[Economic Report of the President (2008)]
[Administration of George W. Bush]
[Online through the Government Printing Office, www.gpo.gov]

 
CHAPTER 5


Tax Policy

Societies face two basic questions with regard to tax policy. The
first question concerns the amount of revenue that should be raised.
That is, what is the appropriate level of taxation? The level of
taxation ultimately reflects views about the appropriate size of
government. If a society believes that the government should play a
large role in the economy, then a high level of tax revenue is
necessary. While taxes are necessary to finance the public sector,
they have a considerable cost to the economy because they distort
incentives and result in lost value of output to society. Without
taxes, individuals would decide where to allocate resources depending
on where those resources are most productive. Taxes give individuals
an incentive to reduce their tax burden by avoiding activities that
are taxed; as a result, decisions about working, saving, investing,
and spending are influenced by tax considearations, resulting in the
loss of output that would have created value for producers,
consumers, and workers. The distortions created by taxes have
important implications for economic growth and the well-being of
Americans.

The second question about tax policy concerns how the tax burden
should be distributed across different members of society and
different types of activities. That is, what is the appropriate
structure of taxation? Different tax structures impose different
costs on the economy in terms of the distortions they create. A more
efficient tax structure raises a given amount of revenue with less
distortion. Different tax structures also give rise to different
distributions of after-tax income, and some distributions of income
may be viewed as more fair than others. A related issue is the timing
of taxes. The use of government debt allows the tax burden to be
spread across time, raising questions about how to tax different
activities and individuals at different points in time.
The key points of this chapter are:
The ratio of Federal taxation in the United States to gross
domestic product (GDP) has fluctuated around an average value of 18.3
percent over the past 40 years; despite the President's 2001 and 2003
tax relief, this ratio was 18.8 percent in 2007, above the 40-year-
average. Under current law, revenues are predicted to grow faster
than the economy in coming years, raising the level of taxation well
above its historical average.
Tax reductions in 2001 and 2003 have considerably lowered the
tax burden on labor and capital income and reduced distortions to
economic decisions. Making these tax cuts permanent can greatly
improve long-term economic outcomes.
In addition to contributing to growth, the tax cuts of 2003
also improved the efficiency of the tax structure primarily by
reducing the double taxation of corporate income.
The business tax structure in the United States still creates
substantial distortions. To attract investment from abroad and
compete more effectively in foreign markets, the United States must
consider how best to address distortions created by the structure of
business taxes, as other countries have done.


The Size of Government: A Historical View
Economists and policymakers have long debated the appropriate role of
the government in a market economy. The government provides some
services-such as national defense and law enforcement-that are
clearly essential for economic growth, but other functions of
government, such as large redistributions of income, are more
controversial. A large public sector imposes a cost on the economy
primarily because the taxes that are required to finance government
programs distort labor supply, saving, and investment decisions,
resulting in lost value of output to society. Thus, our Nation faces
a tradeoff: a larger government can provide more public services and
transfer payments (payments that are not in exchange for goods or
services) to lower-income individuals, but these benefits often come
at the cost of lower economic output and well being.

The cost from tax distortions can be considerable. One recent study
suggests that raising an additional dollar of revenue from the
individual income tax costs the economy approximately 30 to 50 cents.
That is, if taxes increase by $1, taxpayers bear a cost of $1.30 to
$1.50 - the $1 in revenue and 30 to 50 cents from accompanying
distortions. This additional cost of 30-50 cents is known as
deadweight loss. Any government services that are funded with this
revenue would have to have a benefit to society of at least $1.30 to
justify the increase in taxes.

One measure of the role of government is the size of government
spending relative to the economy. Over the past 40 years, Federal
expenditures have averaged 20.7 percent of GDP. Government activities
can be financed by current taxes or borrowing (which will necessitate
higher future taxes or lower future spending). Chart 5-1 shows that
over the past 40 years, the ratio of Federal taxes to GDP has
fluctuated around an average value of 18.3 percent. The ratio rose
well above that level in the late 1960s, the early 1980s, and the
late 1990s. Each of these periods was then followed by several years
in which the ratio fell below its long-term average. Recent swings
have been particularly pronounced with the ratio reaching a post-
World War II high of 20.9 percent in 2000. Tax revenues increased
strongly relative to GDP from 1992 to 2000 as a result of rising real
incomes, increases in capital gains realizations, and the tax
increases of the early 1990s. Tax revenues as a share of GDP tend to
rise when real incomes rise and fall when real incomes fall.
Beginning in 2001, tax revenues began to decline as the economy
slipped into recession and real incomes declined. The ratio of tax
revenues to GDP fell to 16.3 percent (a 40-year low) in 2004. Since
that time, tax revenues have grown faster than the economy, resulting
in a tax-to-GDP ratio of 18.8 percent in 2007, once again above its
40-year average.

While the Federal tax-to-GDP ratio has not exhibited any consistent
trend in the past 40 years, it is projected to grow over the next 10
years. Under current law, the President's tax relief of 2001 and 2003
will expire at the end of 2010. At this time, there will be a
significant increase in the tax-to-GDP ratio. Moreover, even in the
absence of any legislative changes, there is a tendency for the
tax-to-GDP ratio to rise. (While the ratio may not rise every year,
there is an upward trend over time.) In the past, significant tax cuts
(in 1964, 1981, and 2001 to 2003) have maintained the tax-to-GDP ratio
at a relatively stable level. The solid line in Chart 5-2 shows the
projected tax-to-GDP ratio if the President's 2001 and 2003 tax
relief is not extended.







Several factors will contribute to rising revenue in the near term,
including the expiration of the 2001 and 2003 tax cuts, the
Alternative Minimum Tax (AMT), real tax bracket creep, and
withdrawals from tax-deferred accounts.

Expiration of the 2001 and 2003 Tax Cuts
The tax cuts of 2001 and 2003 (discussed in detail below) reduced
individual tax rates on ordinary income, dividends, and capital
gains; increased the child tax credit; reduced the "marriage penalty"
(the additional tax that some couples pay as a result of getting
married); and began a phase-out of the estate tax. These provisions
are set to expire at the end of 2010. If they do, the tax-to-GDP
ratio would climb from the 18.8 percent it reached in 2007 to
approximately 20 percent. Making the tax cuts permanent would lower
this ratio to the 18 to 19 percent range (the dashed line in
Chart 5-2), although the ratio would still continue above the
40-year average of 18.3 percent by the end of the 10-year period
depicted in the figure.







Alternative Minimum Tax

Prior to 1969, a handful of high-income taxpayers used deductions
and  exemptions to substantially reduce or eliminate their income
tax liability. This outcome was perceived as unfair, and to address
this problem, the Alternative Minimum Tax (AMT) was established. In
its current form, the AMT requires taxpayers to compute their tax
liability a second way using a broader definition of income that
reduces or eliminates many of the deductions and exemptions allowed
in the calculation of regular income tax. The taxpayer must pay the
greater of the two tax liabilities. In 1970, only 20,000 taxpayers
were subject to the AMT. However, in recent years, the AMT
increasingly affects middle-income families, primarily because its
parameters are not indexed for inflation. Those who are most
vulnerable include families with many children (giving rise to a
large number of exemptions) and families in high-tax states (giving
rise to a large deduction for state taxes). The solution thus far has
been to pass a series of temporary ``patches'' to limit the scope of
the AMT. The most recent patch keeps the number of AMT filers stable
through 2007 at about 4 million-the same as in 2006-instead of the
increase to 25 million that would have occurred had the patch not
been enacted. The Administration proposes a similar patch for 2008 in
the Budget that will continue to keep the aggregate number of AMT
taxpayers roughly constant. If the AMT is not patched in future
years, the number of taxpayers affected will continue to climb,
resulting in a rising tax-to-GDP ratio. Indexing the AMT parameters
for inflation and extending the tax cuts would lower the tax-to-GDP
ratio below the dotted line in Chart 5-2, unless the revenue loss
from AMT indexation were made up via additional taxes.

Real Bracket Creep
Federal taxes as a whole are progressive, meaning that a family's
average tax rate (total taxes paid as a percentage of income) rises
as its income rises. Recently released estimates suggest that in
2005, taxpayers in the bottom 20 percent of the income distribution
faced an average Federal tax rate of 4.3 percent, while taxpayers in
the top 20 percent faced an average Federal tax rate of 25.2 percent.
(This analysis takes into account individual income taxes, payroll
taxes, corporate income taxes, and excise taxes.) Over time, people's
nominal incomes (not adjusted for changes in purchasing power) tend
to grow. Part of this growth is due only to inflation, but part of
it represents an increase in purchasing power (real income growth) as
productivity improves and we become more prosperous as a nation.
Regular income tax brackets (but not AMT brackets) are indexed for
inflation, which prevents people from moving up to higher brackets
because of inflation (a phenomenon called nominal bracket creep).
However, as people's real incomes grow, they become subject to higher
tax rates. This is called real bracket creep. The implication is that,
even without explicit tax increases, the median income family (that
is, the family whose income places them in the middle of the income
distribution) will face a rising average tax rate over the years
because median incomes are likely to grow faster than inflation.
This will tend to increase the ratio of Federal revenues to GDP.
Withdrawals from Tax-Deferred Accounts
A large amount of individual saving occurs through tax-deferred
savings vehicles, including defined benefit pension plans (which
provide a specified benefit at retirement) and tax-deferred savings
accounts, such as 401(k) plans and traditional Individual Retirement
Accounts (IRAs). Individual and employer contributions to these
tax-deferred savings vehicles are deductible at the time the
contribution is made, and accumulate tax free until retirement. After
retirement, payments from these savings vehicles-including benefits
paid by defined benefit plans and withdrawals from tax-deferred accounts
-are taxable. In comparison, withdrawals from other types of accounts
-for example, ordinary savings accounts and Roth IRAs-do not require
payment of income tax on the withdrawal, and deposits in these accounts
are not tax deductible. At the end of 2002, there was about $9.0
trillion in tax-deferred retirement plans on which tax would be paid
at withdrawal. With the aging of the population that is projected to
occur, there will be an increase in such payments, resulting in
increased government revenue. These withdrawals are different from the
previous three factors for two reasons. First, they cause a temporary
surge in revenue driven by a demographic shift. Second, their impact
will occur over a somewhat longer period than depicted in Chart 5-2.
According to a recent study, these withdrawals are likely to increase
income tax receipts by about 0.25 percent of GDP over the next 25
years, and twice that amount by the end of 75 years.
The factors discussed above-the expiration of the 2001 and 2003 tax
cuts, the expansion of the AMT, real bracket creep, and withdrawals
from tax-deferred savings accounts-are built into the tax code. In
addition to these internal factors, there are also external pressures
for taxes to increase in the future. Total Federal expenditures in
2007 were 20 percent of GDP. However, entitlement programs like
Medicare, Medicaid, and Social Security are facing financial pressures
from rising medical costs and an aging population. Based on current
law, projected benefits under these programs could push entitlement
spending alone to 20 percent of GDP in 2080, compared to 10.6 percent
in 2007. In the absence of needed reforms to reduce projected
spending, this would necessitate unprecedented levels of taxation,
deficit spending, or dramatic reductions in the fraction of economic
activity devoted to other government activities.
The Impact of Recent Tax Reductions
Taxes transfer resources from individuals to the government. The
transfer itself does not represent a net cost to society: any money
given up by taxpayers is gained by the government and can be used to
fund government programs or transfer payments. However, taxes impose
a considerable burden on the economy for several reasons. First,
taxes interfere with the efficient allocation of resources by
changing the rewards from working, saving, and investing. In the
absence of taxes, individuals and firms would allocate resources to
activities where they would be most productive. When taxed,
individuals alter their behavior. For example, high tax rates on
labor income induce individuals to reduce their labor supply, because
the incentive for working is lower. High tax rates on capital income
(the return earned on capital investments) discourage investment in
new capital. A reduction in investment lowers the ratio of capital to
labor and in turn reduces worker productivity and wages. As a result
of these distortions to work, saving, and investment behavior, output
is lost-output that would have created value for producers, consumers,
and workers. This loss of output is called the deadweight loss of
taxation. As discussed above, raising an additional dollar via the
individual income tax imposes a direct cost of $1 on taxpayers (which
merely represents a transfer to the government) and a deadweight loss
of 30 to 50 cents from the lost value of output to society. Second,
high tax rates may also encourage some taxpayers to underreport their
incomes, giving rise to equity concerns and requiring higher taxes on
those who do comply in order to maintain revenue. (While most
taxpayers pay the taxes they owe, there is still a gap between the
amount of taxes that should be paid and the amount that is actually
paid.) Finally, taxes have large compliance costs that reflect the
resources taxpayers use to determine and pay their tax liability
(including the value of time spent keeping records and doing
calculations). In 2004, compliance costs were estimated to be $85
billion for individual income taxes and $40 billion for businesses
other than sole proprietorships.
The tax cuts of 2001 and 2003 significantly lowered the tax burden on
labor and capital income and reduced distortions. The dividend and
capital gains rate cuts enacted in 2003 had an additional benefit to
the economy by improving the efficiency of the tax structure. By
reducing the existing preference for corporate debt financing over
equity financing, these tax cuts reduced the distortion of corporate
finance decisions and improved corporate governance.
Labor Supply
Taxes effectively decrease the wage that workers receive for
providing labor and therefore distort labor supply decisions by
changing the incentive for working. These distortions create
efficiency losses. The tax cuts of 2001 significantly decreased the
tax rates that workers pay on their earned income, thereby reducing
the efficiency losses created by the distortion of their labor supply
decisions.
Individuals decide to work based upon whether take-home earnings
exceed the value of the leisure they forgo (for this discussion,
leisure includes any activities outside the labor market). Take-home
pay declines as the average tax rate, that is, the fraction of income
paid in taxes, rises. Hence, higher average tax rates mean that fewer
individuals choose to work. Moreover, higher marginal tax rates-the
fraction of additional income paid in taxes-reduce the incentive for
working more hours or in a higher-skilled profession. Increases in
both average and marginal tax rates distort labor supply and skill
investment decisions and thus generate efficiency losses.

Individuals vary in their responsiveness to average and marginal tax
rates, so the efficiency losses from taxes differ by group. Studies
show that single mothers and married women are particularly sensitive
to high average tax rates. Their cost of working is higher because of
child care and other home production demands. The 2001 tax cuts
lowered average tax rates at all points of the income distribution,
thereby making work decisions more efficient (that is, closer to what
they would be in the absence of tax distortions). A recent study
suggests that the 2001 tax cuts led single mothers to allocate more
of their time to market work. In contrast, several studies suggest
that men and single women without children are not affected much
by average tax rates when deciding whether to work. The
responsiveness of married women to high average tax rates has been
falling over time as they become more attached to the labor market
(as men have more traditionally been).
High marginal income tax rates may discourage workers from working
more hours, choosing higher-paid occupations, and investing more in
education and other skills that would increase their earnings. To
see why higher marginal tax rates have these effects, imagine a
worker with only a bachelor's degree deciding between a career as a
40-hour-per-week accountant in a small firm paying around $40,000 per
year versus a career as a 70-hour-per-week self-employed consultant
with an MBA earning around $80,000 per year. Suppose that the worker
would pay $4,000 per year in taxes in the accounting job and $18,000
per year in the consulting job. After taxes, the additional income
for the more demanding career would be $26,000 per year. The marginal
tax rate would be 35 percent (see Table 5-1).
Now suppose a change in tax policy reduces taxes for the accounting
job to $1,000 and increases taxes for the consulting job to $21,000.
Instead of a 35 percent marginal tax rate on the additional $40,000
in pre-tax income, there would be a 50 percent marginal tax rate. This
change in tax policy reduces the additional return to the more
demanding career from $26,000 to just $20,000 per year, a 23 percent
drop in the return to the more lucrative career (see Table 5-1).
Factoring in 30 more hours per week working, the greater stress in
the consulting job, and the costs of getting the MBA, this tax
policy change could induce this worker to choose the less demanding
career, thereby creating an efficiency loss. So even if this change
in tax policy is revenue neutral (that is, the policy does not change
overall average tax rates), the higher marginal taxes would reduce
overall economic efficiency because they alter the way wages allocate
workers to jobs and decrease incentives to choose higher-paying
careers with longer hours, greater intensity demands, and more costly
skill investments. The tax cuts in 2001 and 2003 generally reduced
marginal tax rates and reduced these distortions, thereby
encouraging workers to become more productive.
Saving and Investment
When individuals receive income, they can either spend it on current
consumption or save it to fund future consumption. Individual savings
gets channeled into capital investments. For example, an individual
may save by buying financial assets, such as stocks or bonds. Firms
use the funds raised from selling stocks and bonds to finance capital
investments, such as buildings or equipment. These investments
generate income, which individual savers receive in the form of
interest payment on bonds, or dividends and capital gains on stocks.
Investment plays an important role in improving the well-being of
Americans, as increases in the amount of capital per worker result
in productivity increases and economic growth.






An important tax policy issue concerns the treatment of income
generated by capital investments. Taxes on capital income discourage
saving by individuals and investment by businesses. This lowers the
capital-to-labor ratio and harms long-run economic growth. Currently,
when firms earn income from their capital investments, they may be
subject to a firm-level tax on this amount (after subtracting
depreciation and interest costs). In addition, individual savers,
who provide the funds used to finance these investments, pay income
tax on the return on their savings (which includes dividends,
capital gains, interest, and rent). As a result, capital income is
often taxed at both the firm and the individual level, resulting in
double taxation.
Individuals save so they can consume resources in the future, rather
than today. Firms invest so that they will be more productive and
profitable in the future. Taxes on capital income lower the return to
saving and investment, thereby favoring current consumption over
future consumption. For example, suppose a corporation is considering
the purchase of a machine that will be financed by selling
additional shares of stock, and that the rate of return on the
investment-net of depreciation, or the reduction in the value of the
machine-is 10 percent. Suppose further that individual savers are
willing to purchase the shares if they receive a return of at least 6
percent. That is, they are willing to sacrifice $1 of current
consumption (by buying the shares) in exchange for $1.06 of
consumption 1 year from now. The investment is socially beneficial
because it generates a 10 percent rate of return, and the savers
providing the funds would have settled for 6 percent. At the firm
level,the income generated by the machine is subject to the corporate
income tax. If the corporate tax rate is 35 percent, and the firm is
allowed to deduct actual depreciation, then the after-tax return
generated by the machine is 6.5 percent. Suppose the firm then pays
its shareholders the entire 6.5 percent return in the form of
dividends. If the dividend income tax rate is 15 percent, savers
are left with a 5.5 percent after-tax return. The rest of the initial
10 percent return (4.5 percent) goes to the government. Because the
5.5 percent after-tax return is less than the 6 percent that the
individual savers require to be willing to forgo current
consumption, the investment is not made even though the total return
is still 10 percent (4.5 percent to the government plus 5.5 percent
to the savers). Consequently, taxes on capital income distort
saving and investment decisions. Longer time horizons tend to
magnify this distortion because lower after-tax returns get
compounded over time.
Firm-level taxes on capital income vary depending on the
organizational form of the firm. Some business income, including
that of sole proprietorships, Subchapter S corporations, and
partnerships, is taxed under the individual income tax system. These
firms are known as flow-through businesses because they face no firm-
level tax; instead, the firms' income flows through to their owners,
who pay personal income tax on it. On the other hand, Subchapter C
corporations fall under the corporate tax system. C corporations
(hereafter simply referred to as corporations) pay a firm-level tax
on the firm's income after deducting costs including wages, interest
payments, raw materials, and depreciation.
Current U.S. tax policy is a hybrid of an income tax and a
consumption tax. Some capital income is exempt from tax, as it would
be under a consumption tax. For example, at the individual level,
the return to saving through individual retirement accounts (IRAs)
and employer-sponsored retirement plans accumulates free of tax.
According to recent estimates, about 35 percent of the return to
household financial assets effectively receives consumption tax
treatment. The remainder is subject to income tax treatment. At the
firm level, firms can often take advantage of accelerated
depreciation provisions-which allow them to deduct depreciation from
their income before it actually occurs-to lower their tax liability.
Accelerated depreciation lowers the tax burden on investment.
The tax reductions of 2001 and 2003 have significantly reduced the
tax burden on capital income. By lowering individual income taxes,
the 2001 tax cut lowered the top marginal tax rate on flow-through
businesses from 39.6 percent to 35 percent. Individuals also pay
these reduced tax rates on their interest income. The 2001 tax cuts
also included a phased-in elimination of the estate tax (or tax
imposed on assets left to one's heirs). Since the estate tax is a tax
on wealth, if it were permanently eliminated, it could be expected
to increase saving and investment. The tax cuts of 2003 included
cuts in dividend and capital gains taxes. As discussed below, if
these tax cuts are made permanent, they will have a substantial
impact on investment and long-run economic growth.
Corporate Financial Policy and Governance
Tax reforms can result in considerable economic benefits even when
they do not lower the overall tax burden. This outcome is
accomplished by improving the efficiency of the tax structure, so
that the same amount of revenue can be raised with less distortion.
The reverse can be true as well: a revenue neutral change, or even a
tax cut, can reduce well-being if it is poorly structured.
The tax cuts of 2003 improved the efficiency of the business tax
structure by reducing the high tax burden on corporate equity that
results from double taxation. For funding investment in new capital,
firms generally have a choice between debt (issuing bonds) and
equity (retaining earnings or issuing new shares of stock).
Corporations pay tax on their revenue minus their costs. Costs
include wages, interest, raw materials, and depreciation. Corporate
profit is then either paid out to shareholders as dividends, or
reinvested in the company (eventually resulting in capital gains for
shareholders). Shareholders are taxed at the individual level on any
dividends they receive, and on any capital gains they realize when
they sell the stock. Double taxation of corporate income imposes a
particularly high burden on equity-financed corporate investment. In
comparison, because interest payments are deductible to the firm
(and taxable to bondholders), corporate debt is only subject to one
layer of taxation. Therefore, corporations have a strong incentive
to use debt financing, rather than equity financing, for new
investment. The overuse of debt financing increases the chances of
bankruptcy: when a firm has high debt payments, there is a greater
probability that the firm's income will be insufficient to cover
these payments. Bankruptcies subject investors to additional costs
and risks.

The tax cuts of 2003 also reduced the tax bias against paying
dividends compared to retaining earnings. Prior to 2003, long-term
capital gains were taxed at a maximum rate of 20 percent, while
dividends were potentially subject to the top individual income tax
rate (38.6% in 2002). In addition, capital gains income has another
tax advantage over dividend income: taxes are deferred until the
asset is sold. Thus, capital gains can accumulate tax free, while
dividends are taxed when they are paid out. Through compounding, the
difference in tax can be substantial, especially over a long period
of time.
The tax cuts of 2003 lowered the top tax rate on both qualified
dividends and long-term capital gains (capital gains on assets held
for more than a year) to 15 percent. While capital gains still have
a tax advantage over dividends as a result of deferral, the
differential treatment has been reduced considerably. This policy
change appears to have had a marked impact on firm behavior. As shown
in Chart 5-3, the growth in dividend income received by households
increased substantially after 2003. In the 20 years prior to 2003,
dividend income grew at an average rate of 5.9 percent per year;
following the 2003 tax cut, growth increased to an average of 13.7
percent per year. This result has been confirmed in a number of
formal studies. (The 2004 spike in the graph represents a special
one-time dividend paid by Microsoft Corporation.)
This increase in dividend payments reflects the reduction in the tax
bias against dividends. Paying dividends can have a number of
benefits for corporate governance, and there is an efficiency loss
when the tax code discourages firms from using dividends when they
are appropriate. First, dividends can be used to return funds to
shareholders, who can decide how to reinvest them, rather than
leaving funds in the hands of corporate managers. Because a portion
of managers' pay is independent of the firm's performance, managers'
interests generally differ from the interests of shareholders, so
managers may have an incentive to use retained earnings in a way that
does not maximize the value of the firm. Second, paying dividends
can help firms signal their profitability to investors. Thus,
corporate governance may suffer if the tax code penalizes dividends
relative to capital gains.







Significance of Tax Cuts to Individuals
The tax cuts since 2001 lowered taxes overall and across all
income groups. Average Federal tax rates (which include income,
payroll, corporate, and estate taxes) are estimated at 21.7 percent
in 2007, but would have been 23.8 percent in the absence of the tax
cuts (see Table 5-2). For taxpayers in the bottom 20 percent of the
income distribution, Federal tax rates are 3.4 percent, which is
lower than the 3.7 percent they would be in the absence of the tax
cuts. In addition, over 5 million taxpayers in 2007 are projected to
have had their Federal income tax liability completely eliminated by
the tax cuts.






The tax cuts increased the share of Federal taxes being paid by
high-income taxpayers; the top 20 percent of taxpayers are estimated
to have paid 73.0 percent of overall Federal taxes in 2007, but would
have paid a somewhat lower share, 72.4 percent, without the tax cuts
(see Table 5-2). Conversely, the tax cuts decreased the share of
Federal taxes being paid by moderate and middle-income taxpayers;
the second and third quintiles (from 20 to 60 percent in the
income distribution) are estimated to have paid 9.5 percent (2.1
percent plus 7.4 percent) of overall Federal taxes in 2007, but
would have paid 10.0 percent (2.3 percent plus 7.7 percent) without
the tax cuts.
In addition to distorting work and skill investment decisions, the
tax system can also distort marriage decisions. As discussed in
Box 5-1, a progressive tax system cannot simultaneously treat all
families with the same income equally and be marriage-neutral. This
has resulted in a tax system with marriage bonuses (mostly for
couples with dissimilar incomes) and marriage penalties (mostly for
couples with similar incomes), although on net it encourages
marriage (even before the 2001 tax cuts). It should be noted that
both marriage bonuses and penalties distort marriage decisions and
potentially generate efficiency losses. However, if marriage
generates some greater social good that should be subsidized,
marriage bonuses may improve efficiency on net.
The 2001 tax cuts, in general, increased marriage subsidies and
reduced marriage penalties in the tax system by: (1) expanding
the Earned Income Tax Credit (EITC) for married couples only, (2)
expanding the 15 percent bracket only for married couples, (3)
expanding the standard deduction only for married couples, and (4)
doubling the child tax credit and making it partially refundable.
Recent research estimates that the tax cuts, on average, increased
the subsidization of marriage by the tax system by about $1,000 per
year, although the effect for a particular family depends on family
income, number of children, and the share of family income earned by
each spouse. It is estimated that these tax changes should
eventually increase marriage rates by about 1 to 4 percentage points.
Economic Benefits of Lower Taxes

The previous sections focused on specific ways in which taxes can
distort individual behavior. The analysis suggests that recent tax
cuts have reduced distortions to labor supply, saving, investment,
and corporate governance. A recent study projects that the
introduction of the 2003 tax cuts resulted in an immediate increase
in GDP in 2003. But because the cuts are temporary, they will have
less impact on decisions that generate payoffs far in the future than
they would if they were permanent. For example, the decision to
undertake education depends on the effect of education on wages over
potentially long careers. Thus, they can only have a limited impact on
long-term economic
----------------------------------------------------------------------
Box 5-1: Marriage Penalty Basics

It is widely acknowledged that a tax system cannot simultaneously
accomplish the following three goals:

1. Progressivity: average income tax rates rise with family income
2. Family neutrality: families with equal incomes pay equal taxes
3. Marriage neutrality: taxes paid by a family do not depend on
marriage
The inherent conflicts in these three goals can be illustrated by
considering a few examples. Consider a couple without children with
one spouse who earns $60,000 and another who does not work. Under 2007
tax law, that couple pays $5,592 in Federal income taxes, but would
pay a total of $9,236 if they were not married and both were filing
individually. The resulting marriage bonus of $3,644 is generated
because the nonworking spouse serves as a tax deduction for the
higher earning spouse. The current tax system is not marriage-neutral.

Alternatively, suppose that each spouse earns $30,000, resulting in
the same family income of $60,000. Current tax law is family-neutral,
so this couple pays the same $5,592 as above. If the tax system is
changed so that all individuals file separately, each spouse pays
$2,796 for a total of $5,592. That is the same as they would pay on a
family income of $60,000 but is $3,644 less than the combined tax
liability of the family above. A progressive tax system that has all
taxpayers file individually cannot be family-neutral.

Finally, if the tax system is changed so that all taxpayers pay 10
percent on all of their income, taxes are $6,000 for each family
regardless of whether the couple is married or how the earnings are
split between the two spouses. The tax system is marriage and family
-neutral, but it would no longer be progressive, because the average
tax rate would be 10 percent for all taxpayers.
----------------------------------------------------------------------


performance. Making them permanent can substantially improve
economic efficiency. The Treasury Department estimates that if the
tax cuts of 2001 and 2003 were made permanent and paid for by
reductions in future government spending, economic output would
increase by 0.7 percent in the long run. However, the benefits to
the economy might be offset if the extension of the tax cuts results
in additional government borrowing or future tax increases, rather
than spending cuts. The Treasury Department also estimates that if
the tax cuts were made permanent but offset by other revenue raising
tax measures in the future, then economic output would decline by 0.9
percent in the long run. The concern about long-term financing for the
tax cuts is particularly important because of the likelihood of
rising spending pressures in the future. The Office of Management and
Budget projects, for example, that under current law total noninterest
Federal spending could reach 25 percent of GDP by 2080, compared
with 18.2 percent today. The breakdown of projected spending in
Chart 5-4 shows that the main driving force behind this increase is
the growth in spending on entitlement programs, primarily Medicare,
Medicaid, and Social Security, which could reach approximately 20
percent of GDP by 2080. The benefits of making the tax cuts permanent
might also be offset if the tax cuts are financed by a reduction in
efficient government spending (spending whose benefits exceed both
the direct cost to the taxpayer and the deadweight loss).








The Structure of Business Taxes
Despite recent reforms, the business tax structure still creates a
number of distortions in its treatment of capital income. To the
extent that the U.S. tax system resembles an income tax, it
encourages current consumption rather than saving. Beyond this,
however, the tax system imposes differential tax burdens on different
types of investments, thereby leading to a misallocation of resources.
Ideally, firms should undertake investments that generate the highest
rate of return, independent of taxes. If all investment returns are
taxed at the same rate, then the projects with the highest returns
will still be selected (although investment overall will fall because
investment returns overall are taxed). However, if different kinds of
investments face different tax rates, then a lower-return project may
be selected over a higher-return project because the after-tax return
could be higher for the lower-return project.
As noted above, the tax burden on investment is affected by both
firm-level taxes (such as the corporate income tax) and individual
-level taxes on the return to saving (such as dividend and capital
gains taxes). The complexity of the tax code makes it difficult to
measure the true tax burden on investment returns. For example,
corporate earnings are taxed at a maximum Federal rate of 35 percent.
However, that tax burden is reduced by accelerated depreciation,
special tax preferences for certain activities, and the interest
deduction. Also, while some kinds of savings are subject to personal
income tax, other kinds (for example, retirement savings accounts)
accumulate tax free. A standard approach to quantifying the
distortions is to compute the effective marginal tax rate, which
measures the percentage difference between the before-tax and after-
tax returns on a new investment, taking into account the
complexities of the tax code, and both firm- and individual-level
taxes. The effective marginal tax rate is most relevant when a firm
decides whether to undertake a new investment.
Table 5-3 shows the effective marginal tax rates on different kinds
of investments. It is clear from the table that tax rates vary
considerably across investments, depending on the type of capital
involved and the method of financing. Equity-financed corporate
investment faces the highest effective tax rate of 40 percent. This
is still the case even though the tax cuts of 2003 substantially
reduced the double taxation of corporate equity. The tax rate on debt
-financed corporate investment is actually negative, a result of the
interest deduction combined with accelerated depreciation allowances.
Noncorporate investments face a low tax rate because noncorporate
firms are treated as flow-through entities and are not subject to
double taxation. Owner-occupied housing faces a very low tax rate.
The return to an owner-occupied home is the rental value of the home
to the occupant, which is not subject to income tax.

These results suggest several distortions. First, housing is favored
relative to other capital. While there may be reasons to favor owner-
occupied housing, its benefits must be weighed against the value of
other kinds of capital. Second, there is a distortion across different
types of business investment. For example, equipment is lightly taxed
relative to structures and inventories. Third, taxes distort a firm's
choice of organizational form. The corporate form of organization is
unattractive from a tax standpoint, leading firms to become flow-
through entities even in situations in which the corporate form would
allow the most effective use of resources. Finally, there is a
distortion to corporations' financing decisions, with debt receiving
a tax advantage over equity.
There are two broad directions for reform. First, efficiency could be
improved by reducing the disparate tax treatment of different kinds of
investment. There are a number of reforms that could help to achieve
this goal. For example, the Treasury Department estimates that if
special preferences were eliminated, the corporate tax rate could be
reduced from 35 percent to 31 percent and still raise the same amount
of revenue. Further integration of the personal and corporate tax
systems would alleviate the double taxation of corporate income. For
example, some countries in the Organization for Economic Cooperation
and Development (OECD), including the United Kingdom, Canada, and
Mexico address the double taxation of capital income by giving
investors a tax credit for taxes paid at the corporate level.
Second, reducing the tax burden on investment can improve long-run
economic performance by increasing the ratio of capital to labor,
thereby boosting labor productivity and earnings. There are two ways
to reduce the tax burden on investment at the firm level. One is to
reduce the corporate tax rate, and the other is to allow full or
partial expensing of new investment. Full expensing allows the firm
to fully deduct the cost of new investments at the time the






investments are made. A more modest approach would be to allow
partial expensing, under which a firm could immediately deduct a
fraction of the investment's cost. As shown in Box 5-2, full
expensing reduces the firm-level tax on new investments to zero.
----------------------------------------------------------------------
Box 5-2: Expensing versus Corporate Rate Reductions

Consider a firm that purchases a machine for $100. A year later, the
machine produces output worth $50. The firm then sells the machine for
$60. Thus, the return from investing $100 in the machine is 10 percent
(the firm earns $50 + $60 = $110). The firm can finance the initial
$100 investment by borrowing (debt), by reinvesting earnings, or by
issuing new shares.

Assume that the firm either reinvests earnings or issues new shares
(equity financing). Under an income tax, the firm�s net income is $10,
the value of the machine's output ($50) plus the proceeds from selling
the machine ($60) minus the cost of the machine ($100). If the
corporate income tax rate is 35 percent, the firm pays $3.50 in tax on
its $10 income, leaving it with $6.50 after taxes (a 6.5 percent
after-tax return). Thus, an income tax creates a distortion to the
investment decision by lowering the after-tax return on the
investment.

In contrast, full expensing allows the firm to deduct the entire $100
cost of the machine up front. Thus, the firm�s taxes go down by $35
when it makes the investment, and the effective cost of the machine
is $65, rather than $100. The firm earns $50 from the machine�s output
plus $60 from the sale of the machine, and the total income of $110 is
taxed at a rate of 35 percent (because the firm already deducted the
cost of the machine upon purchase). Thus, the tax paid is $38.50, and
the firm's after-tax income is $71.50. The rate of return is
($71.50 - $65) / $65 = 10 percent, which is the same as it would have
been without a tax. Effectively, full expensing makes the government a
partner in the investment: the government pays for 35 percent of the
investment's cost (via the deduction), and receives 35 percent of
its return.

To be most effective in reducing distortions, full expensing would
need to be combined with elimination of the interest deduction.
Suppose interest payments remain deductible under the full-expensing
approach described above and the firm borrows money to fund half of
the machine's cost ($50) at a 10 percent interest rate. The effective
cost of the machine is $65 due to expensing. Therefore, the firm
spends $15 of its own funds ($65 - $50 = $15) for the machine. Next
year, the machine generates $110 of income, and the firm pays $55 to
the lender (principal plus interest). The firm deducts the interest
payment of $5 from its income, resulting in taxable income of $105.
At a 35 percent tax rate, the firm's tax liability is $36.75. The
firm is left with a profit of $18.25, a return of 22 percent on its
initial $15 investment. Thus, the tax on the investment�s return is
negative (the investment receives a subsidy from the government). If
the interest deduction were not allowed, the firm's tax bill would be
$38.50 (just as above), and the profit after repaying the lender $55
and paying taxes would be $16.50, a 10 percent rate of return. With
full expensing and no interest deductibility, there is no distortion
to either the investment decision or the financing decision.

Another alternative is to reduce the corporate rate. Using the same
example as above, consider the impact of reducing the corporate tax
rate from 35 percent to 10 percent. The firm makes its $100
investment, and next year pays tax on its net income of $10. This
leaves the firm with an after-tax return of 9 percent. Since the
after-tax return is still below the before-tax return, there is a
distortion to the investment decision. However, there is less of a
distortion than with the 35 percent tax rate.
----------------------------------------------------------------------

In recent years, other countries have taken the approach of cutting
the corporate tax rate. A tax rate cut affects all capital, both new
and old. In comparison, expensing is targeted to new investment only.
Thus, expensing generates a greater increase in investment for any
given revenue reduction. Another difference between tax rate cuts and
expensing arises because firms sometimes earn returns on their
investments that are above the normal, ordinary return. To illustrate
this, consider the example in Box 5-2, in which a $100 investment
yields a 10 percent rate of return. Suppose that the next best use of
the firm's funds would produce a return of 5 percent. The return of
5 percent represents the opportunity cost of the funds, also known as
the normal return. As long as the investment return is above the
normal return, the firm will undertake the project; thus, taxing any
returns that exceed the opportunity cost of funds (called
supra-normal returns) does not create any distortions. Expensing
exempts only the normal return from taxation; supra-normal returns
are subject to taxation. In the example, $5 of the investment's
payoff represents compensation for the firm's opportunity cost, and
$5 represents a supra-normal return. If the corporate tax rate is
35 percent, full expensing would give the firm a deduction worth $35
this year, and require it to pay a tax of $38.50 next year.
Effectively, the firm is able to defer $35 of tax liability for 1
year. The value to the firm of deferring the tax until next year is
$1.75 (5 percent of $35). However, next year, the firm must pay $3.50
in additional taxes. Thus, the firm has effectively paid a tax of $1.75
(the $3.50 of additional taxes minus the $1.75 value of deferral),
which represents a tax of 35 percent on the $5 supra-normal return.
Note that taxing the supra-normal return does not result in any
distortions, because the firm's decision to undertake the investment
does not depend on the tax. If the normal return were instead 10
percent, then the deferral of tax would be worth $3.50 to the firm,
and there would be no effective tax on the investment return. In
contrast to expensing, a corporate tax rate cut lowers the tax on
both normal and supra-normal returns.
The efficiency of the business tax structure in the United States is
particularly important as other countries undertake major corporate
tax reforms. Capital is mobile across international borders, and the
business tax environment is important in ensuring that the United
States continues to attract investment from abroad, and that U.S.
firms can compete effectively in foreign countries. In the mid-1980s,
the average statutory corporate tax rate (weighted by GDP) across OECD
countries was 44 percent. The U.S. tax reform of 1986, which reduced
the corporate tax rate from 46 percent to 34 percent, made the United
States a relatively low-tax country at the time of the reform. Since
that time, however, the OECD-average corporate tax rate has fallen
below that of the United States. These comparisons refer to statutory
tax rates. The United States has relatively generous accelerated
depreciation provisions and a multitude of business-level exemptions
and deductions that reduce the tax burden on investment below the
statutory rate. However, the effective marginal tax rate on corporate
investment is still high: compared to other G7 countries (France,
Germany, the United Kingdom, Canada, Italy, and Japan), the United
States imposes an above-average marginal effective tax rate on
corporate investment for domestic debt and equity holders in the top
individual income tax bracket. In contrast, the U.S. average
corporate tax rate (the total amount of corporate taxes paid as a
percentage of corporate operating surplus) is low relative to other
countries. This fact highlights the inefficiency and complexity of
the corporate tax system. The marginal tax rate represents the
additional tax burden a firm faces when it undertakes a new
investment; therefore, it is the relevant tax rate for new investment
decisions. This distortion is larger in the United States than in
other countries. Despite the larger distortion, the corporate tax
raises less revenue in the United States than in other countries, as
evidenced by the fact that the average tax rate is lower. The
implication is that investment incentives could be improved without
a reduction in government revenue.
Conclusion
The analysis in this chapter has focused on both the level and
structure of taxation. Over the past 40 years, Federal revenues have
fluctuated around 18.3% of GDP. Under current law, however, tax
revenues are scheduled to rise much faster than GDP in coming years.
Furthermore, over longer periods of time, projected growth in
entitlement spending will put pressure on taxes to rise. Because
taxes distort incentives, these trends have important implications
for economic growth. Extending the tax cuts of 2001 and 2003 would
improve labor supply and savings incentives and result in less
distortion of corporate finance decisions. Combined with control of
entitlement spending, and a long-term solution to the Alternative
Minimum Tax, this can have a beneficial effect on long-run growth.
The tax cuts of 2001 and 2003 have also improved the efficiency of
the tax structure, particularly with respect to the double taxation
of corporate income. However, the structure of business taxation
still creates a number of distortions and puts the United States at
a competitive disadvantage globally. Even revenue-neutral reforms can
result in economic gains if they remove unnecessary distortions.