Would Tax Reform Alter the Economy's Growth?

^Pr^ep^ar^{ed f}^o^r^M^e^m^b^ers^an^d^Com^mⁱ^t^t^ees^of^C^ong^r^e^ss

An argument frequently made by proponents of tax reform is that it would boost economic
growth. “Tax reform” means different things to different people. This report uses the recent
recommendations of the President’s Panel on Tax Reform as a launching point. (Tax reform may
have important effects on the efficiency, equity, and simplicity of the tax system, but these issues
are not addressed in this report.)
The Panel’s proposal is meant to be revenue neutral over 10 years compared with a baseline that
assumes that the 2001 and 2003 tax cuts are not allowed to expire as scheduled and the other tax
proposals in the President’s budget are enacted. This means that the Panel’s proposal raises less
revenue than a current law baseline, such as the CBO baseline, which assumes that expiring tax
provisions are allowed to expire. The proposal would raise, on average, about as much revenue as
a share of GDP annually as the government currently collects. Thus, the proposal would not be
expected to have a stimulative or contractionary effect on aggregate spending in the economy in
the short run. Because budget deficits reduce national saving, it would be expected to cause about
the same degree of crowding out of private investment as is occurring today (and more crowding
out than would occur under a current law baseline).
Because the Panel’s proposal is revenue neutral, it is made up of revenue raisers and revenue
reducers that cancel each other out overall. The largest revenue reducer is the repeal of the
alternative minimum tax (AMT). This revenue loss is so large that the Panel is not left with much
scope to significantly reduce marginal income taxes and remain revenue neutral. Most of the
other tax reductions in the proposal are focused on taxes on capital (or saving). (To remain
revenue neutral over 10 years and cut taxes on saving, the Panel proposes the expansion and
conversion of “back-loaded” tax-preferred saving accounts that raise revenue in the short run but
lose revenue in the long run.) Although many economists have criticized the AMT, it has marginal
rates that are similar to the regular income tax for most affected taxpayers, so repealing it does
not influence incentives to work or save in most cases.
The overall effect on economic growth is likely to be negligible. Any effects on labor supply
would be small and ambiguous because some workers would face higher marginal tax rates, and
others would face lower tax rates. The proposal could potentially have a larger effect on saving,
but the theoretical and empirical evidence is mixed. Since 1980, tax preferred saving accounts
have been expanded and taxes on capital gains and dividends have been reduced; these changes
have not stopped the personal saving rate from declining from over 10% of disposable income to
zero in that time. This suggests that saving may not be very responsive to further tax reductions.
The Panel proposes reducing or eliminating many of the tax expenditures in the current system.
Although these changes may increase economic efficiency (by reducing distortions in market
outcomes), they are unlikely to have more than negligible effects on economic growth, which
depends on increases in labor, capital, and productivity. This report will be updated as events
warrant.

Introduc tion ..................................................................................................................................... 1
Economic Growth: Short Run vs. Long Run...................................................................................2
Total Revenues and Fiscal Policy..............................................................................................2
“Supply Side” Effects and Long-Term Growth...............................................................................3
Effects on Labor Supply............................................................................................................4
Effects on Saving......................................................................................................................6
The Tax Reform Panel’s Quantitative Estimates of Their Proposals’ Effect on Growth...........9
Tax Expenditures...........................................................................................................................10
Efficiency vs. Growth..............................................................................................................11
Conclusion ...................................................................................................................................... 11
Figure 1. Labor Supply from 1990-2005.........................................................................................6
Figure 2. Household Saving, 1960-2005.........................................................................................8
Author Contact Information..........................................................................................................12

An argument frequently made by proponents of tax reform is that it would boost economic
growth. The term “tax reform” means different things to different people. This report uses the ¹
recent recommendations of the President’s Panel on Tax Reform as a launching point. Although
the Panel proposes making many small changes to the tax system, its recommendations
incorporate a few broad concepts that will be evaluated:
• Elimination of the alternative minimum tax (AMT). This would reduce revenues ²
by $1.2 trillion over 10 years. This lost revenue is offset by other provisions in
the proposal that raise revenue.
• Reduction in certain tax expenditures, such as the mortgage interest deduction
and the deductibility of health insurance, and elimination of other tax
expenditures, such as the education credits and deductibility of state and local
taxes.
• Reduction in the number of marginal income tax brackets from six to three or
four.
• Replacement of the standard deduction and personal exemptions with a family
credit.
• Reduction in the taxation of capital through higher contribution limits on tax-
preferred saving accounts, reduced taxation of dividends and capital gains, a
lower corporate income tax rate, and more favorable rules for writing off capital
investment.
Overall, the Panel stated that its proposed tax system would generate revenue equal to a baseline
under which the 2001 and 2003 tax cuts were extended. It would generate less revenue than a
current law baseline (such as the Congressional Budget Office’s) under which those tax cuts
expire as scheduled.
The Panel proposed two separate plans: the Simplified Income Tax (SIT) Plan and the Growth
and Investment Tax (GIT) Plan. Both plans incorporate all of these broad concepts and differ
mostly on the corporate side (the GIT plan would more fundamentally alter the structure of the
corporate tax system). If the Administration adopts a proposal based on the Panel’s
recommendations, the proposal will be presented to Congress. This report will evaluate how each
of these concepts could potentially influence the growth rate of the economy. Tax reform may
also have important effects on the efficiency, equity, and simplicity of the tax system, but these
issues are not addressed in this report.

¹^P^r^e^s^ideⁿ^{t’s A}^d^vⁱ^sory^P^a^ne^{l o}ⁿ^Fe^de^ra^{l T}^a^x^Re^f^o^r^m^,^Sim^p^le^{, Fa}^ir^{, a}^{nd Pr}^o-^G^r^ow^t^h^:^Pr^o^pos^als^to^Fix^Am^e^r^ic^a’^s^Tax
^Sy^s^t^e^m^,^No^v.²⁰⁰⁵^.
²^U^.^{S. D}^e^pa^r^t^m^e^{nt of}^the^T^r^e^a^s^ur^y^,^F^a^ct^S^h^eet^{: T}^h^{e T}^o^l^l^of^T^w^o^T^a^{xes: T}^h^{e R}^e^gu^l^a^{r I}ⁿ^co^{me T}^a^{x an}^d^t^h^{e A}^M^T^{, Ma}^r.^2,
²⁰⁰^5.^Se^e^a^l^s^o^C^R^S^R^e^por^{t R}^S²²¹⁰⁰^,^T^h^{e A}^l^t^e^rna^tⁱ^{ve Mi}ⁿⁱ^mu^m^T^a^{x f}^o^{r Ind}ⁱ^vi^du^a^l^s:^L^e^gⁱ^sl^a^tⁱ^{ve In}ⁱ^tⁱ^a^tⁱ^{ves i}ⁿ^t^h^{e 10}⁹^t^h
^C^ongr^e^s^s^,^by^G^r^eg^g^A^.^Es^eⁿ^w^e^{in. G}^e^ne^r^a^{l inf}^o^r^m^a^{tion on}^the^A^M^{T c}^aⁿ^be^f^oun^{d i}ⁿ^C^R^S^R^e^por^t^R^L³⁰¹⁴^9,^The
^Alte^r^nativ^e^Mⁱ^nim^u^m^T^a^x^for^I^ndi^vⁱ^dua^ls^{, by}^Ste^v^eⁿ^Ma^g^u^ir^e^.^T^h^e^pa^ne^{l di}^d^not^pr^ov^ide^of^fⁱ^cⁱ^a^l^r^e^venue^e^s^tim^a^t^e^s^of
^its^pr^o^pos^a^l^s^.

Economic growth refers to an increase in gross domestic product (GDP), the goods and services
currently produced by the economy. Economists distinguish between short run and long run
sources of growth. In the long run, production is determined solely by inputs of labor and capital,
and how productively those inputs are used. Thus, in the long run, economic growth is
determined by how quickly inputs of labor and capital are increased and how quickly productivity
grows. Most of this report will analyze how various provisions of the tax reform proposal would
affect long-run growth, but first it will briefly analyze the proposal’s overall effect in the short
run.
Although growth in labor, capital, and productivity determines the economy’s average growth
rate over time, most economists believe that it cannot explain the short-term fluctuations
(recessions and expansions) of the business cycle. In the short run, these fluctuations in growth
are caused by changes in aggregate spending (demand) in the economy. Recessions occur when
spending—whether it be consumer, investment, net export, or government spending—falls below
the economy’s productive capacity. As a result, labor becomes unemployed and physical capital
lies idle. When spending exceeds the economy’s productive capacity, price inflation occurs. Any
mismatch between spending and production is strictly temporary: over time, the economy
automatically adjusts to bring the two back in line. Since World War II, recessions have lasted on
average for 10 months.
The government has two tools at its disposal to alter aggregate spending: monetary policy and
fiscal policy. Fiscal policy alters aggregate spending through changes in the budget deficit. When
the government borrows more from the private sector to increase government spending or reduce
taxes, it increases aggregate spending. When the government borrows less, aggregate spending
falls. After its initial expansionary effects, a constant budget deficit over time neither expands nor
contracts aggregate spending. Any potential changes in the aggregate spending caused by changes
in deficit resulting from tax reform could presumably be offset by changes in monetary policy
since the path of deficits would be preannounced in advance.
Fiscal policy influences aggregate spending in the short-term, but it also has long-term effects on
growth. A larger deficit, whether it be caused by tax cuts or higher government spending, must be
financed out of the finite pool of private saving. Private capital investment must be financed out
of the same pool, so deficits are said to “crowd out” capital investment spending by bidding up
interest rates to secure those finite resources. With higher interest rates, there are fewer profitable
investment opportunities. Crowding out could be avoided by borrowing abroad, but then the
additions to the capital stock would be foreign owned, and the future income it produced would
not flow to Americans. Because capital investment increases the productive capacity of the
economy, crowding out leads to a smaller economy than would otherwise have been over the ³
longer run.

³^Se^e^C^R^S^R^e^por^{t R}^L³¹⁷⁷⁵^,^D^o^B^udge^t^D^e^fic^its^Pus^h^U^p^Iⁿ^te^r^e^s^t^R^a^te^s^an^{d I}^s^This^t^h^e^Re^le^v^ant^Q^u^e^s^tion^?^{, by}^Ma^rc
^L^a^bonte^.

Should the Panel’s proposal be considered revenue neutral? It depends what one considers to be
the baseline to which it is compared: current law (in which current tax cuts and AMT relief
expire) or the President’s budget proposal. The CBO baseline uses the current law concept. The
Panel provides no revenue estimates, but over the next 10 years, it has stated that its proposal
would keep revenues on average at the level the Administration proposed in its budget. The
Administration’s budget assumes that the 2001 and 2003 tax cuts will be extended and the
Administration’s other tax proposals (including expanded tax-preferred saving accounts) will be ⁴
adopted. This budget reduces revenues by $1.3 trillion (of which, extending the tax cuts accounts
for $1.1 trillion) over 10 years compared with current law. In the Administration’s budget,
revenues would rise from 16.9% to 17.7% of GDP over five years. The Administration’s budget
does not project overall revenues after five years, but CBO data indicate that revenues would
continue to slowly rise because of real bracket creep and the increase in the number of taxpayers
subject to the AMT. If the Tax Panel’s recommendations maintain revenues at roughly current
levels on average, then there would be little effect on aggregate spending now or in the future,
assuming outlays stay constant as a share of GDP. (By contrast, under a baseline based on current
law, such as the CBO baseline, revenues would rise by about 1.5% of GDP after 2011 because of
the expiration of the 2001 and 2003 tax cuts.) In other words, the tax reform proposals would lead
to the same degree of crowding out compared with current law as the President’s budget.
The Panel’s proposal would result in greater revenue loss in the long run because of its expansion
of tax preferred saving accounts (discussed below). Similar to Roth IRAs, these accounts (and
defined contribution pension plans under the GIT plan) would tax contributions to the account
deposited today but would allow tax free withdrawals of earnings and principal withdrawn in the
future. That means the accounts, sometimes referred to as “back-loaded” accounts, would raise
revenue in the short run but lose revenue when saving is withdrawn, mostly outside the 10-year
budget window. Although there is no estimate of this proposal’s long-run cost, Burman, Gale, and
Orszag estimated that an earlier Administration proposal to expand tax preferred saving accounts
(by less than the Panel proposed) could lead to an annual long-term revenue loss of 0.5% of ⁵
GDP.

In contrast to the short-term effects of fiscal policy on aggregate demand (spending), a change in
tax policy can also affect economic growth over the longer run by affecting the growth rate of
labor, the capital stock, and productivity (often referred to as the “supply side” of the economy).
Capital is financed out of national saving, so the capital stock can be increased in the long run
only if saving is increased. Changes in the growth rate of labor or saving will depend on how
much tax reform changes the incentive to work or save and how responsive labor and saving are ⁶
to changes in incentives.

⁴^{No de}^ta^ile^d^re^v^e^nue^e^s^tim^a^t^e^s^ha^v^e^be^eⁿ^re^le^a^s^e^d^{, so it is}^not^kⁿ^o^w^{n if}^the^P^a^ne^{l’s re}^c^o^m^m^e^nda^tio^{ns w}^{ould a}^lte^r
^r^e^v^e^nue^s^f^r^o^m^theⁱ^r^c^u^r^r^e^{nt pa}^t^h^oⁿ^aⁿ^aⁿ^nua^l^ba^sⁱ^s^.
⁵^L^e^ona^{rd Burm}^aⁿ^{, W}^illia^m^Ga^le^{, a}^nd^P^e^te^{r Orsz}^a^g^{, “}^T^he^A^d^mⁱ^nistra^tion’s^Sa^vⁱ^ng^{s P}^r^op^osa^l^{s: P}^r^e^lim^ina^r^y^Aⁿ^a^l^y^s^is,”
^Tax^Note^s^,^Ma^r^c^{h 3}^,²⁰⁰³^,^p.¹⁴²³^{. T}^h^e^A^d^mⁱ^nis^t^r^a^tion’^s^pr^op^os^a^l^s^m^a^y^ha^v^e^ha^{d a}^la^r^g^e^r^r^e^v^e^nue^los^s^,^h^o^w^e^ve^r^,
^b^ecau^{se t}^h^{ere w}^e^{re n}^o^rest^ri^ctⁱ^oⁿ^s^oⁿ^wⁱ^t^h^d^r^aw^al^{s so}^t^h^{e p}^a^rtⁱ^cⁱ^p^atⁱ^oⁿ^rat^e^{s w}^o^u^l^{d h}^a^{ve l}ⁱ^k^el^y^b^een^hi^gh^er.^S^{ee al}^so
^C^R^S^R^e^por^t^R^L³²²^28,^Pr^o^pos^e^d^Savⁱⁿ^gs^Ac^c^o^unts^:^Ec^on^om^ic^a^nd^Bud^g^e^t^ar^y^Effe^c^t^s^,^by^J^a^ne^G^.^G^r^av^e^lle^a^{nd Ma}^xⁱ^m
^Shv^e^dov^.
⁶^T^h^eo^retⁱ^cal^l^y^,^l^o^w^e^r^m^a^r^gⁱⁿ^alⁱⁿ^co^m^e^t^a^{xes co}^u^l^d^cau^{se w}^o^{rkers t}^o^{save an}^d^w^o^rk^m^o^{re o}^r^l^e^ss.^T^h^ey^co^u^l^d^{save an}^d
⁽^c^onti^nue^d.^..)

This is not the end of the story, however. Current fiscal policy is unsustainable over the long run.
Tax revenue is insufficient to finance spending at present, and spending under current policy is ⁷
projected to increase dramatically because of an aging population and rising medical costs.
Therefore, neither current tax rates nor the tax rates proposed by the Panel to be revenue neutral
can be considered permanent. Any effects of lower taxes today will be more than offset by higher
taxes in the future because the future revenue needs of the government will rise due to current
borrowing. Since many dynamic models of behavioral responses to a change in taxes require an
assumption of fiscal sustainability, the economic effects of current policy cannot be estimated in
these models without an explicit assumption of higher taxes or lower spending in the future.
Over long periods of time, labor and capital inputs per capita cannot be continually increased.⁸
Therefore, technological change is the main determinant of increases in income per capita over
long periods of time. Because there has not been any established connection between the U.S. tax
system and the rate of technological change, any change in tax regimes will make a small
contribution to economic output, limited to the cumulative effects of the initial changes in labor
and capital inputs.
To keep the proposal revenue neutral and compensate for the large revenue loss caused by
eliminating the AMT, the Tax Reform Panel had little scope for significant reductions in marginal
tax rates. Marginal tax rates on labor income for most taxpayers under tax reform would be
roughly similar to the current system. The Panel recommends reducing the total number of
marginal tax brackets by shifting some taxpayers into higher brackets and some into lower
brackets.
Some taxpayers would face a slightly lower marginal income tax rate under tax reform: for
example, the elimination of the 35% bracket means the highest income taxpayers will have their
marginal rates reduced by two percentage points to 33% under the SIT plan and by five
percentage points under the GIT plan. However, very few taxpayers would be affected by this
proposal—only 0.4% of all taxpayers, accounting for 7.5% of total earnings, fell under the 35% ⁹
statutory rate in 2005. Some taxpayers would also face slightly lower marginal rates under tax
reform because the phaseout of deductions and personal exemptions at high income levels under
the current system raises the effective marginal tax rate; under tax reform, the family credit does
not phase out.
On the other hand, some taxpayers would face slightly higher marginal tax rates under tax reform,
particularly under the GIT plan. For example, the elimination of the 10% bracket, which 22% of

^(...c^on^tin^ue^d)
^w^o^rk^m^o^re^sinc^e^sa^vⁱ^ng^a^{nd w}^o^rk^a^r^e^now^m^o^re^re^wa^rding^on^a^ta^k^e^-hom^e^ba^{sis (c}^a^lle^{d a}^“^s^ubstituti^{on e}^f^f^e^c^t^.”^{) But}
^the^y^c^{ould a}^l^{so sa}^v^e^a^{nd w}^o^rk^le^{ss sinc}^e^le^{ss pre}^-^ta^x^inc^o^m^e^{is ne}^e^d^e^d^t^o^e^q^ua^{l pre}^v^io^{us sta}^nda^r^d^{s o}^f^liv^ing^(c^a^lle^{d a}ⁿ
^“ⁱ^nc^om^e^ef^f^e^c^t^.”⁾^T^hus^{, the}^o^r^y^do^e^s^not^hol^{d t}^h^a^t^ta^x^c^u^ts^uⁿ^a^m^biguous^ly^r^a^is^e^e^c^onom^ic^g^r^o^w^th.
⁷^Se^e^C^R^S^R^e^por^{t R}^L³²⁷⁴⁷^,^S^o^cⁱ^a^l^Se^c^u^r^ity^aⁿ^d^M^e^dic^a^r^e^:^The^Ec^o^nom^ic^I^m^pl^ic^ati^ons^{of C}^u^r^r^e^{nt Po}^lic^y^{, by}^Ma^rc
^L^a^bonte^.
⁸^In^t^h^{e n}^e^o^c^l^a^ssi^cal^gro^w^t^h^m^o^d^e^l^,ⁱⁿ^{creases i}ⁿ^saviⁿ^g^/^capⁱ^t^a^l^h^a^{ve o}ⁿ^l^y^{a t}^e^m^p^o^r^ary^e^ffect^oⁿ^gro^w^t^h^b^ecau^{se o}^f
^dimⁱ^nis^h^ing^re^turⁿ^s^to^c^a^p^ita^l.^Ev^eⁿ^tua^lly^{, the}^c^a^p^ita^{l s}^t^oc^k^be^c^o^m^e^s^la^rg^e^e^noug^{h t}^h^a^t^aⁿ^y^f^u^rthe^{r a}^dditio^ns^ha^v^e^no
^e^f^f^e^c^t^{on o}^u^tp^ut.
⁹^C^ong^r^e^s^sⁱ^ona^{l B}^udg^e^t^O^f^fⁱ^ce^,^E^f^f^ectⁱ^{ve Ma}^rgⁱⁿ^a^l^T^a^{x R}^a^t^e^{s on}^L^a^bo^{r In}^co^me^,^N^o^v.²⁰⁰⁵^,^p^.²¹^.

taxpayers fell under in 2005, means some of these taxpayers will now face a 15% marginal rate.
For many taxpayers filing as head of household, marginal rates will be somewhat higher because
that category will be eliminated. Other individuals face slightly higher rates because the proposed
income thresholds for some brackets are higher under tax reform.
The elimination of the AMT means some taxpayers that faced the AMT’s marginal rates will now
face the regular system’s rates instead. For most affected taxpayers in 2006, this would move
them from a statutory marginal rate of 26% or 28% under the AMT to a statutory marginal rate of
25%, 30%, or 33% under the regular tax system (with the Panel’s new marginal rates). Thus, for
most affected taxpayers in 2006, elimination of the AMT would mean a slightly lower or higher
marginal tax rate, and therefore would have little effect on incentives. For some taxpayers with
large families, however, elimination of the AMT would move them from the AMT rates to the
15% tax bracket; the effect on incentives could be significant for this group. Because the AMT is
not indexed for inflation and does not allow personal exemptions, more and more taxpayers with
children in the 15% bracket would fall under it as time went by.
This analysis is based on a comparison to the AMT in its current form, according to current law.
In the past few years, Congress has repeatedly “fixed” the AMT temporarily, so that it has never
affected more than a relatively few individuals in the 15% or 25% bracket under the regular
income tax. The latest fix to the AMT expired at the beginning of 2006, and will be renewed if the
tax reconciliation bill (H.R. 4297) that is currently in conference becomes law. If Congress
continues to extend AMT reform one year at a time (as it does in the reconciliation bill), relatively
few taxpayers would fall under the AMT and see a significant drop in their marginal rates under
tax reform.
Thus, the recommendations would have an ambiguous effect on labor supply—some workers
would face slightly higher marginal taxes on labor, others would face slightly lower. In any case,
given how little labor supply has changed over the past decade and a half, as seen in Figure 1,
there is little reason to think that there would be any significant change in labor supply in
response to these small changes. Despite many changes in tax policy, including major tax
increases in 1990 and 1993 and major tax cuts in 2001 and 2003, the hours worked and
employment-population ratio of men and women have hardly changed at all over that time. The
little variation that is found in the employment-population ratio corresponds to the effects of
recessions in 1990-1991 and 2001, and is therefore unlikely to be related to changes in ¹⁰
incentives.

¹⁰^{For a}^re^vⁱ^e^w^of^the^e^m^piric^a^l^lite^ra^ture^{, se}^e^{CRS Re}^port^RL³¹⁹⁴⁹^,^Issu^{es i}ⁿ^Dyna^mi^{c R}^even^u^e^E^s^tⁱ^ma^tⁱⁿ^g^{, by}^J^a^ne
^G.^Gr^a^v^e^l^l^e^.

Figure 1. Labor Supply from 1990-2005
^Source:^Burea^{u of La}^{bor Statistics}
^Note:^{“Employ-pop (employmen}^{t-population) ratio” is measured}^{by dividing the n}^umbe^{r of em}^{ployed individuals}
^{16 yea}^{rs and older by}^{the total p}^{opulation 16 yea}^{rs and older. “Weekly hours worked” is mea}^sure^{d as the}
^{annual ave}^{rage weekly}^{hours wor}^{ked of workers in all industries.}^{There are}^{no data on hou}^{rs wor}^{ked available}
^{for 20}^05.

The Tax Reform Panel proposes to reduce the taxation of saving and capital investment for both
individuals and corporations. On the individual side, the Panel proposes expanding tax-preferred
saving accounts so that every taxpayer could, in effect, save up to $20,000 per year tax deferred
outside of their pension. This compares with a current limit on individual retirement accounts of
$4,000, although that can be supplemented in the current system by the various other accounts
available. The Panel proposed that the new accounts replace various existing retirement, health,
and medical saving accounts, but because the proposed accounts have fewer restrictions on their
use than existing accounts (e.g., annual withdrawals up to $1,000 could be made penalty free
from one type of account), individuals may find them more attractive.
If tax reform were to increase national saving, it would have a positive effect on capital
investment and economic growth. To determine whether expanding tax preferred vehicles would
raise saving, it is useful to look at the evidence on existing vehicles. Evidence on whether existing

¹¹^{See also}^{CRS Rep}^o^{rt RS}²²³⁶⁷^,^Fe^de^{ral T}^a^x^Re^{form a}ⁿ^d^{Its P}^o^te^ntial^Effe^c^t^{s o}ⁿ^Savⁱⁿ^g^,^b^y^Gr^e^g^g^A.^E^s^eⁿ^weⁱⁿ^.

tax-preferred accounts boost private saving is mixed.¹² Although there is no doubt that significant
sums are saved in these accounts, the issue is whether saving in the accounts represents new
saving that would otherwise have not taken place or existing saving that was shifted out of
nontax-favored vehicles. Account holders have higher saving rates than others, but this is not
evidence that the accounts boost saving because the account holders may be inherently higher ¹³
savers in the first place. About half of IRA-holders currently contribute the full legal limit; this
suggests that these account holders were probably already saving more than the contribution
limit, and the account holdings represent the shifting of existing saving. If this is the case, tax
preferred saving offers an incentive to save less (called an “income effect”) because less saving is
needed to generate future income because of the tax break, and additional saving cannot take
advantage of the tax preference.
If contribution limits were expanded, would saving rise? It could only potentially rise for those
whose total saving is above the current limit but below the proposed limit. For those saving above
the current legal limit, Auerbach points out that an increase of the contribution limit, which
allows existing saving to be transferred to enlarged tax-preferred accounts, represents a windfall
gain to existing saving (because capital gains are not taxed until realized) that costs the ¹⁴
government money without generating new saving. For those who are not currently saving at
the legal limit, expanding saving accounts would not increase the incentive to save since it offers
no new incentive for these savers to take advantage of. If income limits were removed, as the
Panel proposes, saving could only potentially rise for those newly eligible individuals currently
saving less than the proposed contribution limit. Those saving above the proposed limit would
have an incentive to save less, since they would now earn a higher rate of return on existing
saving, but be unable to earn a higher rate of return on new saving.
Some economists argue that tax-preferred saving accounts raise the saving rate, in part, because
withdrawal restrictions prevent people from being “tempted” into drawing down their saving, as
they might in a normal saving vehicle. If this argument is correct, then the Panel’s proposal to
allow up to $1,000 to be withdrawn annually without penalty would reduce the positive effect on
saving from this factor.
Furthermore, private investment depends on national saving, not just private saving, and the
proposal affects both. The proposed “back-loaded” saving accounts (in which contributions are
taxed and subsequent investment earnings are not) reduce tax revenue in the long run, so even if
private saving rose, that would be offset by a decline in public saving. (The overall tax reform
package is intended to be revenue neutral over 10 years, but the cost of the back-loaded saving
accounts would be substantially lower in the first 10 years than in the long run.)
On the corporate side, tax reform would reduce the taxation of business investment, which would
raise the after-tax profitability of investment. Whether this would raise overall investment levels
in the long run depends, again, on how responsive private saving is to higher rates of return. The
level of investment is determined by the equilibrium between the demand for investment

¹²^Se^e^C^R^S^R^e^por^{t R}^L³⁰²⁵⁵^,^Iⁿ^divⁱ^du^{al Re}^tir^e^m^eⁿ^{t Ac}^c^o^unts⁽^I^RAs⁾^:^I^s^s^u^e^s^an^{d Pr}^o^pos^e^d^Ex^p^ansⁱ^oⁿ^{, b}^y^Th^o^m^a^s^L.
^H^ung^e^r^f^o^r^d^a^{nd J}^a^ne^G^.^G^r^a^v^e^lle^.
¹³^Sa^ra^{h H}^o^l^d^eⁿ^,^et^a^l^,^“^T^he^I^ndivⁱ^d^u^a^l^R^e^tir^e^m^eⁿ^t^A^c^c^{ount a}^t^A^g^e^{30: A}^R^e^tr^os^pe^c^tiv^e^,^”^In^vest^m^en^t^Co^mp^an^{y In}^stⁱ^t^u^t^e
^Persp^ective^{, v}^o^{l. 11}^{, n}^o^.^1,^Fe^b.²⁰⁰⁵^{. T}^h^{is e}^s^tim^a^t^e^is^f^o^{r tra}^d^iti^on^a^l^IRA^s^{in 200}^3.
¹⁴^A^l^an^A^u^erb^ach^,^“T^h^e^T^a^{x Re}^f^o^rm^P^aⁿ^e^l^’^{s Rep}^o^r^t^:^Mⁱ^ssi^on^A^cco^m^p^lⁱ^s^h^e^d^?”^Ec^onom^is^ts^’^V^o^ic^e^,^B^E^P^r^ess,^Dec.
²⁰⁰^5.

spending and the supply of saving. Reducing taxes on business investment pushes up the demand
for investment, but, in the extreme example, if saving stayed constant there would be no change
in investment levels and the after-tax rate of return would remain the same (because the before-
tax rate of return would fall through higher borrowing costs until equilibrium was restored).
Profit-maximizing businesses may desire to save more when rates of return rise, but ultimately,
profits accrue to individuals and they could potentially reduce household saving in response to
higher corporate saving. It is possible that investment could rise, even if national saving did not
rise in response to the higher after-tax rates of return on capital, as a result of a greater inflow of
foreign capital. By identity, this would lead to a larger trade deficit, however, which could be
problematic since the trade deficit is already unprecedentedly large.
Measuring the responsiveness of saving to a change in the rate of return empirically is difficult
because saving represents a decision to postpone consumption from the present to the future.
Thus, the decision will be based not only on today’s saving rate but also future saving rates. For
that reason, the responsiveness of saving to rate of return is usually not be measured directly but
instead estimated with highly complex, stylized models. However, these models often use
assumptions that many economists have argued are unrealistic (see the next section), and these
assumptions are important to their predictions. Thus, predictions vary widely from model to
model. Using direct evidence instead is also problematic because many factors besides rate of
return determine saving, including demographics and the state of the economy, and it is difficult
to properly control for all of these variables. Moreover, as a practical matter, there is no single
“rate of return” in the economy that can be observed and compared with saving.
Figure 2. Household Saving, 1960-2005
^Source:^Burea^{u of}^{Economic An}^alysis
As simple evidence of the responsiveness of saving to capital income taxation, the household
saving rate can looked at over time. As Figure 2 shows, household saving has been declining
continuously since the 1980s, despite the steady expansion in tax preferred saving vehicles and
reduction in tax rates on capital income. Although this is not conclusive evidence that tax reform
would not raise the household saving rate (taken literally, the figure suggests it would lower the
saving rate), it certainly does not suggest that tax is the primary determinant of household saving
behavior.

Even if tax reform did not change the overall level of capital investment, it could potentially lead
to greater efficiency within the allocation of capital investment across different classes of assets.
For example, it might reduce the wedge between the taxation of nonhousing and housing assets or
between the taxation of debt and equity. Although this would increase economic efficiency, it
would probably have little effect on economic growth. To the extent that tax reform would shift
investment into more productive types of capital assets, this might lead to a one-time boost to
growth, but the boost would presumably be small since the increased investment in certain assets
would be offset by less investment in other assets.
The Panel briefly states that the SIT plan could raise economic output by up to 0.5% in 10 years
and 1.2% in the long run based on simulations using three different theoretical macroeconomic
models. For the GIT plans, the simulations predict higher output of up to 2.4% in 10 years and
4.8% of GDP in the long run. These estimates are relatively modest: for example, if the economy
were assumed to grow 3% annually as a baseline, then the economy would have grown 34% over
10 years under current policy anyway. Because the panel offers no elaboration on how these
figures were reached besides identifying the models used (they are the neoclassical growth model,
the overlapping generations (OLG) model and the Ramsey model), it is difficult to offer an
evaluation that goes into specifics. However, some general tentative observations can be offered.
It is likely that larger effects were found in the OLG and Ramsey models than in the neoclassical ¹⁵
model, so the following observations will focus on the former two.
These models are highly theoretical and have not been proven to perform well empirically. The
models are favored by some economists because they are logically consistent and tractable, not
because they have proven to be realistic; professional forecasters use a completely different type
of model. Some of the assumptions made in the models have been criticized for being extremely
unrealistic. For example, they require individuals to be able to make highly complex decisions
today that span their entire lifetime (successfully making accurate forecasts of economic variables
in order to do so), they assume individuals act systematically and rationally (and do not, say,
under-save), and they assume that labor supply changes when interest rates change. Further, the
Ramsey model assumes that people have infinite life spans, or at least all treat their descendants’
well-being as equivalent to their own.
The growth response in these models comes from assumptions made by the user about how
sensitive labor and saving are to a change in taxes. As was discussed earlier, the tax reform
proposals reduce the marginal tax rate on labor by only small amounts, so the growth response
presumably comes primarily from the reduction in marginal tax rates on saving and capital.
Presumably, the simulation assumes that saving is relatively responsive to changes in tax rates,
although the dramatic decline in saving as tax rates have fallen since 1980 suggests otherwise.
Although the Panel does not provide enough detail to be certain, their results may be reached by
modeling tax reform as a shift to a consumption tax. This would be misleading because a
consumption tax places a one-time tax on existing saving, and the Panel’s recommendation does
not. By placing a one-time tax on existing saving, which is a lump-sum tax without negative
incentive effects, consumption taxes are able to reduce other tax rates and have larger effects on

¹⁵^For^m^o^r^e^de^ta^ils^{, s}^e^e^C^R^S^R^e^p^o^r^t^R^L³¹⁹^49,^Issu^{es i}ⁿ^Dyna^mi^{c R}^e^ven^u^e^E^s^tⁱ^m^a^tⁱⁿ^g^{, by}^J^a^ne^G^.^G^r^a^v^e^lle^.

economic growth. If the changes were modeled as a shift to a consumption tax, they might also
assume that some growth would derive from a shift to a perfectly clean, broad tax base, whereas
the proposals actually move only part of the way in that direction. For these reasons and others,
Burman and Gale claim that the Panel’s estimates are too high compared with the results found in ¹⁶
other research on the economic effects of a shift to a consumption tax.
The Ramsey and OLG models also contain, to varying degrees, the assumption that people save
more when the government runs a budget deficit in anticipation of higher taxes or lower spending
in the future (because deficits must be reversed in the future). This means that the models cannot
be solved unless the modeler specifies how taxes will be raised or spending reduced to balance
the budget in the future (the Panel does identify the assumption it uses). Therefore, part of the
increase in growth in the models comes from an assumption that individuals will work and save
more now, when taxes are low, so that they can work and save less in the future, when taxes are
high. As noted above, if the tax reform proposal is compared with current law (which assumes
that the 2001 and 2003 tax cuts will expire as scheduled), then the proposal would cause the
“crowding out” of private investment and reduce growth because it increases the budget deficit.
The estimates made by the Panel presumably do not take this effect into account because they are
made compared with a baseline in which the deficit has already been increased by an assumption
that the tax cuts were made permanent.

The Panel’s reform proposal makes many changes to the tax expenditures currently found in the
income tax. Tax expenditures are defined as “revenues losses attributable to provisions of the
Federal tax laws which allow a special exclusion, exemption, or deduction from gross income or ¹⁷
which provide a special credit, a preferential rate of tax, or a deferral of liability.” One of the
Panel’s primary stated goals was to promote simplicity in the tax system, and they single out
reform of tax expenditures as a way to achieve that goal. In most cases, they propose
simplification or reduction of tax expenditures, although in some cases, such as the state and local
tax deduction, they propose elimination.
Economists generally dislike tax expenditures because they distort economic activity. For
example, allowing mortgage interest to be deducted leads to greater mortgage borrowing, at the
expense of other types of borrowing and saving. If markets are functioning properly, these
distortions reduce economic efficiency. Only if there are market failures present can the distortion
caused by a tax expenditure increase economic efficiency. Furthermore, holding government
spending constant, the revenue loss from tax expenditures leads to another set of distortions
because it must be replaced by revenue that is raised through higher marginal taxes on labor or
capital income, which distort the decision to work or save. For example, the Treasury Office of
Tax Analysis estimated that if all tax expenditures were eliminated (so that only the standard
deduction and personal exemptions remained), marginal tax rates could be lowered by one-third ¹⁸
and still raise the same amount of revenue as the current system.

¹⁶^L^e^ona^{rd Burm}^aⁿ^a^{nd W}^illia^m^Ga^le^{, “}^A^P^r^e^lim^ina^r^y^Eva^l^ua^tion^of^the^T^a^x^Re^f^o^r^m^P^a^ne^{l’s Re}^p^o^rt,”^Tax^No^te^s^,^Dec.
^5,²⁰⁰⁵^{, p. 13}⁴⁹^.
¹⁷^C^ong^r^e^s^sⁱ^ona^{l B}^udg^e^t^A^c^{t of}¹⁹⁷^4,^P.L^.^93-³⁴^4.
¹⁸^P^r^e^s^ideⁿ^{t’s A}^d^vⁱ^sory^P^a^ne^{l o}ⁿ^Fe^de^ra^{l T}^a^x^Re^f^o^r^m^,^Sim^p^le^{, Fa}^ir^{, a}^{nd Pr}^o-^G^r^ow^t^h^:^Pr^o^pos^als^to^Fix^Am^e^r^ic^a’^s^Tax
^Sy^s^t^e^m^,^No^v.²⁰⁰⁵^,^p^.⁵²^.

If the reduction in tax expenditures were coupled with marginal rate reductions, then they would
reduce the distortions to work and save, which could potentially raise economic growth
depending on how individuals responded. However, as noted above, the Panel’s proposal on a
whole does not reduce everyone’s marginal rates. The revenue raised by reducing or eliminating
expenditures is instead used primarily to compensate for the revenue lost by eliminating the
AMT. Although many economists see the elimination of the AMT as desirable, it was not a tax
with high marginal rates compared with the regular income tax. Therefore, the Panel’s proposals
regarding tax expenditures appear to have little scope for influencing economic growth.
As opposed to its popular usage, economic efficiency does not involve economic growth, wealth, ¹⁹
or productivity. In fact, evidence shows that efficiency is at loggerheads with these goals.
Generally, an outcome is economically efficient if the marginal cost of producing one more unit
of a good is equivalent to the marginal benefit of consuming one more unit of the good. When
markets function perfectly—which is defined as a market with many buyers and sellers, no
barriers to entry, perfect information, and the costs and benefits of the transaction are completely
borne by the buyer and seller—an economically efficient outcome will occur and government
intervention can only reduce efficiency. A market failure is said to exist when these criteria are ²⁰
not present. Unless a market failure is present, tax expenditures reduce efficiency by inducing
economic activity related to the expenditure at greater levels than would otherwise occur. A tax
expenditure reduces economic growth only if it discourages work, saving, or productivity.
Therefore, the Reform Panel’s proposal to reduce or eliminate many tax expenditures may
enhance economic efficiency without having much impact on economic growth.

Changes to the tax system have the potential to affect economic growth in three ways. First,
changes in overall revenues that are not offset elsewhere change the budget balance, temporarily
affecting aggregate spending in the economy. However, these changes also affect national saving,
which in turn raises or lowers interest rates and private investment. Second, changes in the
taxation of labor can influence incentives to work. Third, changes in the taxation of capital and
saving can influence the saving rate.
The President’s Tax Reform Panel has proposed to keep revenues equal to the levels proposed in
the President’s budget, which includes an extension of the 2001 and 2003 tax cuts and some
smaller new provisions. As a share of GDP, this proposal would keep revenue close to its current
levels. This implies little change in aggregate spending from current policy, and a similar amount
of crowding out of private investment as found under current policy, and more crowding out than
a current law baseline under which the tax cuts expire as scheduled. Therefore, fiscal policy
would continue to have a negative effect on economic growth because of the crowding out effects
of the budget deficit.

¹⁹^F^o^{r exam}^p^l^e,ⁱⁿ^su^ran^{ce m}^a^r^k^et^s^ma^yⁱⁿ^creas^{e e}^f^fⁱ^ci^en^c^y^b^u^t^red^u^c^{e p}^r^ecau^tⁱ^oⁿ^a^ry^saviⁿ^g^,^t^h^ereb^y^re^d^u^cⁱⁿ^{g gro}^w^t^h^.
²⁰^For^m^o^r^e^inf^o^r^m^a^{tion, s}^e^e^C^R^S^R^e^por^{t R}^L³²¹⁶^2,^The^Sⁱ^z^e^an^{d R}^o^le^of^G^o^v^e^r^nm^eⁿ^t:^Ec^o^nom^ic^I^s^s^u^e^s^{, by}^Ma^rc
^L^a^bonte^.

The Panel has proposed to make modest changes in marginal tax rates on labor, with some
taxpayers facing lower rates and others facing higher rates. This implies an ambiguous and—
given the relatively steady pattern of labor supply over time—probably insignificant effect on
labor supply.
The Panel has proposed to make larger reductions in the taxation of capital and saving, through
changes to both the corporate and individual tax systems. If saving were sensitive to changes in
after-tax rates of return, then saving could rise as a result. But empirical evidence is not clear on
how sensitive it is, and the sharp and steady decline in private saving since the 1980s—despite
the reduction in the taxation of capital and expansion of tax-preferred saving vehicles over that
time—offers prima facie evidence that private saving would be unlikely to rise significantly as a
result of tax reform.
The reason that the Panel proposes relatively modest changes in marginal tax rates on labor and
capital overall is because of the need for revenue raisers to offset the Panel’s proposal to eliminate
the alternative minimum tax (AMT). This proposal results in large revenue losses because the
number of taxpayers under the AMT is projected to increase rapidly in the next few years.
Although many economists have criticized the AMT, it has marginal rates that are similar to the
regular income tax for most affected taxpayers, so repealing it does not significantly influence
incentives to work or save in most cases.
Over long periods of time, technological change is the major determinant of income growth per
capita, and there is no established link between the U.S. tax system and the rate of technological
change.
^{Marc L}^a^bon^t^e
^Specialis^{t i}ⁿ^Macroecon^o^mⁱ^c^P^o^licy
^m^l^abon^t^e^@^c^rs^.l^oc.g^o^v^{, 7-}⁰⁶⁴⁰