CRS: Nothing Affects Economic Growth
September 17, 2012
A study by the Congressional Research Service (CRS) is getting a lot of attention, because it finds that tax cuts are not associated with economic growth. Although less reported, the study also finds nothing is associated with economic growth, including all the standard factors such as education, population growth, and government spending. That’s because a) economic growth is notoriously hard to pin down, with many factors difficult or impossible to measure, and b) the CRS methodology has no hope of dealing with this complexity.
First, the CRS study looks at just one country, the U.S., when researchers have at their disposal excellent data from many countries, providing more variation in conditions, tax rates, and economic growth while taking into account massive worldwide events like World War II and its aftermath. Trying to attach annual changes in U.S. GDP in the 1950s to tax rates completely misses the fact that the U.S. emerged relatively unscathed after World War II relative to our competitors, and the U.S. was in a place to capitalize on decades of pent up demand and innovation resulting from the world wide Depression. Researchers that do such cross-country panel studies, such as the OECD, find that personal income taxes are second only to corporate income taxes in their harm to the economy. However, empirical difficulties remain, and many such studies fail to find a significant relationship.
Second, the CRS looks only at personal income tax rates and capital gains tax rates, and tries to relate those to investment, productivity, and growth. The biggest tax on investment is the corporate income tax, which CRS ignores. The capital gains tax, and the dividend tax, is just one additional layer, along with the estate tax. Furthermore, business credits, deductions, and depreciation schedules are important as well. Ultimately one would want effective tax rates, if that could be accurately measured over time, or the cost of capital.
Third, CRS takes no account of the long term effects of tax policy. Their regressions relate one year tax changes to one year growth changes, when in fact saving and investment often takes years and even decades to come to fruition. CRS relates tax rates to investment and saving as a share of GDP, but this too is flawed since it ignores changes in GDP. That is, lower taxes on saving and investment should increase saving and investment, and increase GDP, such that the ratio between them does not change significantly.
Ultimately, the only significant result CRS finds relating to economic growth is that labor productivity increases with more government transfer payments. That is at odds with virtually every study out there, casting further doubt on this approach.
CRS does find that tax cuts are associated with increasing income inequality. But that too is a result of the 1 year time horizon. Longer term, it appears there is no relationship between tax rates and inequality, at least not over the range of tax rates we’ve seen since the 1980s.
The slow-growth Bush-Obama years do provide a seeming difficulty for folks like us, who argue for lower taxes on saving and investment. However, the Bush tax cuts are a political brew that no one in their right mind would argue strictly targets economic growth. Many low-income provisions, such as the 10 percent bracket and the expansion of the child credit, were meant to increase progressivity of the tax code, and they did. These same low-income provisions were also found to have actually decreased economic growth, according to a Treasury study. The Treasury found that the most growth enhancing aspect of the Bush tax cuts was the lowering of the top four marginal rates on personal income, followed by the lowering of tax rates on capital gains and dividends.
Follow William McBride on Twitter @EconoWill