Copyright (c) 1995 Tax Analysts
Tax Notes
MAY 29, 1995
LENGTH: 9023 words
DEPARTMENT: Special Report (SPR)
CITE: 67 Tax Notes 1229
HEADLINE: 67 Tax Notes 1229 - THREE ERRORS IN THE 'NEUTRAL COST RECOVERY SYSTEM' PROPOSAL.
AUTHOR: Johnson, Calvin H.
SUMMARY:
Calvin H. Johnson is a professor of law at the University of Texas.
The neutral cost recovery system in the Contract With America is intended to be the equivalent of 'expensing,' that is, deducting an investment when made. In this article, Johnson finds that the NCRS proposal has three errors:
1. Johnson argues: (a) that the combination of expensing and an interest deduction gives a negative tax or government subsidy, which is better than zero tax by the value of deducting interest. The subsidy would cause waste of capital because it makes investments that are not acceptable, in terms of their real pretax economics, look acceptable to the taxpayer after tax. The negative tax is also tax-rate dependent, giving more subsidy to high-bracket taxpayers than to low, thus tending to drive lower effective rate taxpayers out of the market.
(b) The combination of deduction of inputs against ordinary income and favorable capital gains rates for outputs creates a separate rate-dependent, negative tax.
2. Johnson argues that NCRS assumes, without support, that savings will respond powerfully to expensing. The consensus of the economics profession across the political spectrum, he argues, is that savings will be steady or drop. The proposals will not improve capital formation or productivity, he argues, but will instead just shift tax from those with capital to those without.
3. The third error Johnson finds is in scoring the proposal as a revenue raiser when it should be scored as losing over $300 billion over five years.
This article is a slightly revised version of testimony Johnson gave before the House Committee on Ways and Means on January 26, 1995.
TEXT:
The tax provisions of the Contract With America bill would allow taxpayers to elect a complicated new depreciation system, the neutral cost recovery system (NCRS), for basis in machinery and equipment. The logic of the system is that taxpayers should be given tax savings worth the value of 'expensing' for tangible personal property with a current tax life of 10 years or less. 'Expensing' means deducting the basis of one's investment immediately, as soon as the equipment is set up for use, although no resources have yet been lost. Under the NCRS proposal, the taxpayer in fact takes deductions over the tax life of the property, but NCRS then gives interest and inflation adjustments to compensate for the delay since the investment's start.
The NCRS proposal involves a number of serious errors, judged under neutral, professional standards. Left uncorrected, the errors make NCRS an irresponsible or ideological proposal. In the political excitement of the moment, it is too easy to let professional standards go by the way, but politics needs calm deliberation and dispassionate analysis. We need to slow down and correct these errors now or we will repent at leisure. The mistakes are, however, core errors so that correction might well kill the whole NCRS proposal.
1. Tax shelters. The first error of the NCRS proposal is in creating negative taxes or subsidies by combining expensing with an interest deduction and with lower capital gains rates. The proposal, for instance, allows both expensing of the investment and also an interest deduction on debt that makes the investment possible. Given the expensing, the interest deduction is a mistake. Expensing and interest deduction combined gives a negative tax, better than zero tax by the value of deducting interest. The negative-tax subsidy will cause waste of capital by making investments that are not acceptable, in terms of their real pretax economics, look acceptable to the taxpayer after tax.
The negative tax is tax-rate dependent, moreover, giving more subsidy to high-bracket taxpayers than to low. A bracket-dependent subsidy drives young entrepreneurs, start-up companies, and ailing companies out of the market, simply because they have too low a tax bracket. A bracket-dependent subsidy also leaves high-bracket investors with windfalls they do not turn over to investment.
Separately, the NCRS proposal combines expensing with advantageous capital gain rates. This combination creates a separate rate-dependent, negative tax.
It is the combination of expensing and the interest deduction or capital gains rates that is the killer. Household ammonia is a pretty good cleaner, for instance, and chlorine bleach is a pretty good cleaner too, but pour them in together to clean the bathtub and you will blow out the bathroom.
2. Savings don't respond to incentives. The second error of NCRS is that it assumes, without support, that savings will respond significantly to expensing. Without a significant savings response, NCRS just shifts the burden of tax or deficit from taxpayers with capital to taxpayers with little or no capital. That shift will do the country harm.
We have had our experiment with expensing in the enactment of the accelerated cost recovery system, 'ACRS,' in 1981. ACRS was, like NCRS, intended to allow expensing equivalence for machinery and equipment. The results of the experiment are in. ACRS failed. The 1981 act gave the largest tax incentives for capital in the history of the income tax and savings declined. The consensus of the economics profession across the political spectrum is that the data give no support for a significant positive response of savings to tax rate reductions. The proponents of ACRS were surprised. Congress cut back on ACRS in 1982, 1984, and 1986 because the incentives were too generous. We need to learn from past mistakes.
3. Revenue estimates. The third error in the NCRS proposal is in scoring the proposal as a revenue raiser. NCRS is equivalent to expensing in intent and impact, and if it were scored like expensing, NCRS loses roughly $300 billion over five years.
NCRS is scored as a revenue raiser because it switches the underlying method of depreciation from 200 percent declining balance to 150 percent declining balance. In early years, the switch means that annual depreciation deductions are smaller than under current law. But NCRS also gives the investor interest to compensate for the time that has gone by since the NCRS property was placed in service. Under what is called the 'NCR ratio,' depreciation deductions are increased to give the investor 3.5-percent after-inflation interest and also to compensate for inflation. The NCR ratio in effect treats the federal government as if it owed a debt equal to the value of tax savings from expensing. The debt is repaid, together with compound interest, over the tax life of the property. Most of the interest adjustment costs of the NCR ratio occur, however, after the five-year window and are ignored. With the interest and inflation adjustments, however, NCRS has a present value and real impact on the deficit equal to the value of expensing.
Scoring NCRS as a revenue raiser is dishonest accounting that hides the true cost and impact on the deficit. The logic of the NCR ratio treats the government as owing an interest-bearing debt to the taxpayer equal to the value of expensing. The debt, however, is off- balance sheet debt and is never mentioned in the accounting for the federal deficit. It is misleading to ignore the future tax reductions involved in paying off the debt, so as to treat NCRS as raising revenue. If this kind of accounting were done in the private sector to sell securities or consumer goods, the promoter should be indicted.
________________________________________________________________________________
The Three Errors Explained in Detail
________________________________________________________________________________
Negative Tax: Interest Deduction /1/
NCRS permits expensing of the cost of the investment and also permits the taxpayer to deduct interest on debt financing that makes the investment possible. The combination of expensing and interest deduction means that there is a 'negative tax' or subsidy on the investment, that is, treatment better than mere tax exemption or zero tax. The subsidy is generally worth the value of deducting interest. With borrowing of B, interest of i and a tax rate of t, the subsidy will generally be equal to B*i*t, (and throughout the balance of this paper I shall refer to the subsidy as B*i*t), as shown in the following box.
________________________________________________________________________________
The Algebra Behind the Tax
Subsidy
The negative tax or subsidy can be shown
with simple algebra.
Assume taxpayer Q
borrows amount 'B' paying interest of 'i' to make
an investment A in
equipment that will return profit at return rate
'r' annually.
Assume first that interest,, 'i,' and profit,, 'r,' are
equal,, so that in
absence of tax,, investment A is a break-even
investment. Every year
for the life of the equipment,, A makes B*r
profit and pays
B*i interest. B*r-B*i=0.
Adding tax with both expensing and an
interest deduction,, will
improve Q's
return. With expensing of investments,, taxpayer Q can
expand the amount
invested in A,, relying on tax savings generated by
the investment (at
tax rate 't') to reduce its cost. With expensing,,
borrowing of B
will allow Q to invest amount X such that X-t*X = B.
Rearranging the
algebraic terms,, it follows that the amount invested
X in investment A
under expensing will be B/(1-t). The expensing
allows Q to get
pretax return rate r on augmented investment,, B/(1-
t),, whereas Q
pays interest only on borrowed B itself. The pretax
position every
year is
[B/(1-t)]*r - B*i. (1)
The pretax position in expression (1) is
subject to tax at rate
t, so that Q/s
after tax annual income is
[B/(1-t)]*r - B*i -t([B/(1-t)]*r -
B*i). (2)
which is
equivalent to
(1-t)([B/(1-t)]*r - B*i)+ B*i*t
or B*r-B*i +
B*i*t.
(3)
When return rates and interest rates are
the same,, B*r and B*i
cancel out and
expression (3) becomes
B*i*t. (4)
Expression B*i*t is an after-tax positive
value or subsidy equal
to the tax saved
by deducting interest of i on borrowing B for an
investment in A
that would have no net value (r=i) in absence of tax.
The subsidy,, B*i*t, will allow taxpayers
to accept real returns
from investments that
are materially below the prevailing cost of
borrowed capital.
Setting after-tax annual return,,
expression (3),, equal to
zero:
B*r-B*i + B*i*t = 0. (5)
Rearranging the terms and factoring out
B:
r = i(1-t). (6)
Hence with the subsidy,, an investor will
accept a pretax or
real loss on the
investment,, which however will cover the cost.
There is a pretax
loss:
B*i*(1-t)-B*i or -B*i*t. (7)
But the tax subsidy B*i*t (in bold in (8)
below) will turn the
pretax loss in an
investment able to meet its cost of capital:
-B*i*t + B*i*t = 0. (8)
________________________________________________________________________________
Ordinary use of NCRS with debt would be a paradigm tax shelter. A tax shelter is an investment that is worth more after tax than before tax because of negative tax from artificial accounting losses. A tax shelter saves the investor tax that would otherwise be paid on consumed amounts or other outside income. Within the parlance of the Contract With America bill, 'neutral' means free from tax and, under that criterion, NCRS is not neutral because it goes beyond zero tax. To get back to zero tax, interest deductions must be disallowed or something equivalent.
The subsidy (B*i*t) will allow taxpayers in competition to bid up the price of investment A so that the return, r, stated as a percentage of investment, will drop below the prevailing interest rate, i, to i(1-t). For instance, where pretax interest rates are at 7 percent, then the return to the highest-bracket investors from the NCRS qualified investment needs only be 7 percent times (1-41 percent) or 4 percent.
A tax system that allows investors to profit privately from investments that return less than the prevailing interest rates wastes capital. We saw the waste in the tax shelters prevailing before the 1986 Tax Reform Act. Investors put their money into legal, but economically silly investments such as see-through office buildings, garden apartments that were constructed and then promptly torn down, jojoba beans in Costa Rica, twaddling record masters and the like. We have too little capital to waste it so.
The waste will be most significant for situations outside of the passive activity restrictions, that is, for corporate investors and for investors who actively participate in the business making the investment, but the negative tax will have effect even within the limitations imposed by passive activity rules. /2/ The proposal would limit NCRS deductions to the amount of the investor's business taxable income, but that limitation does not prevent use of the subsidy to zero out all business real income. Where the negative tax is available, the waste from returns below prevailing interest rates is as inevitable as apples falling.
Rate-dependent. The subsidy, B*i*t, also critically depends on the investor's tax rate at which the value of expense and interest deductions are computed. High tax bracket investors, with the highest tax rates, will get the largest subsidy -- perversely because they are considered to be the best bearers of tax. With the subsidy, the highest-tax bracket investors can thus tolerate the lowest pretax real economic return. Lower-bracket competitors such as start-up companies, young entrepreneurs with fresh ideas and ailing companies temporarily in trouble will get lesser or no subsidies and will be driven out of the market. Lower-bracket investors get a smaller subsidy and thus cannot bear the low pretax returns that high-bracket investors can. The table below shows, for taxpayers in various tax brackets, how much real pretax economic return an investor would need from NCRS property, given the B*i*t subsidy.
________________________________________________________________________________
Equals
Tax Interest
Less Tax Required
Bracket of Cost Subsidiary Real Pretax
Investor (i) (B*i*t)
Return (r)
41% 7%
2.9% 4.1%
34% 7%
2.4% 4.6%
28% 7%
2.0% 5.0%
15% 7%
1.05% 5.95%
0% 7%
0 7%
________________________________________________________________________________
As the chart shows, it is the highest-bracket investors who can accept the lowest returns and will pay the most for any given property. For example when interest rates are 7 percent, a 41-percent bracket investor needs only a 4-percent return from the investment and a start-up company facing no tax will need a 7-percent return. All other things being equal, high-bracket investors will bid up the price of the investments, driving down their returns, and driving out lower-bracket investors.
Where the price for NCRS property settles will depend on market factors, especially on how many alternative ways there are to avoid tax. The price will settle at a return sufficient for the effective tax rate of the marginal or last investor necessary to make all NCRS property sell. This price will depend in the market on the supply (basis that can be made to qualify within NCRS) and the demand for tax shelter (income that can not be sheltered from tax by cheaper means). Because of the cornucopia of other tax advantaged plans that swamp the market, /3/ we should expect the tax rate for the marginal investor to be less than the 34-percent statutory tax rate applied to large corporations.
Wherever the market settles, investors in a higher-than-the- marginal tax bracket will get a windfall that they do not turn over to investments. For example if the market settles with a pretax return of 5 percent, investors in the 41-percent and 34-percent brackets will get windfall returns in excess of their cost of capital. The 41-percent bracket investors will get 5 percent after tax whereas 4 percent would be a sufficient return to meet their cost of capital. The extra 1-percent windfall is a variety of waste.
Investors at or above the marginal rate will also tend to become the exclusive owners of NCRS property. Unless they have some material nontax advantages, investors in tax brackets below where the market has settled will be excluded from the market because the return is too low. Assuming again the market settles with a pretax return of 5 percent, then 15 percent and tax-exempt or tax-loss entities will not be able to afford the cost of capital and will not be able to buy the machinery.
It is difficult to see how the negative tax subsidy from the interest deduction can be anything other than an error. The rhetoric behind NCRS argues for zero effective tax rates, but not for a negative tax subsidy. If we were going to give a subsidy for investment as a budget item with a known cost, we would not give the subsidy to the nonmeritorious drivel like jojoba beans and record masters that have qualified for the shelter subsidy in the past. We would also not distribute the subsidy in a way that would drive start-up competitors and temporarily losing companies out of the market and we would not give the greatest subsidy to the richest first, in the exact negation of the pattern by which the tax burden is distributed, when Congress thinks about distribution issues and reaches a fair compromise.
The negative tax for debt-financed NCRS investments is not justified by consumption tax arguments. Expensing has been supported by the consumption tax argument that an income tax is a double tax on capital, /4/ but the argument fails for debt-financed investment because the starting step, a tax on the investor on the capital that makes up the investment, is missing. One cannot defend the negative tax as a subsidy for capital formation since neither borrowing cash nor promising to pay for property in the future is capital formed, but instead is anticapital or use of capital. Consumption tax theorists would correct the error of the negative tax by disallowing the interest deduction, taxing borrowed amounts, or some equivalent measure. /5/
Interest on debt is in theory taxed to the other side, and if debtor and creditor were in the same bracket and paid tax as interest is deducted, there would be just zero tax. But investors arrange themselves into clienteles under which high-bracket investors incur debt, and pension funds or low-bracket investors supply debt, so that only trivial tax is collected on interest. Even if that were not so and the creditors paid tax on interest and passed it back to borrowers, interest rates would rise, but still leave a tax-dependent subsidy on the debtor's side, giving windfall to some and excluding others just because of their tax bracket. /6/ The combination of expensing and interest deduction is a mistake.
Remedy. The remedy to bring tax up to zero is to disallow deduction of interest on debt, up to an amount of debt equaling the NCRS investment. It is neither administrable nor fair to attempt to 'trace' debt to specific uses. Money is fungible. Financing from any source goes into a common pool from which the company needs are withdrawn. Debt incurred for any purpose will allow the taxpayer to increase NCRS basis, even if the occasion for the debt was not NCRS investment. Two taxpayers with identical balance sheets should pay the same tax, without regard to the order or occasion of their borrowing. /7/ The necessary remedy follows section 263A(f) (construction interest) and identifies interest on the taxpayer's first debt as nondeductible.
In theory interest should be disallowed on an amount of debt equal to the current value of NCRS. A less complete solution would be to identify the target debt according to the basis of NCRS property under a slow, economic depreciation schedule.
Negative tax: capital gain. Negative tax subsidies also arise on NCRS property because some part of the return from sale of the property would be eligible for lower-rate capital gains tax. As a matter of well-established tax theory, expensing of an investment is equivalent to zero tax on the return from the investment. With expensing, taxes do not reduce the pretax percentage return the investor gets. /8/ When expensing of the investment is combined with advantageous capital gain, the overall tax effect is negative and increases the taxpayer's return from the investment. The following box gives numerical examples of the negative tax, both when there is no recapture and when there is full recapture of the prior deductions upon sale of the NCRS property.
________________________________________________________________________________
Numerical examples of negative
tax with
expensing and capital gain.
1. No capital gain. If an investment is
expensed,, the before-
tax rate of return
is the same as the after-tax rate of return.
Expensing does not
reduce the taxpayer's return. Assume for instance
an investment with
$ 100 input and $ 131.08 revenue in four years. In
absence of tax,,
the investment has a return ('internal rate of
return')* like
compound annual interest at 7 percent,, because $ 100
will grow to equal
$ 131.08 in four years at 7 percent compounded
annually:
$ 100 * (1+7%) /4/ = $ 131.08. (9)
Rearranging equation (9),, 7 percent is
also the discount rate
that will make the
present value of the revenue on termination equal
the present value
of the cost:
-$ 100 + $ 131.08/(1+7%) /4/ = 0. (10)
Because 7 percent is the discount rate
that will make the
discounted value
of costs and returns from the investment equal zero
on net,, 7 percent
is the 'internal rate of return' from the
investment. /9/
Expensing the $ 100 input immediately
means that the taxpayer
gets a tax savings
(here assumed to be at 40-percent tax rates) which
reduces its cost
(by 40 percent). When tax is imposed on the revenue
at the same tax
rate,, the net result is that the taxpayer's return
is not reduced by
tax:
-$ 100(1-40%) + $ 131.08(1-40%)/(1+7%)
/4/ = 0. (11)
Expression (11) is an illustration of the
thesis that expensing
is like a zero tax
on the percentage return. The investment gives 7
percent return
both before and after tax.
2. No recapture. The most dramatic
negative tax occurs for
investments like
timber,, business intangibles,, and R&D where there
is no recapture
and the entire return can qualify as capital gain.
Assume no
recapture and capital gains tax rates of 20 percent (half
the assumed
ordinary tax rate). All of the $ 131.08 received at the
end of the
investment is then capital gain,, reduced by 20 percent
(i.e. to 80
percent) to $ 104.86. The investment now has an annual
return of 15
percent because 15 percent is the discount rate that
makes the net
present value equal zero:
-$ 100(1-40%) + $ 131.08*(1-20%)/(1+15%)
/4/ = 0. (12)
Expensing of the input combined with
capital gain for the output
has increased the
return from a pretax 7 percent to a posttax 15
percent.
Expensing plus capital gain would also
allow inferior
investments to
compete with better ones,, posttax. Assume an
investment that is
a guaranteed money loser,, requiring an input of
$ 100 and giving a
return of only $ 88.31. Nonetheless the investment
gives 4.2-percent
return after tax,, which equals 7 percent (1 minus
40 percent) or
exactly what a 40-percent bracket taxpayer will get at
the given
prevailing interest rates on normally taxed bonds or CDs:
-$ 100(1-40%) + $ 88.31*(1-20%)/
[1+7%(1-40%)] /4/ = 0. (13)
Without recapture,, expensing will cause
wasted capital by
turning
investments that lose money on their economic merits ($ 100 to
$ 88.31) into
investments that are acceptable to taxpayers who get the
subsidy.
3. Some capital
gain. Traditionally 'full' recapture remedies have
been limited to
making only gain up to the original cost basis ($ 100)
ordinary income.
The NCRS proposal does not increase recapture. Under
such recapture,,
for the 7-percent investment of $ 100 in and $ 131.08
out,, the revenue
output up to $ 100 would be subject to ordinary tax
of 40 percent at
termination,, but the extra $ 31.08 'gain' would be
subject to tax at
20 percent. The regime increases the return from 7
percent pretax to
9 percent after tax:
-$ 100(1-40%) + [$ 100*(1-40%)+$ 31.08
(1-20%)]/(1+9.05%) /4/ = 0. (14)
A tax regime that increases the
taxpayer's return after tax will
also waste capital
because it will make inferior investments
acceptable after
tax. The following investment gives only a 3.2-
percent real
return before tax,, because it grows only from $ 100 to
$ 113.42 in four
years. /10/ But to a 40-percent bracket taxpayer,,
the 3.2-percent
return will look as good as the prevailing 7-percent
return:
-$ 100(1-40%) + [$ 100*(1-40%)+$ 13.41
(1-20%)]/[1+7%(1-40%)] /4/ = 0. (15)
A tax regime that makes a pretax
3.2-percent investment look as
good as a pretax
7-percent investment will cause waste of capital.
________________________________________________________________________________
To prevent the negative tax, all gain from NCRS property must be ordinary income. /11/ An incomplete solution, which would prevent some but not all negative tax, would be to provide that recaptured depreciation deductions are increased by 3.5-percent interest compounded, just as the NCR ratio increases the depreciation deduction by 3.5-percent interest.
The Failure of Expensing Incentives
The second error in the NCRS proposal is the assumption that it will produce a significant increase in savings. The purpose of NCRS is to provide an incentive for the formation of capital (savings). The increase in capital, if it were a feasible result, would increase productivity and economic growth and would create more and better paying jobs. Without the hoped-for increase in savings, however, NCRS will not increase productivity, growth, or jobs. NCRS will then just shift the burden of tax or deficit from taxpayers with capital and onto taxpayers with little or no capital. Shifting the burden of tax and deficit downward, onto the low-capital taxpayers who start with a lower standard of living pretax, will do the country harm.
The assumption that savings will respond significantly to expensing has no support in the economic literature or the underlying economic data. If economics is a science, it is a science because it is empirical, that is, it tries to disprove its assumptions on the basis of the evidence. We need to look at the record.
We have had our grand experiment with allowing expensing of machinery and equipment. The results of the experiment are in. In 1981, in the enthusiasm at the start of the Reagan administration, Congress adopted a tax system for depreciable property, the ACRS, intended like NCRS, to give the equivalent of expensing for machinery and equipment. ACRS reached for expensing-equivalence by giving an immediate 10-percent investment tax credit, whereas NCRS would reach expensing by giving an interest-like augmentation to deductions. ACRS in 1981 even went considerably beyond expensing in value because in computing expensing-equivalence ACRS used an interest rate that was materially higher than the real interest rate the government must pay. /12/ The purpose of ACRS, like that of NCRS, was to give incentive to capital formation, i.e., to savings. /13/
Proponents of ACRS in 1981 argued that savings would respond dramatically to the new incentive. ACRS was part of the 'supply-side economics revolution' which sought to shift the focus of public policy away from fiscal stimulus, through government spending and deficits, and onto incentives for the production of goods. /14/ Some proponents argued that the economic response to tax cuts might be large enough to make the 1981 tax cuts self-financing. The Laffer Curve, for instance, showed a range where economic response to tax cuts would be so extraordinary that the government revenue would rise by participating in the new higher level of economic activity. /15/ The 1980 Reagan campaign projected that additional economic growth from tax nicentives would lead to federal surpluses (in addition to the surpluses to be had from cutting federal spending) but only in the long term. /16 More moderate advocates doubted a response that was self-financing, /17/ but they did assume a high response, arguing that the 'elasticity of savings in response to the real after-tax rate of return was relatively large and elastic.' /18/
Judged on its own terms, ACRS was a failure. Savings rates declined, notwithstanding the most dramatic cut in effective tax rates on capital in history. Personal savings, for instance, dropped from 7.9 percent of disposable personal income in 1980 to 6.4 percent of disposable personal income in 1985, to 4.2 percent of disposable income in 1990. /19/ Supply-side economics renounces deficit stimulus, /20/ but the annual deficit ballooned. The Reagan campaign projected that the deficit would shrink to $27 billion in 1980 and become a surplus of $93 billion by 1985. /21/ The deficit in fact grew in size, under the 1981 tax cuts, from $74 billion in 1980 to $220 billion by 1986. /22/ Proponents of ACRS were surprised, even shocked. Norman Ture, one of the primary architects of ACRS as a Treasury official in 1981, was quoted as saying that the drop in savings is 'not only a disturbing result,...it is very surprising.' /23/
On terms not allowed by Reaganomics, the deficit created by ACRS caused a fiscal stimulus to the economy. The fiscal stimulus of deficit spending created economic activity and made everyone riding upward feel good. The high deficits, however, were the target or enemy of supply-side depreciation, not its purpose. President Reagan, for instance, had promised that 'we are putting the false prosperity of overspending, easy credit, depreciating money and financial excesses behind us.' /24
The conclusion that investable savings did not respond significantly and positively in reaction to the 1981 incentives is a strong consensus shared by the economics profession across the political spectrum. Robert Hall of the conservative Hoover Institute, for instance, has concluded that consumers do not defer consumption in response to higher expected interest returns and that apparent findings to the contrary were in error. /25/ Joel Slemrod of the respected National Bureau of Economic Research and the University of Michigan and the editor-in-chief of the National Tax Journal tells us that the economic research from the 1980s has concluded that savings are unresponsive to lower tax. /26/ Tanzi and Sheshinski of the International Monetary Fund have found that there was no important increase in U.S. savings rates after ACRS. /27/ Jonathan Skinner and Daniel Feenberg of the National Bureau of Economic Research report that the consensus in the economic literature is that any positive response of savings to interest rate increase is 'fragile and fleeting.' Long term, since the 1970s, they find, savings have reacted negatively to increases in after-tax returns. /28/ Barry Bosworth and Gary Burtless of the Brookings Institution show the decline of different kinds of private savings throughout the 1980s and advise that 'government policy-makers should act under the presumption that income tax incentives for saving are likely to fail.' /29
Tax incentives do have a powerful influence on the timing, form, and channeling of investment, but those effects do more harm than good. Taxpayers, for instance, delay or anticipate sales to take advantage of shifts in capital gains rates. /30/ Taxpayers will also rapidly change the legal or financial forms by which investment is undertaken. In 1986, for instance, when Congress changed the relation of corporate and individual tax rates, S corporations became better investment vehicles than regular C corporations. /31/ Corporations responded dramatically, scrambling to make S elections. /32/ Finally, taxpayers will move fixed capital already in existence to achieve tax savings. When IRAs and Keogh plans were liberalized in 1981, taxpayers reacted far more dramatically than expected, but apparently without increasing their total savings. /33/
Movement in the form or direction of savings does not increase overall capital and will usually cause waste by inducing the use of the fixed savings in inferior investments. Tax incentives can affect channeling, short-term shuffling, and financial manipulations, but they will not cause capital formation nor improve the efficiency of its use. The hard thing to do is to increase real savings and underlying savings did not respond to the 1981 incentives.
Explaining negative and sluggish reactions. The results from the data that underlying savings react sluggishly or negatively to tax cuts are plausible in terms of normal economic behavior, even if it seems implausible at first.
Savings drop when taxes go down, first, because of 'target savers' who quit saving earlier when their after-tax return rate increases. Assume, for instance, that Saver A needs $150,000 for tuition payments in 10 years for his child. When after-tax return rates are 12 percent after tax, A must set aside $8,548 a year for 10 years. Reduce A's return to 3 percent after tax and A will need $13,085 per year. Take away the difference (9 percent) in tax and watch A get really desperate and also increase savings by $4,500 per year per child to meet his targets.
Taxpayers who are saving for a car or down payment on a house or retirement security are similarly target savers who will contract savings rates when after-tax returns on saving go up and increase savings when tax rates go down.
Decreased taxes decrease savings, secondly, by reason of what has been called the 'mail-box effect,' /34/ that is, taxpayers consume a high proportion of cash received. Under traditional theory -- the permanent income theory -- rational, far-sighted consumers should average their resources evenly over their life-times. They should not increase consumption simply because more cash is realized, unless the cash also signals a real increase in wealth or the permanent flow of income. /35/ Under the model, cash is no reason to increase consumption. /36/ But econometricians are finding increasing evidence that consumers are not very far-sighted in their savings behavior. Consumption is closely related to receipts, averaged over a time horizon of just a few years, and it is more volatile than allowed by life-cycle model. Consumers keep only enough savings to buffer uncertain incomes in the next few years and they consume the rest. /37/ Under these findings, a tax cut decreases savings because it increases the cash that consumers have in hand.
The sluggish response of capital formation to tax incentives also occurs because so much of savings goes forward even when real interest rates are negative. Much saving is done to create a buffer or to average out consumption. Money that you have for subsistence in case of trouble is very much more valuable than money spent on luxuries during good times.
It is thus rational to save money for a rainy day, even if the return rates are negative. Pharaoh had a dream that Joseph said meant Egypt would have seven fat years and then seven drought years. Joseph said save. Mice ate the grain so that it shrank by 10 percent to 15 percent per year, so that the real rate of interest on stored grain was a negative 10-15 percent. Still, Joseph was no dummy. The rationality of saving, even in the face of expected negative interest rates also explains, for instance, why in high-inflation Latin American countries, the wealthy often save in dollars. Dollars give no interest coupon and dollars do lose value, although not at the rate of the local currency. Saving still goes on.
The rationality of saving suggests that even when the interest is reduced to the negative range, there is little effect on savings. /38
Revenue Estimates
The third error in the NCRS proposal is in scoring the proposal as gaining revenue for the government. NCRS is intended to be equivalent to expensing in net present value terms and cost to the government. It is misleading accounting to score NCRS with a revenue estimate loss less than what would be scored for expensing. Scored honestly, I estimate that NCRS will lose over $300 billion over the five-year budget window.
Under the Budget Act, revenue estimates are prepared for tax proposals by the staff of the Congressional Budget Office and the Joint Committee on Taxation to analyze how a tax proposal will affect federal deficits. /39/ The federal deficits are large enough now that no one would wish them larger. Revenue estimates limit how much can be given away by competing tax proposals.
In substance, the NCRS proposal treats the federal government as owing an obligation at the time and amount determined by expensing and requires the government to pay interest on the obligation by increasing annual deductions. In fact, NCRS gives taxpayers deductions over the tax life of the property under the pattern of 150-percent declining balance method of depreciation, /40/ but the proposal increases the annual deductions by a compound interest factor, the 'NCR ratio,' to compensate the taxpayer for the delay since the property was placed in service. The NCR ratio augments the annual deduction to cover both inflation, now at 3.2 percent, and an after-inflation 3.5-percent interest amount. The two adjustments mean an annual interest factor of 3.5 percent + 3.2 percent or 6.7 percent, or slightly higher than the current 6-percent federal interest costs on short-term federal obligations, in which the obligee is similarly protected from rises in inflation.
Since NCRS treats the government as owing an interest-bearing debt equivalent to expensing, the revenue cost from NCRS should be scored as the equivalent of the revenue loss from expensing. In analyzing the distributional impact of the burden of proposed tax changes, the staff of the Joint Committee on Taxation has adopted a principle to maintain consistency in the measurement of 'tax proposals that are expected to have equivalent economic effects on the taxpayer.' /41/ It is, for instance, possible to create a zero effective rate on pension fund investments either by excluding contributions from tax or by excluding distributions from tax and the present value of either alternative would be the same. The staff analyzes both as having the same economic impact, although in the former case the tax benefits are front-end loaded and in the latter they are back-end loaded. The principle of consistency applied to revenue estimating would mean that NCRS would be scored the same as its intended equivalent, expensing. Using available data, I estimate that NCRS would lose over $300 billion during the next five years if NCRS were scored as an expensing-equivalent. /42/
Because of an artificial five-year convention in revenue estimating, /43/ however, NCRS is in fact scored as if it were a revenue raiser. Current law allows depreciation deduction under the 200-percent declining balance method and switching to 150-percent declining balance means more undiscountedgovernment revenue for five years, even with the augmentation by the NCR ratio. Most of the costs of the NCR-ratio interest factor occur beyond the five-year window. Revenue costs that occur after five years are ignored.
Scoring NCRS as a revenue raiser is misleading, dishonest accounting that hides the true cost. NCRS will not decrease the deficit or federal debt, as it is scored. NCRS will increase the federal debt by more than $300 billion over five years. A public policy is responsible only if it measures the costs of a proposal and compares the costs fairly and forthrightly with the benefits. If the costs are hidden or ignored, then the costs and benefits can not be responsibly compared. The hiding of costs by pushing costs to beyond the five-year window will distort decisionmaking. Unfortunately, the first victim to be misled will be Congress itself.
________________________________________________________________________________
FOOTNOTES:
________________________________________________________________________________
/1/This section is based on Johnson, 'Tax Shelter Gain: The Mismatch of Debt and Supply-Side Depreciation,' 61 Tex. L. Rev. 1013 (1983), which has more extensive arguments on some issues.
/2/Section 469 provides that deductions from any passive activity may be used only against taxable income from all passive activities. Widely held corporations and active participants in a business are exempt from the limitations. Even for nonexempt taxpayers, deductions representing artificial accounting losses may be used against consumed amounts or amounts that represent real economic income. The passive activity rules are also under constant political assault, in part, because they represent such bad theory. There is nothing wrong with being passive nor with being active nor with the combination. The assumption behind the exemptions that businessmen and public corporations would not use artificial deductions is also not warranted.
/3/Harvey Galper and Gene Steuerle, 'Tax Incentives for Savings,' 3 Statistics of Income Bulletin, 1, 4 (Spring 1984) estimate that 80 percent of the assets held in wealth of individuals has an effective tax rate of zero, which implies that no high-bracket taxpayer would accept very much lowered pretax returns from NCRS investments.
/4/See, e.g., Norman Ture (later under secretary of the Treasury urging ACRS), 'Statement,' 1 General Tax Reform: Panel Discussions bef. Comm. on Ways & Means, 93d Cong., 1st Sess. 160,162 (1973) (benefits for depreciation just reduce the double tax on savings).
/5/David Bradford, 'The Economics of Tax Policy Toward Savings,' in The Government and Capital Formation 42-50 (George Von Furstenberg, ed. 1980)(disallowing interest). Other advocates of consumption tax would cure negative tax by including borrowing in taxable income. Dep't. of Treasury, Blueprints for Basic Tax Reform 124 (1975); Nicholas Kaldor, 'Comment,' in What Should be Taxed: Income or Expenditures? 151 (Joe Pechman, ed. 1980).
/6/See discussion, Johnson, 'Tax Shelter Gain,' 61 Tex. L. Rev. 1013, at 1039-1049 (1983)
/7/See William Klein, 1962 Wisc. L. Rev. 608, 611-612.
/8/The thesis that expensing is the same as no effective tax on the percentage income from an investment is one of the base theorems of modern tax economics. See, e.g, Cary Brown, 'Business-Income Taxation and Investment Incentives,' in Income, Employment and Public Policy: Essays In Honor of Alvin H. Hansen 300 (1948); Dep't of the Treasury, Blueprints For Basic Tax Reform 123-24 (1977).
/9/As shown in text, internal rate of return is the interest or discount rate (or 'solving rate') that will make the present value of the costs and returns equal to zero. The internal rate of return has some anomalies -- there are sometimes more than one solving rate and the solving rate may have misleading implications about terminal results when interim returns cannot be reinvested. The anomalies do not apply to the examples in text, however, so that internal rate of return is equivalent in the examples to the annual interest rate, assuming a compound schedule.
/10/$100 * (1.032) /4/= $113.42
/11/NCRS property, however, must continue to be subject to the limitation on capital losses, which have a different rationale. See, e.g., 'Johnson, Deferring Losses with an Expanded Section 1211,' 48 Tax L. Rev. 719 (1993).
/12/ACRS, as adopted in 1981, was equivalent to expensing for corporations (46-percent bracket) at a 24-percent pretax interest rate. See Johnson, 'Tax Shelter Gain,' 61 Tex. L. Rev. 1013, 1022 n. 46 (1983) (equating expensing and ACRS at 13.1 percent after tax which translates into 24-percent pretax interest at 46-percent tax rate). Even long-term federal bonds in 1981 gave only 13- to 14- percent interest. Council of Economic Advisers, Economic Report of the President 1993, at 428 .
/13/See Pamela Gann, 'Neutral Taxation of Capital Income: An Achievable Goal?' Law & Contemp. Probs., Autumn 1985, at 77, 97-108 (giving overview of arguments for capital formation).
/14/See Paul Craig Roberts, 'Supply-Side Economics,' in 4 New Palsgraffs Economic Dictionary 615 (1987).
/15/Jude Wanniski, 'Taxes, Revenues and the 'Laffer Curve,'' Public Interest, Winter 1978, at 3, reprinted in Arthur Laffer & Jan Seymour (ed.), Economics of the Tax Revolt 7 (1979). See also Arthur Laffer, Statement, 'The Economics of the President's Proposed Energy Policies: Hearings before the Joint Economic Committee,' 99th Cong., 1st Sess. 16, 18 (1977) ('[I]t is not only conceivable but entirely possible that [the proposed $2.8 billion tax!] will lead to reduced overall revenue....').
/16/Reagan Bush Committee, 'Ronald Reagan's Strategy for Economic Growth and Stability in the 1980s,' Table 2 (News Release, Sept. 9, 1980, Tax Notes Microfiche ed. Doc 80-7296) (projecting surplus of $121 billion in 1985 if $92 billion in cuts in federal spending were passed).
/17/Don Fullerton, 'On the Possibility of An Inverse Relationship Between Tax Rates and Government Revenues,' 19 J. of Pub. Econ. 3 (1981), concentrating on labor responses, found it possible but unlikely that total revenues would increase. Donald Keifer (Library of Congress, Congressional Research Services), An Economic Analysis of the Kemp/Roth Tax Cut Bill H.R. 8333, Cong. Rec. Aug. 2, 1978, H7777-7787, using privately developed economic models with an aggressive supply-side orientation, nonetheless, projected nothing close to elasticity of 1-percent saving increase per 1- percent tax decrease. Paul Craig Roberts, an important supply-side advocate, argues that self-financing was never the goal, but he made the argument only after it was clear that self-financing did not happen. Supply-Side Economics, supra note 14.
/18/Mai Nguyen Woo (Research Associate IRET), 'Taxation, Savings, and Labor Supply: Theory and Evidence of Distortions' in Essays in Supply Side Economics 117, 131 (David G. Raboy, ed. 1982). Similarly, Michael Boskin, who became Reagan's chairman of the Council of Economic Advisors found a significant positive response (0.4 elasticity) by savings to after-tax interest, 5 'Taxation, Saving, and the Rate of Interest,' 86 J. of Pol. Econ. S3 (1978), but see Lans Bovenberg, 'Tax Policy and National Saving in the United States: A Survey,' 42 Nat. Tax J. 123129 (1989) ('others have found it difficult to reproduce his results.'); Robert Hall (Hoover Institute), infra note 25 (findings that consumers save when expected interest rises are in error) and other sources cited, infra notes 19- 29.
/19/Dept. of Commerce, Statistical Abstract of the United States, 1994 No. 695 (drawing from BEA, National Income and Product Accounts). Personal savings remained low at 4 percent of disposable personal income in 1993.
/20/See, e.g., Paul Craig Roberts, 'Supply-Side Economics,' supra note 14, at 615.
/21/Reagan Bush Committee, supra note 16 (projecting $121 billion surplus by 1985, including $95 billion in expenditure cuts); Office of Management and Budget (OMB), U.S. Budget, Fiscal Year 1982, M3 (1981) (projecting modes surplus by 1984 without 1981 tax incentives).
/22/Dept. of Commerce, Statistical Abstract of the United States, 1994 Table No. 505.
/23/Kilborn, 'Americans Saving Less Now Than Before the 1981 Act,' New York Times, Sept. 6, 1983 at A1, col. 3.
/24/Budget Message of the President, 'U.S. Budget, Fiscal Year 1983' at M5. See also Paul Craig Roberts, 'Supply-Side Economics,' supra note 14, at 615.
/25/Robert Hall (Hoover Institute) 'Intertemporal Substitution in Consumption,' 96 J. of Pol. Econ. 339, 350 (1988)
/26/Joel Slemrod, 'Do Taxes Matter? Lessons from the 1980's,' 82 Am. Econ. Rev. 250, 251-252 (1992)
/27/Tanzi & Sheshinski, 'Fund Study Suggest Answer to the U.S. Savings Puzzle,' 13 IMF Surv. 353, 366-67 (1984).
/28/Jonathan Skinner and Daniel Feenberg, 'The Impact of the 1986 Tax Reform Act on Personal Saving,' NBER Working Paper No. 3257 at 10-17 (1990).
/29/Bosworth and Burtless, 'Effects of Tax Reform on Labor Supply, Investment and Saving,' 6 J. of Econ. Perspectives 3, 14-23 (1992).
/30/Leonard Burman, Kimberly Clausing & John O'Hare, 'Tax Reform and Realizations of Capital Gains in 1986,' 47 National Tax Journal 1 (1994) (seven times more capital gains were realized in December 1986 than in December 1985 even though rates were the same because in 1986 sellers knew rates would rise in 1987); Fredland, Gray and Sunley, 'The Six Month Holding Period for Capital Gains: An Empirical Analysis of its Effect on the Timing of Gains,' 21 Nat. Tax J. 467, 471 (1968) (investors sell property in the month after meeting the holding period requirements for capital gain at a rate that was eight times the rate of sales in the month before qualifying for the preferential rates).
/31/The terminal value of a $1 investment in a partnership, sub S corporation or C corporation debt is [1+R(1-tp)]n and the terminal value of a $1 investment in equity of a corporation, terminated by a capital gain transaction is [1+R(1-tc)]n- cg*([1+R(1-tc)]n-1), where R is the prevailing pretax return on capital, n is the years until termination of the investment and tptc, and cg are the maximum tax rates on individuals, corporations, and capital gain respectively. Myron Scholes & Mark Wolfsen, Tax and Business Strategy 57-59 (1992). Prior to 1986 with R = 10 percent, n = 25, tp = 50 percent, tc = 46 percent, and cg = 20 percent, the partnership, debt or subS would give $2.65 per dollar invested and the C corporation stock would give $3.18. After 1986, with tp = 33 percent, tc = 34 percent, cg = 28 percent, and R and n the same, the partnership form would give income of $5.05 and the C corporation stock would give $3.83.
/32/Robert Leonard, 'A Pragmatic View of Corporate Integration,' Tax Notes, June 1, 1987, p. 889 (In first two weeks of 1987 there were 220,000 subchapter S elections, compared with 70,000 elections for all of 1985).
/33/Eric Engen, William Gale and John Karl Scholz, 'Do Saving Incentives Work?' Brookings Papers on Economic Activity 1994-1, 85, 150.
/34/Douglas Bernheim, Introduction in National Savings And Economic Performance 1,5 (Douglas, Bernhein and John Shoven, eds. 1991).
/35/A rational far-sighted consumer should smooth his consumption so that his expected lifetime wealth is consumed at steady, average rate over his whole life. Wide variations in amount consumed per year will not maximize utility of the dollars spent. Franco Modigliani, 'The Life-cycle Hypothesis of Saving, the Demand for Wealth and the Supply of Capital,' 33 Social Research 160 (1966).
/36/Auerbach and Hassett, 'Corporate Savings and Shareholder Consumption,' in National Savings and Economic Performance 75 (Douglas Bernhein and John Shoven eds. 1991)
/37/Barry Bosworth, Gary Burtless and John Sabelhaus, The Decline in Saving: Evidence from Household Surveys, Brookings Papers on Econ. Activity, 1991-1, 226 (summarizing the research);. Christopher Carroll and Larry Summers, 'Consumption Growth Parallels Income Growth: Some New Evidence,' NBER Working Paper No. 3090 (1989) (permanent income or life-cycle hypothesis is inconsistent with the grossest features of the data on actual consumption patterns); Flavin, 'The Adjustment of Consumption to Changing Expectations about Future Income,' 89 J. of Pol. Econ. 974, 1006 (1981) (data rejection of the permanent income or life-cycle hypothesis is significant); Campbell and Mankiw, 'Permanent Income, Current Income, and Consumption,' NBER Working Paper No. 2436, 32 (1987) (estimating that 40-50 percent of income is received by individuals who consume according to current income rather than according to wealth or permanent income).
/38/The fact that savings go on under negative interest rates implies that much of the positive interest received for saving is 'rent.' The premium return or 'rent' on savings means, by definition, that the rent or premium could be taxed away without changing the amount saved. The maximum revenue from capital (as shown, for instance, in the Laffer Curve) should be at a point in which after- tax interest is negative. Labor income, by contrast, could never be taxed into the negative range without suppressing the income in full. The rationality of savings even in the face of negative returns suggests that there might be efficiency gains possible by shifting taxes from labor and onto capital.
/39/Rules of the House of Representative, XXI,c.8, adoption reported 137 Cong. Rec. h.5 (Jan. 3, 1991) (tax bill out of order unless accompanied by revenue estimate by Congressional Budget Office with Joint Tax Committee). See Joint Committee on Taxation, 'Discussion of Revenue Estimation Methodology and Process' (JCS 14- 92) (Aug. 13, 1992).
/40/The 150-percent declining balance method of depreciation first computes what fraction of basis would be allowed in the first year if basis were allowed in equal portions in each year of the tax life. (e.g. 1/n, where n is the tax life). It then increases the fraction by 150 percent to 1.5/n. The fraction 1.5/n is multiplied by the remaining or adjusted basis from the end of the prior year to determine the depreciation deduction for the current year. A constant percentage of an ever decreasing adjusted basis would never allow the taxpayer to recover the entire basis, so the taxpayer is allowed to deduct remaining basis in equal amounts over the remaining years when that amount is bigger (1.5/n 1/m where m is remaining years of the tax life).
/41/Staff of the Jt. Comm. on Taxation, 'Methodology and Issues Measuring Changes in the Distribution of Tax Burdens' 4 (JCS 7- 93) (July 14, 1993).
/42/I reach an estimate of revenue loss of roughly $330 billion composed of $775 billion revenue cost from expensing, less $445 billion loss from depreciation under current section 168. My estimate assumes $3.1 trillion investment in NCRS eligible property in 1995- 1999, assuming producer durable equipment of $528 billion in 1994 extrapolated forward at 5.4-percent annual growth. Bureau of Economic Analysis, Survey Of Current Business 3 (Nov. 1994) and a 25-percent average tax rate for users. My rough estimates can not be expected to correspond to Joint Committee, Congressional Budget Office or Treasury figures, which start from a different, nonpublic tax data base.
/43/The five-year window is a mandate of the 1995 Balanced Budget Act (Joint Committee on Taxation, 2 U.S.C.A. section 632(e)(6)(1995) and the House Rules, XXI,cl. 8, but that mandate does not make the accounting less misleading or the costs less hidden.
________________________________________________________________________________