THE BEAR'S
LAIR The
Laffer Curve key By Martin
Hutchinson
Younger readers probably don't
remember this, but before 1980 much of the world
was subjected to extraordinarily high marginal tax
rates - 91% in the United States before 1964, 98%
in Britain from 1974-79 (and 135% in 1968). The
Laffer Curve, which states that tax revenues are
zero at marginal tax rates of 0% and 100% and peak
somewhere in the middle, reportedly sketched on a
cocktail napkin by Arthur Laffer in 1974, was thus
an enormous intellectual breakthrough, even if in
retrospect it appears blindingly obvious.
However, in the nearly 40 years since
then, there has never been agreement on the Laffer
Curve's most important parameter: the tax rate at
which revenue is maximized.
The Laffer
Curve is so intuitively obvious that even the left
has been forced to acknowledge it. Their defenses
against its policy implications have come in two
forms: either suggesting that the
peak revenue is reached
only at very high tax rates, or like Barack Obama
in his 2008 presidential campaign or French
socialist presidential candidate Francois Hollande
recently, suggesting that the moral imperative of
high tax rates on the rich is such that taxes
should in some circumstances be raised even if
government revenue is thereby reduced.
Hollande's proposed 75% income tax clearly
comes into this category. Conversely, Obama's
initial proposal of an increase in the capital
gains tax rate from 15% to 20% probably doesn't.
The high US and British tax rates before 1980 were
unquestionably of this type.
In British
Chancellor of the Exchequer's George Osborne's
budget speech of March 21, he suggested that the
50% top income tax rate on incomes over 150,000
pounds (US$240,000) introduced by the Labour
government in 2009 was similarly
counter-productive.
Certainly the decline
in UK self-employed tax payments in January, for
the 2010-11 tax year, strongly suggested that for
the self-employed, at least, the 50% tax rate was
counterproductive - although it may still have
yielded net revenue from those with salary income,
notoriously more difficult to hide (legally or
illegally) from the taxman.
All this
assumes the Laffer Curve is a smooth curve rising
to a single well-defined peak and then declining.
This is intuitively attractive, but there is no
reason for this simplicity to hold in all cases,
since a multitude of factors affect both taxpayer
behavior and economic outputs.
For
example, the very high tax rates from the 1930s to
the 1970s may have suppressed top management
remuneration, since boards of directors were not
able to get useful additional effort out of
managers by paying them extra cash. Pharaonic
perquisites would frequently avoid tax, but it was
much more difficult to tie them effectively to
performance. Hence management played a lot of
golf, and economic output maybe suffered. In the
real world, boards of directors' incentives are
not very closely aligned with shareholder reward,
and management's own incentives are different
again.
There are a number of factors on
which the Laffer Curve peak might conceivably
depend:
One is the efficiency of tax
collection. In a very inefficient system, such as
Russia, high marginal rates of tax are completely
ineffective, because the costs of tax evasion are
relatively small, and consist mostly of paying off
the right officials. Tax collection becomes an
instrument of state control: Those with good
political connections negotiate a favorable tax
deal, whereas those who have alienated the
authorities (such as Mikhail Khodorkovsky) find
the tax code used to impose draconian criminal
penalties.
In such a system, the 13% flat
tax introduced by Russia in 2001 produced a huge
increase in revenues, indicating that the Laffer
Curve optimum tax rate there was not much above
13% and may even have been below it.
At
the opposite extreme, Britain in the 1950s and
1960s, a land of extraordinarily high compliance
with legal demands (at least among the richer
sections of the population) probably had a Laffer
Curve optimum at a higher marginal tax rate than
other societies. While the top tax rate reduction
from 98% to 60% in 1979 increased revenue, as
Laffer would suggest, the further 1988 reduction
to 40% may have reduced it.
That suggests
that in law-abiding un-entrepreneurial Britain, at
least short-term tax yields were maximized at a
marginal rate above 50%. The apparently different
experience in 2009-12 with a 50% top marginal rate
doubtless reflects the changes in Britain as much
as anything else; the society is more
heterogeneous, and the social forces reinforcing
legal compliance are much weaker.
Another
factor affecting the Laffer Curve peak is the type
of tax. Capital gains tax yields are typically
maximized at considerably lower marginal rates
than income tax yields, because the alternative of
simply not selling the investment concerned is
always available. The 1978 US move from a 49%
capital tax rate to 28% increased the tax's yield,
and there is evidence that the 1997 tax rate
reduction from 20% to 15% at least did not reduce
the tax yield much.
Similarly, inheritance
tax rates of 40-50% are far too high to maximize
revenue, since the tax is relatively easy for the
ultra-rich to avoid. Given the damage this tax
does to family-owned small businesses, it's likely
that revenue would be maximized by a rate below
20%.
There are in fact two points of
Laffer Curve tax yield optimization: the
short-term optimum, at which the tax yield in any
given year is optimized; and the long-term optimum
at which the yield is maximized on the basis of
discounting to infinity the yield from the tax,
taking into account the economic effect of levying
it and the effect on other tax revenues of its
economic disincentives.
For minor taxes,
the two optima are close together. Extensive
efforts by British chancellors of the Exchequer to
find Laffer Curve maximizing tax rates for
alcohol, tobacco and petrol duties appear to have
produced general agreement on the best levels.
These optimum levels are relatively higher (for
tobacco and alcohol) than in the 18th century,
when enforcement was poor and smuggling rife, but
lower than in the 1960s, when lower population
mobility and better law compliance limited the
potential for illegal imports from continental
Europe or elsewhere.
Value-added tax (VAT)
in this respect is attractive to taxmen (as are
consumption taxes in general), because the
disincentive effect of higher rates is minor, and
even the economic effects of high VAT rates are
not too severe. Thus the Laffer Curve bends down
only gradually.
In Mediterranean
countries, notably Greece recently, it would
appear that VAT levels above 20% lead to so large
a "black economy" that the taxes' yields decline
(Hungary's 27% VAT may also be too high to be
effective, for example.) However in countries like
Germany, Scandinavia and the United Kingdom, where
tax compliance is high and tax enforcement
effective, we may well not have seen the Laffer
Curve optimum, which may occur only at VAT rates
of 40-50%, higher than any yet experienced in
practice. Apart from Hungary's 27%, Sweden's
25% is the highest VAT rate in the European Union.
In general, high VAT rates are more effective in
large tax areas (because of the potential for
smuggling) so the European Union is ideally set up
for them. No doubt in the near future we will see
the EU tax apparatus test the revenue limits of
VAT with a 40-50% rate, at which point human
knowledge will be advanced, at the expense of
unfortunate European taxpayers.
For
capital gains tax, income tax and estate tax, the
short-term and long-term Laffer Curve optima are
potentially further apart. Capital gains tax
affects capital formation and investment, the
engine room of the economy, so although compliance
can be quite high and the taxes are fairly
difficult to evade, the long-term Laffer Curve
optimum is no more than 20%.
Rates higher
than this adversely affect economic performance,
even if they produce short-term revenue. In the
UK, capital gains tax was increased in 2010 only
to 28%, after a study concluded that an increase
to 40%, as had originally been proposed, would
lose revenue. In the longer term, even 28% may be
too high.
For estate duties, the picture
is quite clear. These taxes are easier to avoid
than income tax, or even capital gains tax, so the
short-term Laffer Curve optimum is at
correspondingly lower rates, probably around
15-20%. It is difficult to assess the adverse
economic effect of the taxes, because it occurs
only on generational transfer, but as they form a
one-off capital tax of 20% or more the adverse
effect can certainly be severe, as is the
difficulty for heirs in paying them.
In
any case, since their yield is relatively modest,
the optimum long-term Laffer Curve estate duty
rate may well be zero, since with a zero estate
tax rate the yield of other taxes is increased by
the economic benefits from not destroying small
businesses. Even in the short term, there can be
no justification for estate duty rates above 20%.
For income tax, the position is between
the two. The yield of income tax is high, so
unlike estate duties a very low or zero rate is
not warranted even in the long term (unless public
spending itself is reduced to Calvin Coolidge-era
levels of the 1920s, which alas is unlikely).
However, the disincentive effect is severe when a
50% all-in rate (including state and local taxes)
is passed, so even the short-term yield optimum in
the United States is almost certainly at an all-in
rate no higher than 45-49%.
Given the
differences in state income tax rates, the Laffer
Curve optimum federal income tax rate will be
higher in New Hampshire, where there is no state
income tax, than in California or New Jersey,
where the marginal rate of state income tax
approaches 10%. Overall, it's likely that the Bill
Clinton-era top federal income tax rate of 39.6%
(minus the Medicare supplement to appear in 2013)
is about the short-term optimum for the nation as
a whole, with high-tax states suffering
economically as their overall tax rate is higher
than the optimum.
However, the long-term
Laffer Curve optimum is considerably lower than
this, probably around the 28% top rate of the 1986
tax reform.
In this context, it should be
noted that tax deductions that do not affect the
marginal rate have little economic effect overall,
but simply divert activity into their chosen area.
It is thus well worth abolishing the home mortgage
interest tax deduction and the charitable donation
tax deduction, and reducing marginal tax rates
correspondingly. The state and local tax
deduction, however, affects marginal tax rates as
well as receipts, and hence is less obviously due
for the chop.
Laffer Curve theory thus has
clear recommendations as to the form a US tax
reform should take. Capital gains tax rates should
be no higher than 20% and estate tax rates no
higher than 15-20%, though there is a case for
abolishing the tax altogether. Marginal income tax
rates should be no higher than 25-28%, and
deductions should mostly be eliminated.
If
that produces insufficient revenue to feed the
state Leviathan, then a VAT can be introduced,
with a rate of no more than 10% (making federal
and state VAT/sales taxes no higher than 20%).
Less fashionably, modest tariffs, at a rate no
higher than 10% that does not greatly distort
trade, are also a possibility.
At all
costs, the left should avoid imposing punitive
income tax rates, because they will sharply reduce
tax compliance. Do that for long, and you will
find that a Russian-style flat tax of 12-15%
becomes the only effective way of raising revenue.
Martin Hutchinson is the author
of Great Conservatives (Academica Press,
2005) - details can be found on the website
www.greatconservatives.com - and co-author with
Professor Kevin Dowd of Alchemists of Loss
(Wiley, 2010). Both are now available on
Amazon.com, Great Conservatives only in a
Kindle edition, Alchemists of Loss in both
Kindle and print editions. (Republished with
permission from PrudentBear.com.
Copyright 2005-12 David W Tice &
Associates.)
Head
Office: Unit B, 16/F, Li Dong Building, No. 9 Li Yuen Street East,
Central, Hong Kong Thailand Bureau:
11/13 Petchkasem Road, Hua Hin, Prachuab Kirikhan, Thailand 77110
