This is by Ethan Kaplan of the University of Maryland (via email):
Does Taxing the Wealthy Hurt Growth?, by Ethan Kaplan: What is the impact of taxation on growth? In theory, a country without taxation will have difficulty providing basic public goods such as roads and research that are fundamental for economic growth. However, many politicians and some economists argue that once basic public goods are provided for, increases in taxation have a negative impact on growth. According to this argument, this is especially true for taxes on the very wealthy, who are likely to save their income and channel that savings into entrepreneurship or other investment. Much of the argument over tax policy in the United States is focused on whether the rich should be taxed at a higher or lower rate than they are today. The argument in favor of higher rates is that income inequality is at extremely high levels and the government should focus more on redistribution and also that the rising national debt is also potentially harmful to growth. The argument against higher rates is that raising taxes on wealthy would disincentivize the people most likely to create economic growth and thus jobs. In a climate where jobs are scarce, the argument goes, this is a particularly bad economic idea.
This debate, however, is largely based on ideology rather than evidence. Unfortunately, it is quite difficult to figure out the impact of taxation on growth. Changes to the tax codes usually pass Congress when other things are happening to the economy. For example, the 1982 tax cuts, which dropped the top marginal tax rate from 69% to 50%, were passed towards the end of a large recession. Moreover, the impact of taxes on growth can change over time as the economy changes.
Nevertheless, looking at the raw correlation between top marginal tax rates and growth can be helpful for getting a rough sense of the likely impacts of higher taxation on growth. One recent paper by Pikkety, Saez, and Stantcheva looks at the correlation between top marginal tax rates and growth and finds the growth is higher when top marginal tax rates are higher. I restrict myself to the historical experience of the United States and go back to 1930. In particular, I took real chained per capita GDP growth from 1930 to the present from the Bureau of Economic Analysis' (BEA) website. The correlation over this period between the top marginal tax rate and output growth is strong and positive as can be seen below:
A rise in the top marginal tax rate from 0 to 100 percent is correlated with a rise in per capita growth of 5.85 percentage points per year. One reason that this simple correlation might overstate the impact of the marginal tax rate on growth is that the top growth years were in the early 40s when the government was spending heavily and when the country was finally recovering from the Great Depression. If we look only at the post war period (after 1946), a rise from 0 to 100 percent in the top marginal tax rate is associated with an increase of only 2.69 percentage points of growth. Moreover, the statistical significance of the relationship becomes marginal, as the p-value rises from 0.017 to 0.122. On the other hand, if we look at the time period encompassing 1960 to the present, a rise in the top rate from 0 to 100 percent is correlated with a rise in per capita growth of 3.03 percentage points of growth per year, and the relationship becomes more statistically significant (with a p-value of 0.064 percent). Finally, if we look only at the years since 1980, a rise from 0 to 100 percent in the top marginal tax rate is associated with an increase in growth of 3.87 percentage points. In this case, the relationship is statistically insignificant (with a p-value of 0.392 percent), in part because the sample size is small.
While we cannot say that there is a robust significant positive relationship between tax rates and growth, it is still interesting that regardless of when we start the sample, higher top marginal tax rates are associated with higher not lower growth. Moreover, a narrative reading of postwar US economic history leads to the same conclusion. The period of highest growth in the United States was in the post-war era when top marginal tax rates were 94% (under President Truman) and 91% (through 1963). As top marginal rates dropped, so did growth. Moreover, except for 1984, a recovery year, the highest per capita growth rates since 1980 were all in the late 1990s, after the top marginal tax rate had been increased from 28% under President Reagan to 31% under the first President Bush and then 39.6% under President Clinton. One possible reaction to this finding is that what matters more than the top marginal tax rate on income is the capital gains tax rate but growth has also been higher when the capital gains tax rate has been higher.
So, what does this tell us? Of course, it would be silly to make the argument that increasing top marginal rates from 0 to 100 percent increased per capita growth by almost 6 percentage points per year. No doubt there are other factors that could confound the relationship between tax rates and growth. However, the changes in top marginal tax rates over the period are quite large so it seems likely that if raising top marginal rates did have a large negative impact on growth, we should be able to see it in the correlations. Thus, it also seems silly to argue that higher taxes on the rich have a large negative impact on growth, given that historically growth is, if anything, positively correlated with the top marginal rate.
What does this mean for public policy? Given the large rise in inequality in the United States over the past 40 years, if the historical evidence tells us that it is unlikely that taxing the wealthy has a large negative impact on growth (and it might even have a positive impact), shouldn't we increase rates on the wealthy from their current top rates of 35%?
p.s. the data used to analyze the time series is available on my website: econweb.econ.umd.edu/~kaplan