Using Business Tax Cuts to Stimulate the Economy

^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

Increased interest in providing business tax cuts to stimulate the economy followed the terrorist
attacks of 2001, which heightened concerns about an economic slowdown. Among the tax
proposals discussed were a corporate rate cut and an investment subsidy. A March 2002 tax cut
contained temporary partial expensing for equipment. Interest in this issue continued, including
proposals by President Bush for reductions in taxes on corporations through temporary dividend
relief, which were enacted in May 2003. The temporary bonus depreciation expired at the end of
2004. Dividend relief was extended through 2010 in legislation passed in 2006. Temporary bonus
depreciation was also part of a recent fiscal stimulus package adopted in 2008 (P.L. 110-185).
Some economists doubt the efficacy of fiscal policy in general even when a stimulus is needed,
especially in an open economy and given the difficulties of achieving proper timing. Also, deficit
financing of a tax cut has potential negative long run effects because it crowds out investment; a
stimulus designed to increase investment spending (rather than consumption spending) would, if
successful, reduce that negative effect. Investment subsidies had largely been abandoned as
counter-cyclical devices over the last two decades, in part because of lack of evidence from
statistical studies relating investment spending to the cost of capital. Some recent empirical
evidence has found some larger effects, at least with some studies, although not enough to suggest
that all of the tax cut is spent (especially with corporate rate reductions). Moreover, the average
behavioral response identified in these studies may be larger than responses during a downturn
when many firms have excess capacities, and planning lags may make investment responses
poorly timed. Recent studies of the 2002 temporary investment stimulus tended to find it a
relatively ineffective stimulus measure.
An investment subsidy has more “bang-for-the-buck” than a corporate rate cut (or dividend
relief), since the latter benefits existing as well as new capital. A corporate rate cut is estimated to
produce as little as two-thirds of the investment induced by an investment credit with an
equivalent revenue loss. The historically most common investment subsidy is the investment
credit, although the same effect could be achieved with accelerated depreciation or partial
expensing. A temporary investment credit would be more effective than a permanent one, and a
temporary investment credit could also be made incremental. (It is not really possible to structure
a permanent incremental credit.) One disadvantage of a permanent investment credit is that it
distorts the allocation of investment and can easily produce negative tax rates. A 10% investment
credit would produce negative tax rates in excess of 100% for short-lived assets. Arguments have
been for a corporate tax rate cut because of estimated large effects on the stock market. These
calculations are overstated because they do not account for the adjustment process and of interest
rate increases. Given the uncertainty about the size of stock market effects or their beneficial
effect on the economy, there is a case for not considering stock market effects an important factor
in choosing an investment subsidy. This report will be updated to reflect major legislative
developments.

When Is Fiscal Policy Needed?.......................................................................................................2
What Fiscal Stimulus Is Most Effective?........................................................................................3
Empirical Evidence on the Effectiveness of Investment Incentives................................................4
The Optimal Design of Business Tax Subsidies..............................................................................5
Bang for the Buck.....................................................................................................................5
Type of Business Tax Cut: Corporate Rate vs. Investment Subsidy...................................6
Temporary vs. Permanent Subsidies...................................................................................7
Incremental Subsidies.........................................................................................................8
Consequences of Permanent Tax Changes on the Allocation of Capital..................................8
Effects on the Stock Market............................................................................................................9
Conclusion ..................................................................................................................................... 10
Appendix A. Measuring “Bang for the Buck”...............................................................................12
Appendix B. Measuring Marginal Effective Tax Rates.................................................................14
Author Contact Information..........................................................................................................14

ncreased interest in providing business tax cuts to stimulate the economy followed the
terrorist attacks of September 11, 2001, which heightened concerns about an economic ^th
slowdown. Among the tax proposals that were discussed in the 107 Congress were a I

corporate rate cut and an investment credit. An investment credit could be considered on either a
temporary or a permanent basis. On October 24, 2001, the House passed H.R. 3090, which
included temporary partial expensing (for three years), a provision similar to an investment credit.
The bill also contained some other provisions, including a repeal of the corporate alternative
minimum tax and an extension of net operating loss carrybacks.
Proposals developed in the Senate contained business tax cuts as well. The Finance Committee
Chairman’s plan included several tax cuts for business: allowing 10% of investments placed in
service to be expensed, an increase in expensing dollar limits for small businesses, and some
other provisions. The Finance Republican’s plan would provide 30% expensing and repeal the
corporate alternative minimum tax, but without refunding accumulated credits immediately as
was the case in H.R. 3090. The 10%, one year, expensing provision was included in the version of
H.R. 3090 that was approved by the Senate Finance Committee on November 8, 2001. This bill
did not pass the Senate; however, Majority Leader Daschle proposed a streamlined bill (using
H.R. 622 as a vehicle) that would include a 30% one year expensing provision. A final version of
H.R. 3090 approved in early March of 2002 included the House expensing provision (two years
at 30%). This provision was referred to as bonus depreciation.
Further action was taken in the 108^th Congress. At the beginning of 2003, President Bush
proposed corporate tax relief in the form of an exclusion for dividends and a step up in basis. This
proposal provides tax reductions for corporate income to the investors rather than the firm. The
effects would probably be similar to corporate rate cuts in the long run, but may have different
effects in the short term. In particular, such cuts might be more likely to be translated into
spending because they would be received directly by individuals. In H.R. 2, enacted in May 2003,
a provision allowing a 15% maximum tax rate on both dividends and capital gains was enacted;
the provision expires in 2008. This bill also increased the bonus depreciation rate to 50% and
extended it through 2004.
Bonus depreciation expired at the end of 2004. As the economy recovered, interest in short run
stimulus measures diminished. Short run stimulus again became a topic of concern in the
aftermath of Hurricane Katrina. Dividend relief was extended in legislation passed in 2006,
through 2010, although not on the grounds of short-term stimulus.
Recent problems in the mortgage markets and concerns about the economy have prompted new
interest in fiscal stimulus. The stimulus proposal recently adopted (P.L. 110-185) included two
major components: an individual income tax rebate and a temporary bonus depreciation similar to
that enacted for 2002-2004.
This report discusses issues associated with the use of business tax subsidies. First, is fiscal policy
appropriate? Second, how successful are subsidies likely to be and what form might they take to
be most effective? Finally, what other consequences might flow from the use of business tax
subsidies, especially if they are to be permanent?
Investment subsidies have typically been provided through an investment tax credit. The
investment tax credit was originally introduced in 1962 as a permanent subsidy, but it came to be
used as a counter-cyclical device. It was temporarily suspended in 1966-1967 (and restored
prematurely) as an anti-inflationary measure; it was repealed in 1969, also as an anti-inflationary

measure. The credit was reinstated in 1971, temporarily increased in 1975 and made permanent in
1976. After that time, the credit tended to be viewed as a permanent feature of the tax system. At
the same time, economists were increasingly writing about the distortions across asset types that
arose from an investment credit. The Tax Reform Act of 1986 moved toward a system that was
more neutral (within the limits of empirical estimates of depreciation) and repealed the
investment tax credit while lowering tax rates.
Investment subsidies can also be provided through accelerated depreciation, but these measures
tend not to be used for counter-cyclical purposes. At least one reason for not using accelerated
depreciation for temporary, counter-cyclical purposes is because such a revision would add
considerable complexity to the tax law if used in a temporary fashion, since different vintages of
investment would be treated differently. An investment credit, by contrast, occurs the year the
investment is made and, when repealed, only requires firms with carry-overs of unused credits to
compute credits. An exception to the problem with accounting complexities associated with
accelerated depreciation is partial expensing, that is allowing a fraction of investment to be
deducted up front and the remainder to be depreciated. This partial expensing approach also is
neutral across all assets it applies to, but the cash flow effects are more concentrated in the
present (and revenue is gained in the future). A temporary partial expensing provision, allowing

30% of investments in equipment to be expensed over the next two years, was included in H.R.

3090 in 2002 and expanded to 50% and extended through 2004 in tax legislation enacted in 2003.

It expired in 2004. The new stimulus proposal, H.R. 5240 (P.L. 110-185) also included temporary
bonus depreciation for 2008.
Historically, corporate rate changes have tended not to be used for counter-cyclical purposes. The
recent interest in the corporate tax rate cuts appears associated with arguments about the effects of
tax changes on stock markets. (A similar argument has been made for a capital gains tax cut.)
The repeal of the corporate alternative minimum tax, included in an earlier version of H.R. 3090
in 2001, is similar to a corporate tax rate in some respects, but its effects on marginal investments
are uncertain and could possibly discourage investment.

Many economists expressed uncertainty about the desirability of an additional fiscal stimulus
during the 2002-2003 period and while many economists supported a stimulus more recently in
2008, some uncertainty remained. Moreover, economists have, in general, become more skeptical
of fiscal policy as a counter-cyclical tool, particularly through the mechanism of tax cuts. Because
of lags in decision-making and administrative lags in getting tax cuts to individuals, a fiscal
stimulus enacted through a tax cut can be poorly timed. Finally, in an open economy with flexible
exchange rates some of the fiscal stimulus can be offset through a decline in net exports, as
increased interest rates attract capital from abroad and bid up the price of the dollar.
Concerns about the path of the economy led to proposals for a fiscal stimulus, including support ¹
from the Chairman of the Federal Reserve Board, as long as the stimulus occurs quickly. While
the economy is not in recession, slow growth is projected and investment incentives as well as

¹^Se^e^G^l^e^{nn Som}^m^e^r^v^ille^{, “}^B^e^r^na^nk^e^Ba^c^k^{s Fisc}^a^l^Stim^{ulus if}^Quic^k,”^{ABC News}^{, J}^a^nua^r^y^17,²⁰^08,
^http:^//a^bc^ne^w^s^.g^o.c^o^m^/^B^u^sⁱ^ne^s^s^/^wⁱ^r^e^Stor^y^?^id=⁴¹⁴⁸⁶⁴^8.

individual tax rebates were under consideration. The final proposal included temporary bonus
depreciation for 2008.

The objective of a fiscal stimulus is to increase spending, and fiscal policies can differ in the
extent to which they induce spending. While a fiscal stimulus delivered through direct spending
has a relatively straightforward effect, a fiscal stimulus delivered via a personal tax cut tends to
have a more muted effect on the economy, because only part of it will be spent. The smaller the
share spent, the smaller the stimulus, although for most types of tax cuts, the presumption is that
much of the tax reduction will be spent. There are, however, theoretical reasons to believe that
higher income individuals will spend a smaller fraction of a tax cut, and empirical evidence to ²
support that view. Indeed the propensity to save makes capital gains tax cuts, which have been ³
under discussion, questionable candidates for stimulating aggregate demand. There is also a fear
that consumers who feel uncertainty about the future will not spend a tax cut.
Note that there is a tension between short run and long run fiscal policy. Measures that increase
consumption are expansionary in the short run, but may detract from growth in the long run
because deficit finance causes aggregate savings to fall (unless the economy is at such a low rate
of employment that the stimulus induces sufficient output to offset the loss in savings). That is,
government spending and tax reductions financed by deficits tend to crowd out investment.
However, if the policy could be in a form that would stimulate investment spending, this negative
effect on long run growth might be avoided. Government investment spending, such as spending
on infrastructure, is one possibility that could provide a short run stimulus without detracting
much from long term growth (and can even enhance long-term growth if the productivity of the
government investment is greater than the productivity of private investments). However, it is
often difficult to enact and implement such spending in a timely fashion. Another policy aimed at
expanding investment is a subsidy to private investment spending: if the revenue loss (which adds
to the deficit) were spent on investment, there would be no crowding out. There is, however, a
caveat: if the type of business subsidy is one that is permanent and not neutral in the long run, the
economy will also sustain a permanent loss in efficiency from the mis-allocation of capital.
The success of such a policy hinges on the effectiveness of investment subsidies in inducing
spending. It is difficult to determine the effect of a business tax cut, and particularly the timing of
induced investment. A business tax cut is aimed at stimulating investment largely through
changes in the cost (or price) of capital. If there is little marginal stimulus or if investment is not
responsive to these price effects in the short run, then most of the cut may be saved: either used to
pay down debt or paid out in dividends. Some of the latter might eventually be spent after a lag.
That is, if a tax cut simply involved a cash payment to a firm, most of it might be saved,
particularly in the short run. Business tax cuts (of most types) also have effects on rates of return
that increase the incentive to invest, and it is generally for that reason that investment incentives
have been considered as counter-cyclical devices.

²^A^re^c^e^{nt study}^tha^t^fⁱ^{nds e}^v^ideⁿ^c^e^of^hig^h^e^r^m^a^r^g^ina^l^pro^p^eⁿ^sitie^s^{to sa}^v^e^a^m^ong^w^e^a^lthy^indivⁱ^dua^l^s^{is Jona}^thaⁿ
^Mc^Ca^rthy^{, “}^I^m^p^e^r^f^e^c^t^Insura^nc^e^a^{nd Dif}^f^e^ring^P^r^ope^nsitie^{s to}^Coⁿ^s^um^e^Ac^{ross Indiv}ⁱ^d^u^a^l^s,”^J^ourⁿ^a^{l o}^f^M^one^t^a^r^y
^Ec^onom^ic^s^,^V^o^l.³⁶^{, N}^o^.^{2 (}^N^ov^e^m^be^r¹⁹⁹⁵⁾^.
³^Se^e^C^R^S^R^e^por^{t R}^S²¹⁰¹^4,^Ec^oⁿ^o^mⁱ^c^an^{d Re}^v^e^nue^Effe^c^t^s^of^Pe^r^m^ane^nt^an^{d Te}^m^por^ar^y^C^apit^a^{l G}^a^ins^T^a^x^C^u^ts^{, by}
^J^aⁿ^e^G.^Gr^a^v^e^l^l^e^.

Why, then, might a business tax cut be considered, and how might alternative tax cuts be
evaluated? First, we consider the empirical evidence which might determine whether a business
subsidy should be considered at all. Then we discuss the “bang for the buck,” which determines
how much of each dollar of cost might be spent. Then we discuss other advantages or
disadvantages

Despite attempts to analyze the effect of the investment tax credit, considerable uncertainty
remains. Time series studies of aggregate investment using factors such as the tax credit (or other
elements that affect the tax burden on capital or the “price” of capital) as explanatory variables ⁴
tended to find little or no relationship. A number of criticisms could be made of this type of
analysis, among them the possibility that tax subsidies and other interventions to encourage
investment were made during periods of economic slowdown. A recent study using micro data
found an elasticity (the percentage change in investment divided by the percentage change in the ⁵
user cost of capital) for equipment of -0.25. A widely cited study by Cummins, Hassett, and
Hubbard used panel data and tax reforms as “natural experiments” and found effects that suggest ⁶
a price elasticity of -0.66 for equipment.
This last estimate is a much higher estimate than had previously been found and reflects some
important advances in statistical identification of the response. Yet, it is not at all clear that this
elasticity would apply to stimulating investment in the aggregate during a downturn when firms
have excess capacity. That is, firms may have a larger response on average to changes in the cost
of capital during normal times or times of high growth, when they are not in excess capacity.
Certainly, one might expect the response to be smaller in low growth periods.
An additional problem is that the timing of the investment stimulus may be too slow to stimulate
investment at the right time. If it takes an extensive period of time to actually plan and make an
investment, then the stimulus will not occur very fast compared to a cut in personal taxes that
stimulates consumption. Indeed, the stimulus to investment could even occur during the recovery
when it is actually undesirable.
There is some evidence that the temporary bonus depreciation enacted in 2002 had little or no
effect on business investment. A study of the effect of temporary expensing by Cohen and ⁷
Cummins at the Federal Reserve Board found little evidence support for a significant effect.
They suggest several potential reasons for a small effect. One possibility is that firms without

⁴^A^su^m^m^a^r^y^of^{this e}^a^r^l^y^lite^ra^ture^c^aⁿ^be^f^oun^{d in}^se^v^e^ra^{l sourc}^e^s^{. For a}^no^n-te^c^hⁿⁱ^c^a^l^sum^m^a^r^y^,^se^e^Ja^ne^G^.
^G^r^a^v^e^lle^,^The^Ec^onom^ic^Effe^c^t^s^o^f^Taxⁱ^{ng C}^api^tal^Iⁿ^c^o^m^e^,^C^a^m^b^rⁱ^dg^e^,^MI^T^P^r^e^s^s^,¹⁹^94,^p^p^.¹¹^8-¹²⁰^.
⁵^Ro^b^e^rt^S^.^C^hⁱ^rⁱⁿ^ko,^S^t^even^M.^F^azzarri^,^an^d^Aⁿ^d^r^ew^P^.^M^e^y^e^r.^“^H^o^w^Res^pons^iv^e^is^B^u^sⁱ^ne^s^s^C^a^p^ita^{l For}^m^a^{tion t}^o^its
^{User Co}^st^{? A}ⁿ^E^x^p^l^o^r^atⁱ^o^{n w}ⁱ^t^h^Mⁱ^cro^Dat^a?”^J^ourⁿ^a^{l o}^f^P^ublic^E^c^onom^ic^s^v^o^l.⁷⁴⁽¹⁹⁹⁹⁾^,^pp^{. 5}³^-⁸^0.
⁶^J^a^s^{on G}^.^C^u^m^m^ins^,^K^e^vⁱ^{n A}^.^H^a^s^s^e^{tt, a}^{nd R}^.^G^l^eⁿ^H^ubba^r^d^{, “}^A^R^e^c^ons^ide^r^a^tio^{n of}^B^e^ha^vⁱ^or^U^s^ing^T^a^x^R^e^f^o^r^m^s^as
^Na^tura^{l Ex}^pe^rim^e^nts.”^Br^ook^ings^Pape^r^s^oⁿ^Ec^o^nom^ic^Ac^tiv^ity^{, 1}⁹⁹⁴^{, n}^o^.^1,^pp^{. 1-}⁷²^.
⁷^D^a^r^r^y^l^C^ohe^{n a}ⁿ^{d J}^a^s^oⁿ^C^u^m^m^ins^,^{A Re}^tr^os^pe^c^tiv^e^Ev^al^ua^tio^{n o}^f^t^h^e^Effe^c^t^s^of^Te^m^por^ar^y^Par^tⁱ^a^{l E}^x^pe^ns^ing^,
^Fina^nc^e^aⁿ^{d Ec}^o^nom^ic^s^Dⁱ^s^c^us^sⁱ^on^Se^rⁱ^e^s²⁰^06-^19,^Fe^de^r^a^{l R}^e^s^e^r^v^e^B^o^a^r^{d, W}^a^s^h^ing^t^on,^D^.^C^.^A^p^rⁱ^{l 20}⁰⁶^{. T}^h^e^y
^co^m^p^aredⁱⁿ^vest^m^eⁿ^tⁱⁿ^creases^f^o^{r sh}^o^r^t^e^{r l}ⁱ^ved^an^d^l^oⁿ^g^{er l}ⁱ^v^ed^asset^s^(l^oⁿ^g^{er l}ⁱ^v^ed^asset^s^recei^v^e^d^{a l}^a^{rger i}ⁿ^cen^tⁱ^v^e)
^an^dⁱⁿ^vest^m^eⁿ^t^cl^o^s^{er t}^o^expⁱ^ratⁱ^on^t^o^t^e^st^t^h^{e ef}^f^ect^s.

taxable income could not benefit from the timing advantage. In a Treasury study, Knittel
confirmed that firms did not elect bonus depreciation for about 40% of eligible investment, and
speculated that the existence of losses and loss carry-overs may have made the investment
subsidy ineffective for many firms, although there were clearly some firms that were profitable ⁸
that did not use the provision. Cohen and Cummins also suggested that the incentive effect was
quite small (largely because depreciation already occurs relatively quickly for most equipment),
reducing the user cost of capital by only about 3%, that planning periods may be too long to
adjust investment across time, and that adjustment costs outweighed the effect of bonus
depreciation. Knittel also suggests that firms may have found the provision costly to comply with,
particularly because most states did not allow bonus depreciation.
A study by House and Shapiro found a more pronounced response to bonus depreciation, given
the magnitude of the incentive, but found the overall effect on the economy was small, which in ⁹
part is due to the limited category of investment affected and the small size of the incentive.
Their differences with the Cohen and Cummins study reflect in part uncertainties about when
expectations are formed and when the incentive effects occur.
Cohen and Cummins also report the results of several surveys of firms, where from 2/3 to over
90% of respondents indicated bonus depreciation had no effect on the timing of investment
spending.
Overall, bonus depreciation did not appear to be very effective in providing short-term economic
stimulus. It is possible, however, that a stimulus during current times, when losses are not as large
as they were in 2002-2004 when the economy was already in a recession, could be more
successful.

In this section we consider several issues surrounding the optimal design of business tax
subsidies.
It follows from a principal rationale for choosing an investment subsidy (to prevent longer run
loss of productivity due to deficit finance) that one would seek the most powerful incentive per
dollar of revenue loss. Three issues arise in evaluating the ratio of induced spending to revenue
loss: the general type of tax incentive, whether one can increase the incentive per dollar of
revenue loss by making the subsidy temporary , and whether one can be successful by restricting
the subsidy to marginal investment.

⁸^Ma^tthe^w^Knitte^l,^C^o^r^por^ate^Re^s^pons^e^t^o^B^oⁿ^u^s^D^e^pr^e^cⁱ^ati^on:^Bo^nus^D^e^pr^e^cⁱ^ati^oⁿ^for^Tax^Ye^ar^s²⁰⁰²^-²⁰⁰⁴^{, U}^.^S^.
^D^e^pa^r^t^m^e^{nt of}^T^r^e^a^s^u^r^y^{, O}^f^fⁱ^ce^of^T^a^x^Aⁿ^a^l^y^s^is^W^o^r^k^ing^P^a^pe^r⁹⁸^,^Ma^y^2007.
⁹^C^h^rⁱ^s^t^o^phe^r^H^ous^e^aⁿ^{d Ma}^tt^he^w^Sha^p^ir^o,^T^e^mpo^r^a^r^{y I}ⁿ^vest^m^en^t^T^a^{x In}^cen^tⁱ^{ves: T}^h^e^o^{ry Wi}^t^h^E^vⁱ^d^en^{ce f}^r^o^m^B^oⁿ^u^s
^Dep^r^eci^a^tⁱ^oⁿ^,^N^a^tio^na^{l B}^u^r^e^a^u^of^Ec^on^om^ic^R^e^s^e^a^r^c^h^W^o^r^k^ing^P^a^pe^r¹²⁵¹^{4, C}^a^m^b^rⁱ^dg^e^,^Ma^s^s^{., Se}^pte^m^be^r²⁰⁰⁶^.

Generally, an investment credit (or other subsidy confined to investment, such as accelerated
depreciation) has more “bang for the buck” than a corporate rate cut (or dividend relief) because
it does not reduce taxes on the flow of income to existing capital assets. The size of both the
absolute amount and the difference between an investment credit and a corporate rate cut depend
on the durability of the investment. As derived in the appendix, a corporate rate cut has an
effectiveness compared to an investment credit, given economic depreciation, based on the ratio:
(g+d)
(r+d)
where g is the normal growth rate, r is the after tax rate of return, d is the economic depreciation
rate, and u is the corporate tax rate. Setting g, r, and u to 0.025, 0.05, and 0.35 respectively, this
measure suggests a 69% ratio for structures (assuming d equals .03) and a 88% ratio for
equipment (assuming an average depreciation rate of 0.15). As d gets very large, the value
approaches 100% and as it gets very small, the value approaches 50%. Thus, a corporate rate cut
will stimulate from almost ⅔ to less than 90% as much investment per dollar of revenue loss as
an investment subsidy directed at the same type of investment. Note also that the superior
performance of the investment credit relative to the corporate rate cut is less pronounced for short
lived assets. This effect occurs because investment is larger relative to the existing capital stock
because of the need to replace the stock more frequently. Corporate assets also include non-
reproducible capital (e.g. land) that receives a tax reduction with no investment increase, which
also reduces the stimulus per dollar of revenue loss. The relative size of the stimulus also depends
on the ratio of growth to return and on pre-existing subsidies; at the extreme, with expensing, the
corporate rate cut will have no effect.
At the same time, short lived assets may have a somewhat larger absolute “bang for the buck,” for
an investment credit since the size of the stimulus per dollar of revenue loss is e/(1-uz), where e is
the investment demand elasticity and z is the present value of tax depreciation (which is larger for
short lived assets). For the typical equipment investment, if e is -0.25, each dollar of revenue loss
from an investment credit produces 35 cents of investment. For structures (assuming the same
elasticity), the increase is 28 cents. Even at the high elasticities estimated at -0.66 these increases
would not be dollar for dollar: the equipment investment would increase by 92 cents and the
structures by 74 cents. (Note, however, that the elasticities could vary across assets, and in
particular could be smaller for structures.)
Another aspect of an investment credit vs. a corporate rate cut is the degree of certainty about the
subsidy. If businesses fear that lower corporate rates will subsequently be raised, they will have
less of an incentive to invest. Indeed, a perception that corporate rates could be increased could
actually have negative effects on investment (as discussed below). Investment credits, however,
are allowed at the time of the investment and are certain, since tax benefits would be highly
unlikely to be retroactively disallowed.
Some other types of corporate tax changes can be likened more to an investment subsidy or to a
rate reduction. An acceleration of depreciation (allowing costs of investments to be deducted
more quickly), or allowing expensing (deducting the entire cost when the expenditure is made)

for some fraction of investments is like an investment credit. A repeal of the alternative minimum ¹⁰
tax provides a benefit for existing assets, but mixed effects on investment, since the marginal
tax burden on investments under the minimum tax can be greater than or less than the burden
under the regular tax. For firms permanently on the minimum tax, the tax burden on new
investment is actually smaller than the tax burden under the regular tax, so that repealing the
minimum tax would actually discourage investment in this case. A modification of the minimum
tax by allowing accelerated depreciation methods (lives are already equated or virtually equated)
would be like an investment subsidy for firms that remain under the alternative minimum tax but
could have mixed effects if the change caused firms to switch to the regular tax. It would be less
likely to discourage investment than a repeal of the tax. Expanding the net operating loss carry-
back periods (or investment tax credit carry-back periods if such a credit were enacted) has a
large cash flow effect which is like a corporate rate cut, but it would also allow more firms to
benefit more fully from investment subsidies and existing accelerated depreciation.
Note that dividend relief is similar to a rate cut in that it affects the return to current investment. If
provided to individuals in the form of an exclusion or lower rate rather than to the firm as a
deduction, the incentive may be lessened since there is no direct and immediate effect on the firm.
Capital gains relief for individuals is even less effective because provides benefits for income
accumulated in the past as well as current income.
Since investment subsidies act through changes in price, there have been attempts to increase the
“bang for the buck.” Two methods have been proposed (and could be combined). The first is to
make the investment tax credit temporary. Theoretically, a temporary investment subsidy, like the
investment credit, would have a more pronounced effect on investment in the short run than a
permanent one, and, of course, would cost much less. Like a temporary sale, demand should shift
and firms should move planned investment spending forward. It is, however, very hard to find
good empirical evidence of this effect, in part because the same problems that have plagued
earlier empirical studies remain, among them the fact that temporary subsidies have been enacted
during a downturn. And, assuming that investment is only shifted, the crowding out issue still
remains.
Another of the difficulties with temporary subsidies is that in order for them to have a more
powerful effect that permanent subsidies, investors have to believe that such subsidies will,
indeed, be temporary. Although the bonus depreciation enacted in 2002, was allowed to expire, it
was extended and expanded in 2003, and historical experience teaches otherwise: temporary
subsidies have a tendency to become permanent.
As noted above, the empirical evidence suggests that even a temporary subsidy was not very
effective as an economic stimulus, although the reasons for that are not entirely clear and studies
face many difficulties. It is possible that such a stimulus would be more effective currently
without the overhang of losses from a recession.
Note that a temporary corporate rate cut will have the opposite effect relative to a permanent rate
cut: it will have little effect on new investment. In fact, a temporary rate cut could discourage

¹⁰^T^h^e^a^lte^rna^tiv^e^mⁱ^nim^u^m^ta^x^{is a}^ta^x^s^y^ste^m^tha^tⁱ^m^pose^s^aⁿ^a^lte^rn^a^tiv^e^ta^x^w^{ith a}^low^e^{r ra}^te^a^{nd a}^b^r^oa^de^{r ba}^se^tha^t
^m^u^{st be}^pa^{id if}^hig^h^e^r^t^h^aⁿ^t^h^e^re^g^u^la^{r ra}^te^.

certain types of investment because, due to accelerated depreciation, deductions are larger in the
early part of an investment’s life than should be the case to reflect economic depreciation, while ¹¹
they are too small in the later part.
In the past, policy-makers have also looked into the possibility of incremental investment
subsidies that would focus more of the effect on the margin. Such a subsidy was discussed at the
beginning of President Clinton’s administration.
A subsidy could be more focused on the margin by applying it only to investment in excess of a
fixed base. While a temporary incremental subsidy is feasible, it is virtually impossible to design
a permanent incremental subsidy. This issue is a complicated one which is discussed in ¹²
considerable detail in a CRS study written at that time.
A final issue in choosing a subsidy is the potential effect on the allocation of capital. A lower
corporate tax rate may (up to a point) increase economic efficiency because currently corporate
income is taxed more heavily than other types of capital income. A corporate rate is also neutral
across different types of assets. Moreover, a corporate tax rate reduction cannot go to the point
that investments are subject to negative tax rates.
An investment credit, however, has the potential for distorting investment (because it favors short
lived assets and is generally applied only to certain categories of investment). It can also easily
produce negative tax rates. Moreover, with the present value of depreciation so high because of
low inflation rates, and the tax rate much lower than in the past, historical levels of investment
credits will produce negative tax rates.
Consider five-year property, which accounts for about 44% of investment. Depending on the time
of year the investment is made, we estimate the present value of depreciation deductions
assuming a 7% nominal discount rate (reflecting a 5% real return and a 2% inflation rate) is
between 0.86 and 0.89 per dollar of investment, depending on the time of year the investment is
made. To avoid negative tax rates any investment credit that does not have a basis adjustment
(that is, depreciation is allowed on the entire investment not just the cost net of the credit) cannot
exceed 3.85% at the higher present value and 4.9% at the lower value. With a basis adjustment,
the credit can range between 5.59% and 7%. The zero tax rate is reached when the present value
of tax depreciation deductions multiplied by the tax rate plus the credit exceed the tax rate times ¹³
the investment.

¹¹^A^pr^opos^a^l^t^o^e^x^pl^{oit t}^h^is^e^f^f^e^c^t^by^eⁿ^a^c^ting^a^f^u^tur^e^c^o^r^p^or^a^t^e^r^a^t^e^c^u^{t c}^oul^{d w}^o^r^k^{in t}^h^e^o^r^y^{, but}^is^unlik^e^l^y^{to be}
^v^e^r^y^ef^f^e^c^tiv^eⁱ^f^fⁱ^r^m^s^doubt^tha^t^t^h^e^r^a^te^c^u^{t w}^ould^a^c^t^ua^lly^ta^k^e^pl^ace.^It^w^o^u^l^d^al^soⁿ^o^t^p^r^o^vⁱ^d^{e f}ⁱ^{rms w}ⁱ^t^h^cash^f^l^o^w
^t^o^m^a^k^eⁱⁿ^creasedⁱⁿ^vest^m^eⁿ^t^s.
¹²^{See CRS Rep}^o^r^{t 9}³^-2⁰⁹^,^In^cremen^t^a^l^Iⁿ^vest^m^en^t^Su^b^sⁱ^dⁱ^e^s^{, by}^J^a^ne^G^.^G^r^a^v^e^lle^.
¹³^T^h^{ese est}ⁱ^m^a^t^e^s^m^e^a^s^u^r^{e effect}ⁱ^v^{e t}^a^{x rat}^e^{s at}^t^h^{e f}ⁱ^r^m^l^e^vel^.^Rat^e^s^w^o^u^l^dⁱⁿ^{crease b}^ecau^{se o}^f^t^h^{e i}^ndi^vi^d^u^a^l^l^e^vel^t^a^x
^b^y^sh^areh^o^l^d^e^rs,^bu^t^w^o^u^l^d^d^{ecrease b}^ecau^{se o}^f^t^h^{e d}^e^d^u^c^tⁱ^oⁿ^o^f^t^h^{e i}ⁿ^f^l^atⁱ^oⁿ^p^o^r^tⁱ^oⁿ^o^fⁱⁿ^t^e^rest^b^y^{a f}ⁱ^r^m^wⁱ^t^h^{a h}ⁱ^gh^er
^ta^x^r^a^te^s^.^T^h^e^ne^{t e}^f^f^e^c^t^s^de^pe^{nd i}ⁿ^pa^r^t^on^w^h^e^t^he^r^inv^e^s^t^m^e^nts^m^a^de^thr^oug^h^pe^nsⁱ^{on f}^uⁿ^d^s^aⁿ^{d I}^R^A^s^a^r^e^c^ons^ide^r^e^d
⁽^c^onti^nue^d.^..)

If a 10% investment credit without a basis adjustment were enacted, these assets would have a
marginal effective tax rate of -122% to -196%, that is, a negative effective tax rate. (The
computation of effective tax rates is explained in the appendix.)
Accelerated depreciation methods can be designed to be more neutral, and there are several types
of investment subsidies that are relatively neutral at least across the assets they apply to
(including partial expensing, allowing credits only for investment in excess of depreciation, or
varying credits with asset durability). However, the largest distortion that has typically occurred is
between assets eligible for credits or accelerated depreciation and assets that are not eligible,
primarily equipment in the former case and structures in the latter case. (Partial expensing in H.R.

3090 is restricted to equipment but is not permanent.)

Some arguments have been made that the effect of a business tax change might help the economy
by raising the stock market price. It is not clear what consequences a rise in the stock market
induced by tax changes might have as an independent influence on the economy, although such a ¹⁴
rise might help to maintain consumer confidence in the economic outlook.
Some simple calculations of this effect have been discussed based on comparing tax rates. For
example, cutting the corporate tax rate from 35% to 25% (a very large cut), which would lead to
an increase of around 15% based on the ratio of (1-t*)/(1-t), where t* is the new tax rate of 0.25
and t is the old tax rate of 0.35. This calculation is based on a rational expectations view of the
world (rather than any empirical measures of the supply and demand responses in the stock
market that some economists would prefer to use). Even so, this analysis is a partial equilibrium
one that does not take account of three important effects: the expected adjustment of the capital
stock in response to change, adjustment costs of such changes, and the possibility of higher
interest rates. We discuss these in turn.
Consider a simple adjustment process where a geometric adjustment path occurs. In that case,
beginning at time zero, the rate of return t years into the future given a fixed after tax rate of
return r will be:
p(t) = r/(1-t*) +[ r/(1-t)-r/(1-t*)]e^-g^t
and the after tax rate of return be:
r(t) = p(t)(1-t*)
If this expression is integrated (summed), after being discounted at rate r, from time zero to
infinity (the same approach that produces the original 15% price increase), the increase in the

^(...c^on^tin^ue^d)
^m^a^r^g^ina^l^{. For}^uⁿⁱⁿ^c^o^r^p^or^a^t^e^d^bus^ine^s^s^e^s^,^r^a^te^s^c^ould^be^hig^h^e^r^or^l^o^w^e^r⁽^a^bs^olute^v^a^lue^of^the^ne^g^a^ti^v^e^r^a^te^s^low^e^r^or
^hig^h^e^r⁾^de^pe^ndi^ng^on^the^m^a^r^g^ina^l^ta^x^r^a^te^of^the^indⁱ^vⁱ^dua^l;^the^y^w^oul^{d pr}^oba^bly^be^low^e^r^,^tha^t^is^{, h}^a^v^e^a^hig^h^e^r
^subsidy^ra^te^.
¹⁴^For^a^dis^c^us^sⁱ^oⁿ^f^o^c^u^s^e^{d s}^p^e^cⁱ^fⁱ^c^a^lly^{on the}^divⁱ^de^{nd r}^e^lie^f^pr^o^pos^a^l^{, s}^e^e^C^R^S^R^e^por^{t R}^L³¹⁸^24,^Divⁱ^de^nd^Tax^Re^lie^f:
^E^f^f^ect^{s o}ⁿ^E^c^oⁿ^o^mⁱ^c^R^eco^very,^L^oⁿ^g^-T^{erm Gro}^w^t^h^,^an^d^t^h^e^S^t^o^ck^Ma^rket^{, by}^J^a^ne^G^.^G^r^a^v^e^lle^.

stock market price will be [(1-t*)/(1-t)][r/(r+g)]. If g is, say, equal to r, then the initial effect on
the stock market will be only half as big. This value will also begin to fall as time goes on until it ¹⁵
returns to its original value.
Secondly, the value will be reduced if the firm faces adjustment (temporary loss of productivity
due to alteration of the productive processes).
How quickly this adjustment path occurs and the magnitude of adjustment costs is something we
know relatively little about in an empirical sense, but it will clearly reduce, perhaps significantly,
the initial increase in value.
The second problem with the analysis of stock market effects is that it not only assumes no
adjustment costs but assumes there are no other effects in the economy that might alter the rate at
which earnings are discounted. But the interest rate will almost certainly rise. Indeed, in a simple
world, with only one asset and a fixed savings rate (even ignoring the resource claims on the
deficit by assuming the revenue is made up elsewhere), a tax cut simply increases the after tax
rate of return, and raises the discount rate just enough to offset the rate cut. The result is that there
would be no effect on the stock market. This criticism is perhaps a more serious one which could
greatly reduce any potential effect on the stock market. But, without a general equilibrium
analysis, such an effect cannot be estimated.
Indeed, one could argue that the future tax cuts are causing the stock market values to decrease by ¹⁶
increasing future interest rates (and also discouraging investment for the same reason).
Note that, according to this type of analysis, the effect of an investment credit on the stock market
can either raise or lower stock prices. While profits rise in the short run, in the long run the value
of financial assets relative to the physical value of the firms assets falls. If the investment credit
were 10%, for example, the asset value would fall by 10%. With two opposing forces in play, the
initial effect on the stock market is uncertain according to rational expectations theory, which may
be one of the reasons that some analysts have proposed a rate cut rather than an investment credit.
For those economists who doubt that a rational expectations approach is the best method of
predicting the effects on the stock market in the short run, however, the inclination would be to
expect an initial rise in the stock market from either policy as more investors choose to buy
stocks. The magnitude of these effects is, however, very difficult to determine, but is likely not to
be very large relative to the other short run swings in the market. Such a view would suggest that
considerations of the stock market not play an important rule in evaluating alternative counter-
cyclical tax instruments.

This analysis suggests that fiscal stimulus may not necessarily be desirable currently and that a
business tax subsidy may not necessarily be the best choice for fiscal stimulus, largely because of

¹⁵^W^{ith im}^pe^rfe^c^t^c^o^m^p^e^tition,^the^v^a^lue^mⁱ^g^h^{t se}^ttle^a^t^a^dif^f^e^r^eⁿ^{t le}^v^e^{l, de}^pe^ndi^ng^oⁿ^t^h^e^na^ture^of^th^e^ind^u^stry^a^nd
^h^o^w^eq^u^ili^b^r^iu^m^{is reach}^ed^{; th}^ese^co^m^p^lex^{ities are b}^e^y^{ond the}^s^c^op^e^of^this^pa^pe^r^.
¹⁶^T^h^{is la}^tte^{r poi}^{nt w}^a^{s m}^a^de^by^Christop^he^{r Ca}^rroll,^“^D^on’t^Cut^Ca^p^ita^{l G}^a^{ins T}^a^x^,^”^Baltimore^S^uⁿ^,^S^e^pte^m^be^r^25,
²⁰⁰^1.

the uncertainty of its success in stimulating aggregate demand. If such subsidies are used,
however, the most effective short run policy is probably a temporary investment credit.
Permanent investment credits, while more effective than corporate rate cuts in the short run, will
distort the allocation of investment in the long run.

The rental price of capital, with an investment credit (without a basis adjustment) is:
(1) ()()()crduzku=+−−−11
where c is the price of capital, r is the after tax rate of return to corporate investment, u is the
corporate tax rate, d is the economic depreciation rate, z is the present value of depreciation
deductions and k is the investment tax credit. If we take logs and differentiate this expression
with respect to k, assuming an initial rate of the credit of zero, we obtain the following expression
for the percentage change in the cost of capital:
−1
(2) ()dccuzdk/=−1
We denote the elasticity of investment with respect to the price of capital as
(3) ()()dIIedcc//=−,
where e is the elasticity and I is investment. By substituting (2) into (3) and rearranging, we
obtain:
eI
(4) ()didk=
uz−1
We can also relate this incentive to the revenue cost, which is −kI, hence the ratio of stimulus
per dollar of revenue loss is ()euz/1−. This stimulus is bigger for shorter lived assets
(assuming the elasticities are the same).
Compare the effect of a rate reduction. Differentiating (1) with respect to u we obtain, dividing by
c and substituting into (4), we obtain the effect on investment of a rate reduction:
eI z−−()1
(5) dIuuzdu=−−()()11
How does this compare with revenue? The only way to make a fair comparison between an
investment credit that appears only in the first year and applies to new investment, and a rate
reduction that lasts over the entire life of the investment but also benefits existing capital is to
compare investment per present value of revenue cost. The present value of the cost of an
investment credit if investment grows at rate g and is discounted at rate r is ()−−Idkrg/. The
effect on today’s investment per present value of revenue loss is therefore ()()erguz/()−−1.
Taxable income is the rental price of capital c (the marginal product of capital) times the capital
stock minus the flow of depreciation. The flow of depreciation per unit of investment today, in a
steady state, is the same as the present value of depreciation formula discounted at the growth rate
rather than the rate of return. Thus the cost of a rate reductions is du (cK - Izg) where zg is the

depreciation formula discounted at the growth rate g. Since I equals ()dgK+ the effect on
today’s investment per present value of revenue loss is:
() ( )()e g d z r g uz r d uz g d zg u+−− − +−−+ −()()/( )()( )( )(11 1 1
Therefore the ratio of the cost of an investment credit to the cost of a corporate rate reduction is
somewhat complicated, but when depreciation is economic depreciation, the ratio simplifies to:
()CoIC gd+
(5) ()=
CoCRCrd+
[where CoIC = Cost of Investment Credit, CoCRC = Cost of Corporate Rate Cut].

A marginal effective tax rate is determined by a discounted cash flow analysis, where the internal
rate of return with and without taxes is compared. This type of measure can take into account all
of the timing effects which are the crucial features of certain tax preferences, including
accelerated depreciation and deferral of taxes on capital gains until realization.
In the case of a depreciating asset, the relationship between pre-tax return and after tax return in
the corporate sector is determined by the rental price formula:
(6) ()()()()rrsuzmkud^pf=+−−−−111/
where r^pis the pre-tax real return, r^fis the after tax discount rate of the firm, d is the economic
depreciation rate, u is the statutory tax rate of the firm (equal to the corporate tax rate for
corporate production and equal to the individual tax rate for non-corporate production), z is the
present value of depreciation deductions for tax purposes, k is the investment tax credit rate, and
m is the fraction of k that reduces the basis for depreciation purposes. The value of depreciation is
discounted at the nominal discount rate,r^f⁺^π , where π is the rate of inflation. This formula
applies to investments in equipment and structures which are subject to depreciation. The
effective tax rate is measured as ()rrr^pf^p−/.
^J^aⁿ^e^{G. Grav}^elle
^Senⁱ^o^r^Sp^{ecialist i}ⁿ^Ecoⁿ^o^m^ic^P^o^licy
^j^g^rav^e^l^l^e^@^c^rs^.l^oc.^g^ov^{, 7-}⁷⁸²⁹