Taxes, growth, and Mr. Clinton - negative impact of more taxes

Raising taxes kills economic growth. Why won't Mr. Clinton's advisors tell him?

Economists have started devoting increasing attention to the problem of economic growth - in particular, to the impact of government policy, especially taxation and spending. Their findings win give us an idea of the likely effect of the kinds of policies being proposed by the Clinton Administration.

Throughout most of the postwar period, economic growth was assumed to be purely a function of the productivity of economic inputs, such as labor and capital. This is known as the neoclassical growth model; its principal developer was Professor Robert Solow of MIT, one of Clinton's economic advisors. Much of the research on growth during this period simply involved calculating the contribution of each particular input in ever greater detail.

The weakness of the neoclassical model is that a large share of the differences in growth rates between countries cannot be explained by it. This is because in the model many of the factors we normally expect to have an impact on growth, such as capital investment, only affect the level of economic output, not the rate of growth. Consequently, the model grossly understates the role of government policy. In the neoclassical model, differences in growth rates between countries were assumed to be transitional. As technology diffused and the capital stock in Europe and Japan was restored following the destruction of World War II, growth rates were expected to converge. The persistently high growth rates in Japan and other Asian economies, however, forced economists to rethink their assumptions. Newer growth models now find important roles for policy in areas such as human capital (i.e., education) and research and development (i.e., technology).

The development of these new growth models also allows economists to investigate other areas that affect growth, such as taxation, spending, investment, and trade policies. The results of this research show clearly that high taxes and government spending, and restrictions on trade and investment, win reduce growth.

One of the first papers to question the passive role of government policy in the neoclassical growth model was written by Robert King and Sergio Rebelo and published in the prestigious Journal of Political Economy in October 1990. King and Rebelo showed that taxation has a substantial impact on long-run growth rates through its effect on the formation of capital, both human and physical. They found that a 10-percentage-point increase in the income-tax rate reduces long-run growth by 1.6 percentage points. Thus the cost of the tax to the economy vastly exceeds the amount of the tax itself (whereas the standard growth models suggest a zero impact on growth from such a tax increase).

Professor Robert Barro of Harvard has carried this line of research forward in a series of papers looking at a large crosssection of countries. Barro has found a clear negative correlation between increases in government spending and economic growth. This is because higher spending leads to higher taxes, which depress private investment. This reduction in private investment is not offset by higher public investment, which appears to have little impact on growth.

A 1991 paper published by the Federal Reserve Bank of Cleveland confirms these results. Economists Charles Carlstrom and Jagadeesh Gokhale looked at total government spending on goods and services in the U. S. between 1946 and 1989, which rose from 13.7 per cent to 22.1 per cent of GDP. They found that this increase in spending has led to a decline in output of 2.1 per cent per year.

More recently, economists Eric Engen and Jonathan Skinner of the National Bureau of Economic Research found an even higher negative correlation. Reviewing the experience of 107 countries between 1970 and 1985, they found that a 10-percentage-point increase in taxation reduced real growth by 3.2 percentage points per year in the short run. In the long run, a balanced-budget increase in both taxation and spending of 10 percentage points reduces growth by 1.4 percentage points per year.

The negative effect of expanding government is illustrated in the graph, which shows the increase in Federal Government spending as a share of GDP since the 1950s and the concomitant decline in real growth of GDP. As the graph shows, when spending averaged 18 to 19 per cent of GDP, growth averaged better than 4 per cent per year. As spending grew to more than 20 per cent of GDP, growth fell to 2 to 3 per cent per year. With spending now at more than 23 per cent of GDP, we are seeing a further decline in economic growth, which has averaged less than 1 per cent per year so far in the 1990s.

The Clinton Effect

What does this research tell us about the likely effects of the Clinton economic program? According to the Congressional Budget Office, the Clinton plan would have the effects shown in the table. As it indicates, in the first three years of the Clinton program there is no reduction in spending whatsoever - in fact, spending rises slightly - while taxes rise almost a percentage point. This suggests a decline in economic growth of about one-third of a percentage point over what would otherwise occur, based on Engen and Skinner's results. This may not sound like much, but it amounts to about 10 per cent of all the growth the Clinton Administration expected to get this year and may explain why it recently reduced its economic growth projection for 1993 down from 3.1 per cent to 2.5 per cent.