
All items in 2003 dollars.
Table
7. Wages and Salaries as a Percent of Income, by Cash Income Class, 2005
1
____________________________________________________________________________________________________________________________________________________________
All Tax Units 65 and Over 4 Age Under 65
______________________________________________________________________________________________________________________________
Tax Units 3 Wages Tax Units 3 Wages Tax Units 3 Wages
Cash Income Class and and and
(thousands of 2005 Salaries Salaries Salaries
____________________________ ____________________________ ____________________________
dollars) 2 Number Percent as a Number Percent as a Number Percent as a
(thousands) of Total Percent (thousands) of Percent (thousands) of Percent
____________________________________________________________________________________________________________________________________________________________
Less than 10 19,560 13.5 43.5 4,083 13.8 4.7 15,477 13.5 56.3
10-20 25,611 17.7 48.6 7,774 26.2 5.3 17,837 15.5 67.0
20-30 19,953 13.8 61.5 4,450 15.0 6.7 15,503 13.5 76.9
30-40 15,289 10.6 67.5 2,570 8.7 7.0 12,719 11.1 79.7
40-50 11,738 8.1 67.8 2,043 6.9 11.3 9,696 8.4 79.6
50-75 20,700 14.3 67.1 3,918 13.2 15.5 16,782 14.6 79.1
75-100 11,936 8.3 68.0 1,969 6.6 16.9 9,967 8.7 78.1
100-200 14,432 10.0 66.6 2,014 6.8 17.1 12,418 10.8 74.6
200-500 3,797 2.6 53.0 664 2.2 14.1 3,133 2.7 61.4
500-1,000 642 0.4 40.2 120 0.4 11.6 523 0.5 46.8
More than 1,000 335 0.2 27.9 68 0.2 10.7 267 0.2 32.0
All 144,573 100.0 58.5 29,690 100.0 12.7 114,884 100.0 68.6
____________________________________________________________________________________________________________________________________________________________
Source:
Urban-Brookings
Tax
Policy
Center
Microsimulation Model (version 0305-3a).
(2) Tax units with negative cash income are excluded from the lowest income class but are included in the totals. For a description of cash income, see
http://www.taxpolicycenter.org/TaxModel/income.cfm
(3) Includes both filing and non-filing units. Tax units that are dependents of other taxpayers are excluded from the analysis.
(4) For married couples, at least one spouse is age 65 or over.
1
The progressivity of the estate tax is understated somewhat in the table
because it is distributed in terms of cash income. Some people who are
quite wealthy can have very modest cash incomes --for example, because
most of their income is in the form of unrealized capital gains. If
households are ranked in terms of economic income (including the imputed
income generated by unrealized assets), then 98 percent of the estate
tax falls on the highestincome 5 percent of households.
2
See McIntyre, Bob. 2004. "Overall Tax Rates Have Flattened Sharply
Under Bush: Total Federal, State & Local Rate on Richest Now Only
Slightly Higher than on
Middle
Ranges
," Citizens for Tax Justice, April 12. Available at http://www.ctj.org/pdf/fsl2004.pdf.
3
See Leonard E. Burman, Elaine Maag, and Jeff Rohaly, 2002, "The
Effect of the 2001 Tax Cut on Low- and Middle-Income Families and
Children," available at: http://www.taxpolicycenter.org/publications/template.cfm?PubID=410465.
4
See William G. Gale, Peter Orszag, and Isaac Shapiro, 2004,
"Distribution of the 2001 and 2003 Tax Cuts and Their
Financing," Tax Notes, June 21, pp. 1539-1548.
5
The threshold is indexed for inflation.
6
See Nada Eissa and Hilary W. Hoynes, 2004, "Taxes and the Labor
Market Participation of Married Couples: The Earned Income Tax
Credit," Journal of Public Economics, Vol. 88, pp.
1931-1958.
7
See Nada Eissa and J. Liebman,1996, "Labor Supply Responses to the
Earned Income Tax Credit," Quarterly Journal of Economics,
Vol. 111, pp. 605-637; and B. Meyer and D. Rosenbaum, 2001,
"Welfare, the Earned Income Tax Credit, and the Labor Supply of
Single Mothers," Quarterly Journal of Economics, Vol. 116,
pp. 1063-1114.
8
Very low-income households without children qualify for a small EITC,
but not the CTC. It is worth an average of $229 for recipient
households; only 3 percent of childless households qualify.
9
See Leonard E. Burman and Deborah Kobes, 2003, "EITC Reaches More
Eligible Families Than TANF, Food Stamps," Tax Notes, March
17, p. 1769.
10
Most EITC recipients use paid preparers to file their tax returns. See
Elaine Maag, 2004, "Tax Preparation for Low-Income Households,
Knowledge of the EITC," Tax Notes, August 2, p. 555. This is
in part a function of the complexity of the EITC relative to the
functional capacity of some recipients and partly due to the popularity
of refund anticipation loans offered by some tax return preparers.
11
Janet Holtzblatt and Janet McCubbin, 2004, "Issues Affecting
Low-Income Filers," in Henry J. Aaron and Joel Slemrod, eds., The
Crisis in Tax Administration (Washington, DC: The Brookings
Institution Press): 148-188.
12
For a discussion, see Leonard E. Burman and Mohammed Adeel Saleem, 2004,
"Income Tax Statistics for Sample Taxpayers, 2003," Tax
Notes, January 19, pp. 413-418. The Table shows the marginal tax
rate on earnings. Marginal tax rates on other forms of income would
often be different. Marginal tax rates are calculated by increasing
income by a small amount and calculating the increment in tax
liabilities after credits per dollar of additional income. The marginal
increase in income is the maximum of $100 and the minimum of one percent
of AGI and $1,000. This is done to smooth out some kinks in explicit and
implicit tax rate schedules. The effective marginal tax rates might not
add up exactly because of rounding or because the formulae for them are
not exactly continuous.
13
See Leonard E. Burman, 2005, "The Expanding Reach of the Individual
Alternative Minimum Tax: Testimony submitted to the United States Senate
Subcommittee on Taxation and IRS Oversight of the Committee on
Finance," May 23, available at http://www.urban.org/UploadedPDF/900812_Burman_052305.pdf.
14
That is an inevitable consequence of revenue-neutral tax reform so not
necessarily an impediment.
15
Jonathan Barry Forman, Adam Carasso, and Mohammed Adeel Saleem,
forthcoming, "Designing a Work-Friendly Tax System: Options and
Trade-offs," Tax Policy Center Discussion Paper Number 20.
16
Many researchers have commented on the implausible ratio of consumption
to income for those with very low incomes. Income is probably
underreported, especially for low-income households, which is a special
risk because the focus of the survey is consumption rather than income.
It also excludes gifts from friends and relatives.
17
In fact, our tax system is far from a comprehensive income tax, so
conceivably a shift to a comprehensive consumption tax could be
accomplished with much more modest rates, but that assumes that the
political pressures for exemptions such as for fringe benefits, mortgage
interest, charitable contributions, and so on, could be avoided under a
consumption tax. It is worth noting, in that context, that the President
has insisted that any tax reform retain incentives for homeownership and
charitable contributions.
18
Although this equivalence holds in the long run under certain
circumstances, there are significant differences in the two tax bases in
the short run. A new consumption tax would increase the price of all
consumer goods or reduce the real value of old capital, placing a large
burden on older people. A new wage tax would effectively exempt all
capital income from tax, effectively granting a large windfall on older
people who are living off of their accumulated savings.
19
See Eric Toder, 1995, "Statement of Eric Toder, Deputy Assistant
Secretary (Tax Analysis), Department of the Treasury, Before the Senate
Budget Committee," February 22.
20
See David Altig, Alan J. Auerbach, Laurence J. Kotlikoff, Kent A.
Smetters, and Jan Walliser, 2001, "Simulating Fundamental Tax
Reform in the United States," American Economic Review, Vol.
91, pp. 574-595; and Don Fullerton and Diane Lim Rogers, 1993, Who
Bears the Lifetime Tax Burden? (
Washington
,
DC
: Brookings Institution Press).
21
See Altig, et al., 2001.
22
See William C. Randolph and Diane Lim Rogers, 2001, "The
Implications for Tax Policy of Uncertainty About Labor-Supply and
Savings Responses," National Tax Journal, Vol. 48, pp. 429-446.
23
See P.A. Diamond and J.A. Mirrlees, 1971, "Optimal Taxation and
Public Production I: Production Efficiency," American Economic
Review, Vol. 61, pp. 8-27.
24
See Joel Slemrod, 1990, "Optimal Taxation and Optimal Tax
Systems," Journal of Economic Perspectives, Vol. 4, pp.
157-178.
25
See Christophe Chamley,1986, "Optimal Taxation of Capital Income in
General Equilibrium with Infinite Lives," Econometrica, Vol.
54, No. 3, pp. 607-622; and Kenneth L. Judd,1985, "Redistributive
Taxation in a Simple Perfect Foresight Model," Journal of Public
Economics, Vol. 28, pp. 59-83.
26
See Larry E. Jones, Rodolfo Manuelli, and Peter Rossi, 1997, "On
the Optimal Taxation of Capital Income," Journal of Economic
Theory, Vol. 73, pp. 93-117.
27
See Kenneth L. Judd, 1997, "The Optimal Tax Rate for Capital Income
is Negative," NBER Working Paper No. 6004.
28
Shinichi Nishiyama and Kent Smetters, Forthcoming, "Consumption
Taxes and Economic Efficiency with Idiosyncratic Wage Shocks," Journal
of Political Economy.
29
See William G. Gale, 2005, "The National Retail Sales Tax: What
Would the Rate Have to Be?", Tax Notes, May 16, pp. 889-911.
30
More specific proposals to simplify the tax system are in Leonard E.
Burman and William G. Gale, 2001, "A Golden Opportunity to Simplify
the Tax System," available at: http://taxpolicycenter.org/publications/template.cfm?PubID=7599.
