Except
for my very
first posting, I have avoided political subjects in this blog, a result of
my newfound apolitical stance.
However, I am at least temporarily re-entering the political arena to
attack what I call the Mantra: the conservative incantation that high taxes,
especially in the upper brackets, reduce GDP growth.
In fact, massive data indicate that the
effect is in the other direction: high tax rates, at least in the top brackets,
correlate with increased GDP
growth. That may seem
counter-intuitive, so let me present some of the data. Table I shows the top marginal federal
tax rate on ordinary income for the one hundred years since we have had an
income tax; I use the top rate as a proxy for all higher brackets since they
move up and down together and because the topmost rate is the one most often
challenged by the Mantra.
 |
|
Table I: Top Marginal
Federal Tax Rates by Year
[Source:
Tax Policy Center: Urban Institute and Brookings Institution.
Highlighting and columns
showing presidencies from 1947 on are added.]
|
The
highlighting in Table I shows the very few periods, totaling 16 years, when the
top rate was less than the 35% rate of the past ten years: the four initial
years of the income tax before our entry into World War I (7-15%); the end of the Roaring
Twenties (24-25%; we all know where that led!); the last year of the Reagan
presidency (28%); and all four years of the Bush-I presidency (28-31%). For another 74 years, the top rate exceeded that of the last ten years,
in most of those years being 50% or higher.
Next, Chart I shows the ten-year
trailing average of GDP growth. The curve's leftmost point is the average
annual GDP growth from 1948-57; its rightmost point is the average from
2002-11. For the moment
disregarding the year-to-year fluctuations and looking only for broad trends,
what is remarkable about this chart is that the early years show GDP growths
distinctly higher than the later years.
That is exactly the reverse of what the Mantra advertises, because top
rates before 1982 were a high 70-91%, and those after 1982 were a much lower
28-50%. Especially telling, I
think, is the plummeting of GDP growth since the Bush II tax cuts of ten years
ago.
 |
|
Chart I: Ten-Year Trailing-Average GDP Growth
[Source: Key Trends in Globalisation] |
Finally,
Chart II gives more detail for recent years, 1986 to present. Each point on the curve gives the
five-year forward average of GDP growth
rather than a ten-year trailing average.
The chart additionally notes the years of tax increases and decreases
during the period. In line with
our previous observations, GDP growth more often than not increased after an
increase in taxes and decreased after a decrease in taxes, not the reverse.
 |
|
Chart II: Five-Year
Forward-Average GDP Growth
[Source: New York Times, September 15, 2012] |
What's going on? In an effort to drill down more finely
into the data in Table I and Chart I, I constructed a scatter plot from
them. For each year starting with
1957, I computed the ten-year trailing average of tax rates in Table I, and
paired it with the same year's ten-year trailing average of annual GDP growth
from Chart I, each pair then defining a point in Chart III.
 |
| Chart III: Ten-Year Trailing Averages: GDP Growth vs. Top Tax Rates |
Chart III corroborates the qualitative
conclusions I have noted. By
scanning the array of points, which generally sweep upward and to the right,
you can easily see how higher top tax rates and higher GDP growth rates tend to
go along with each other. The
trend line shown, which best fits the data points, captures this effect. (For those with a statistical bent, the
correlation coefficient of the data, which measures the degree of their
clustering around the line, is +0.50; if it had been +1.0, all the points would
have lain exactly on a straight line with a positive slope.)
Although the slope of the best-fit line
is modest (showing on average only a 0.1 percentage-point increment of GDP
growth rate for each 6.5 percentage-point increment in top tax rate), some
pairs of points tell a different
story. For example, the best
decade's average GDP growth rate exceeded the worst decade's by a whopping 3.25
percentage points, even though the average top tax rate during the former
decade was 2.3 times larger than that
during the latter. No pair of
points shows anything like that disparity in the opposite direction.
High taxes on ordinary income are not
the only ones denounced by the Mantra.
There are also capital gains (CG) taxes, which are now lower than at any time except 1912-15 and 1922-33. A similar analysis for them yields comparable results,
shown in Chart IV. The trend line
has a slightly steeper slope, showing on average a 0.25 percentage-point
increment in GDP growth rate for each 6.5 percentage-point increment in the top
CG tax rate, but the data are more scattered around the line, with a weaker
correlation coefficient of only +0.26.
There is much the same large disparity between the best and worst
decades' data points.
 |
| Chart IV: Ten-Year Trailing Averages: GDP Growth vs. Top CG Tax Rates |
By whatever measure, then, analysis
discredits the Mantra. GDP growth
rates tend to go up with increasing top tax rates, rather than the other way
around as the Mantra contends.
Of course, I've only established a
positive correlation, not causation; other factors affecting GDP growth may
explain the correlation between it and high tax rates. (For example, hot weather may cause ice
cream to melt and lawns to die, and analysis would therefore show a strong
correlation between the two effects; but no one would say that dying lawns
cause ice cream to melt or vice versa.) However, I strongly argue that there is a causal link at work here, especially
nowadays.
In my estimation, as low top tax rates
help siphon ever more of the nation's wealth to the wealthiest, that excess
wealth hasn't been chiefly deployed for GDP-enhancing activities like building
businesses, as conservatives claim.
Rather, I think much of it has been diverted to personal ends that do
little to enhance and sometimes destroy GDP growth. I call them the three Bs.
The least pernicious of the Bs are the baubles, the accoutrements of wealth, like übercars (mostly made abroad, barely affecting U.S. GDP),
lavish vacations (again, largely abroad), masterworks of art and rare antiques
(minimal effect on GDP), and extravagant mansions (mildly stimulative). Far worse are the bundles of assets sequestered in offshore tax havens, which
have no effect on the U.S. economy.
Worst are the bubbles that
excess wealth stokes when it rashly chases profits by risky investing—such fads
as mortgage-backed securities and other derivatives—massively destroying GDP
when the bubbles burst, as in 2008.
On the other hand, when in the past
higher top tax rates have restrained the diversion of excess capital to the
wealthiest, that capital has usually been deployed much more productively on
infrastructure and other core governmental activities. A key example is the construction of
the interstate highway system, started under President Eisenhower in 1956 and
continued for two decades thereafter with tax money received when top
ordinary-income tax rates were 70-91% and CG rates were 25-35%. It was an
enormously stimulative project, no doubt contributing greatly to the bulge in the ten-year trailing average GDP growth rates from 1966-75 shown in Chart I.
When tax rates went down to today's levels, starving the government of
needed revenue, such infrastructure projects as well as core government
activities like education began to suffer, to the detriment of GDP growth.
Government is not the enemy depicted by
conservatives; it is an integral and vital part of a healthy economy. Sufficient taxation is required for it
to accomplish its necessary role, which in its fulfillment also increases GDP,
to say nothing of helping balance the budget.
To paraphrase a certain
Scotswoman: Out damn'd Mantra! out
I say!