UNR Economics Working Paper Series
Working Paper No. 06-008
Does the Party in Power Matter for Economic Performance?
Department of Economics /030
University of Nevada, Reno
Reno, NV 89557-0207
(775) 784-6850│ Fax (775) 784-4728
In this brief paper, I consider whether five common political beliefs have any basis in fact. Does
the economy grow faster when Republicans are in charge? Does the size of the government
actually keep expanding? If so, is this growth correlated with Democrats being in charge? Does
bigger government lead to slower growth? Finally, is it accurate to characterize Democrats as
the “tax and spend” party? While correlation is not causation and theoretical relationships are
complex, the data on U.S. economic performance during the postwar period does not appear to
support any of these beliefs, and in fact tends more to support the alternative hypotheses.
JEL Classification: H00, H50, H60
Keywords: deficits, government spending, economic growth, political parties
A significant number of voters in the United States appear to believe that the economy has
historically performed better when the Republican party is in charge of the country. The Club
for Growth, for example, supports Republican candidates, especially those who endorse reducing
the size and economic role of government, on the assumption that these policies favor faster
economic growth. At least since the Truman Administration, Republicans have been very
effective at convincing many voters that the Democratic party follows policies that expand the
role of government, and that these policies reduce economic growth. But just as 2004 polls
showed that a large number of voters believed that Saddam Hussein’s government participated in
the terrorist attacks of 9/11, the commonly held beliefs about party policies may make perfect
sense, but are nonetheless unsupported by any evidence.
In this brief paper, I review standard government statistics on economic performance
during the postwar period in order to determine whether these beliefs have any basis in fact. The
Republicans may claim to be more competent at economic management, but does the economy
grow faster when Republicans are in charge? Does the size of the government actually keep
expanding, and if so are Democrats the cause? Does bigger government lead to slower growth?
Is it accurate to characterize Democrats as the “tax and spend” party? To those who have never
considered the data before, the results may seem surprising.
Does the economy grow faster when Republicans are in charge?
The Department of Commerce’s Bureau of Economic Analysis (BEA, 2006) has
collected annual data since 1929 for the National Income and Product Accounts. If we use these
data through 2005 to compare economic performance during Democratic and Republican
administrations, the contrast, as shown in Table 1, is dramatic: the economy has grown
significantly faster under Democratic administrations, and more than twice as fast in per-capita
Including the Great Depression and the Second World War in the comparison period is
likely, however, to distort the results. Thus, for the rest of this paper I will use data since 1949,
after President Truman’s first term. Furthermore, additional data, such as monthly
unemployment data from the Bureau of Labor Statistics (BLS, 2006), is available after 1948.
A brief review of the data for the Clinton Administration and the current Bush
Administration shows striking differences in performance. Though there is disagreement about
the causes, the economic data strongly supports that the economy performed better during the
Clinton Administration than under the current Bush Administration, as illustrated in Table 2.
Under President Clinton, real GDP grew faster on average than under President Bush, even
though we use data through the end of 2005 in order to include the recent economic recovery,
and the contrast is even greater if we adjust for population growth. The difference in the average
unemployment rate was small, but the unemployment rate fell from 7.3 percent to 4.2 percent
during the eight years of the Clinton Presidency, an average of 0.4 percent per year, while under
President Bush unemployment rose to 6.2 percent by mid-2003 before falling back to 4.7 percent
as of the most recent data for April, 2006.
Is the current Bush administration an anomaly? As Table 3 shows, data from 1949 to
1992 suggest that even before 1993, the economy did not perform better under Republican
administrations. Growth was higher under Democrats, while unemployment was, on average,
lower and falling.
In Table 4, I combine the variables discussed above for an overall comparison, together
with several additions. Adjusting for inflation with the Consumer Price Index (CPI), real weekly
wages (available only through 2002) and the Dow Jones Industrial Average both grew faster on
average under Democrats, even when we use logarithmic growth rates for the latter to adjust for
its variability.1 The BLS’s index of multifactor productivity (available through 2001) grew
under Democratic administrations at twice the rate as under Republican administrations. As a
share of GDP, even corporate profits were higher under Democrats, most likely because the
economy did better.
Are these differences merely random? Table 4 also shows t-statistics, and in general a t-
statistic greater than 2 (or less than -2) implies that the difference is statistically significant from
zero at the 95 percent level of confidence (as marked with “*”). A positive (or negative) t-
statistic means that the value is higher (or lower) for Republicans.2 The differences in growth,
unemployment, and the corporate profit share are all statistically significant, and support the
argument that the economy may actually perform better under Democrats. The differences in
weekly earnings, stock market growth, inflation, and multifactor productivity all favor the
Democrats as well, but these differences are not statistically significant.
Is there a lag effect? It is a reasonable argument that economic performance early in a
new administration is likely to be the result of policies followed by the prior administration. For
example, many economists believe that the recession of 1981-82 was caused more by the change
in monetary policy enacted by Federal Reserve Board Chairman Paul Volcker, a Carter
appointee, rather than any specific policy of President Reagan. I therefore tested whether
lagging the effect of the administration on growth might support the argument that the economy
actually performs better under Republicans. For quarterly growth rates of real GDP, the results
are shown in Table 5 for up to sixteen quarters, along with the t-statistics.3 Lagging up to three
quarters, the Republican effect remains negative and significant. After a year the lagged effects
become insignificant, and then become positive for 6 to 10 quarters, though the difference from
zero is small and insignificant. After 10 quarters, the lagged effect remains insignificant but
goes negative again. Even with up to four years of lagged effects, there is thus no evidence that
the economy performs better under Republicans.
Is it the Administration or Congress that matters for economic growth? Because
Congress writes the laws and sets the budgets, it is reasonable to argue that the Legislative
Branch is more responsible for the state of the economy than the Executive. To test this, I
estimated the different effects on real growth rates of Republican control of each house of
Congress as well as for a Republican administration. These results were not significantly
different from zero. I then tested for the effect of having both houses and the administration
under simultaneous control of either party, relative to the effect of simply having a Republican
administration. The effect of having Democratic control of government on growth was positive
on growth and the effect of having Republican control of government was negative, but neither
effect was statistically significant. The party of the administration is a better explanatory
Statistical correlation does not scientifically prove causation, and these results do not
prove that the economy does better under Democrats. But we can reasonably conclude that these
government statistics provide evidence that directly contradicts the argument that the economy
does better on average under Republican administrations. With lagged effects and other causes
considered, the difference may be insignificant, but the economy may actually perform worse
Does the size of the government keep expanding, and are Democrats the cause?
In 1929, purchases by the Federal government accounted for only 2 percent of GDP, half
of which was spent on national defense, and by 2000 this share had grown to around 7.5 percent
of GDP. Total current spending by the Federal government grew much faster, from 2.5 percent
of GDP in 1929 to 19 percent of GDP in 2000. Together with state and local governments, total
government spending rose to almost a third of GDP, and many voters assume the trajectory of
government spending is ever upwards.
The perception of continuously increasing government spending is not entirely accurate,
however, as Figure 1 helps to illustrate. The role of the federal government grew significantly
during the Roosevelt Administration, first as a result of the efforts to deal with the effects of the
Great Depression, and then from the rapid military spending of the Second World War. Once the
postwar economy stabilized, however, the federal government’s share of the economy grew
much more slowly. In a paper prepared for the Joint Economic Committee, Gwartney, et al.
(1998) noted that between 1960 and 1996, government grew more slowly in the United States
than in any other developed nation (i.e., the member states of the Organization for Economic
Cooperation and Development). Of these developed countries in 1996, the United States also
had the smallest relative size of the combined federal, state, and local governments.
In relative terms, the size of the federal government has certainly been higher than at the
end of the Clinton Administration. In 1975, federal expenditures peaked at 21 percent of GDP,
and then fell during the Carter years. Federal spending peaked twice at 23 percent, first during
the Reagan Administration’s military buildup and next during the first Bush Administration.
Overall, federal expenditures increased as a share of GDP by an average of 0.2 percent per year
between 1949 and 1975, but between 1975 and 2000 this share declined by an average of 0.1
percent per year. The federal government also employed less of the labor force. After 1955,
federal employment rose more slowly than the overall labor force, falling from 4.3 percent of the
labor force to 2.8 percent in 1990. During the Clinton Administration, the number of federal
workers actually declined, and as a share of the labor force it fell to 2.1 percent by 2000. At the
same time, state and local government spending stabilized after 1975 to between 10-11 percent
of GDP, while the state and local government’s share of the labor force slowly declined.
How do the political parties compare on the growth of federal expenditures? During the
Clinton Administration, federal spending fell from 23 percent of GDP to 19 percent, while
during the first five years of the current Bush Administration, the share of federal spending rose
by 1.5 percent. For the entire postwar period, as Table 6 shows, federal expenditures were
higher on average under Republicans, and the difference was statistically significant. As a share
of GDP, federal expenditures rose by an average of 0.3 percent per four-year term. Federal
expenditures grew slightly faster during Republican administrations, but the difference was not
One may nonetheless object to the conclusion that the size of government has not
increased more under Democratic administrations, since Democrats (Roosevelt and Johnson) are
responsible for the creation of both Social Security and Medicare, two of the largest components
of federal spending, along with national defense spending and interest on the national debt. It is
difficult to find objective measurements of these future obligations for comparison, but the
argument is valid, even if the partisan comparison has been weakened by the recent unpaid
expansion of Medicare drug benefits. Clearly, these current expenditures, like interest on the
national debt, are not discretionary. Both Social Security and Medicare are programs created by
Democrats that carry an implicit future debt, and though these programs were designed to be
self-sustaining, higher life expectancies, rising medical costs, and slowing birthrates have made
them harder to sustain under current projections. But to the extent that society wants these
popular programs, government needs to figure out how to pay for them, and to be fiscally
responsible we must either increase contributions, reduce benefits, or increase government
savings to prepare for the future.
Does bigger government lead to slower growth?
Underlying the concern over the size of the government’s share of the economy is the
belief that a larger role for government reduces economic growth. Most economists would agree
with this argument if it represents an extreme form of government which dominates all aspects of
economic affairs. In this scenario, the private incentive is quashed, productivity is stagnant,
incomes are low, and growth is fueled only by high rates of investment through forced savings.
But does such a description actually apply to the United States?
Barro (1990) has made a persuasive case that the relationship between the size of
government and economic growth is shaped like an inverted-U, with low growth resulting from
both too little and too much government. That is, both too little government and too much
government lead to slower rates of economic growth, but there is optimal size of government that
is not zero. Gwartney, et al. (1998) agreed, though they considered the size of government in the
first half of the 1960s to be the optimal level and argued that income would have been 20 percent
higher in 1996 had government stayed the same relative size.
Is the United States on the downward-sloping side of the inverted U? That is, is the
marginal effect of growth in government bad for economic growth? For the economies of
Western Europe, most economists would agree that the marginal effect of government on growth
is clearly negative, but the evidence is weaker that this is also true for the United States.
Gwartney, et al., for example, chose to make their case using the period 1960-65 as their growth
baseline, a half-decade with the highest growth rates in postwar American history and, until the
Kennedy tax cuts went into effect in 1964, a period characterized by a maximum marginal tax
rate of 91 percent on unearned income. They then compared this to the period 1990-95, a half-
decade which began with a recession, to argue that growth in government reduced economic
growth. Yet even as they wrote their paper in 1998, the economy was beginning to once again
grow at a rapid pace. Their comparison was thus not fairly representative, and their estimates
need to be taken with a grain of salt.
In considering the historical experience for the U.S., using data provided by Maddison
(2003) and summarized in Table 7, I find that the average growth of real GDP did decline over
time, which lends credence to the argument that the rising size of government after FDR slowed
economic growth. But population growth rates were higher before the Second World War, and
in fact the average real growth rate per capita was faster after 1950, when the government was
significantly larger. Furthermore, the standard deviation of growth was smaller, suggesting that
the growth of government was not only correlated with faster growth, it was also correlated with
more economic stability.
Certainly there are many ways in which government can reduce economic growth.
Private business is usually much better than government at providing incentives to work harder
and smarter, and a good government must recognize and rely on it. Unless they are used to
discourage inefficient behavior, taxes and subsidies can distort private incentives, and
government spending can crowd out the private sector. Free markets work very well much of the
time, as long as private firms are competitive, buyers and sellers have enough information about
what they are buying so neither is taken advantage of, and third parties are not affected. But
these conditions are not always met by private markets without government oversight, and there
are many types of “public” goods that private markets cannot provide efficiently, if at all. These
public goods may include such things as transportation infrastructure, social insurance, financial
regulation, basic scientific research, and education, and even perfect markets do not necessarily
result in a fair distribution of income. Government can help create a stable climate for private
investment in which people trust that their investments will be protected and have faith that the
future will be brighter. Certainly government could often perform some of these functions
better, but if less government was always good for the economy, then Somalia would be one of
the world’s wealthiest countries instead of one of its poorest.
Is it accurate to characterize Democrats as the “tax and spend” party?
While current federal expenditures did not grow faster under Democrats, the data do
show that current federal receipts have grown significantly faster under Democrats. In Table 8,
federal receipts are shown to be higher under Democratic administrations, and as a share of GDP
they increase, on average, under Democrats and decrease under Republicans. These differences
are statistically significant.
Is there a good explanation for the increase in tax receipts under Democrats? While
Republicans would offer their own interpretation, one could easily argue that Democrats have
behaved with more fiscal responsibility. The average federal budget deficit has been
significantly larger under Republicans, and smaller under Democrats. Looking at rate of
changes, the deficit tended to expand under Republicans and shrink under Democrats. Indeed, it
is well known that the Clinton Administration’s first major act was to increase taxes on the
wealthiest Americans to help balance the budget, a tax increase that passed the Senate by a 51-50
vote and left many Republicans warning that an economic collapse would follow.
However, an even better explanation is that tax receipts rise with growth, and Democratic
administrations have tended to preside over faster growth. Figure 2 shows the relationship
between federal receipts and GDP since 1929, adjusting for inflation and population growth. For
the most part, the relationship between the two appears to be almost constant, with an estimated
elasticity significantly greater than one. Not only do tax receipts rise with GDP growth, but real
tax receipts per capita rise at a faster rate than real GDP per capita.