Personal and corporate tax rates often take center stage in debates about the ability of states to attract top talent and new business. Indeed, a wide variation in rates exists both across space and over time. Pennsylvania, for instance, has a corporate income tax of about 10 percent whereas Washington has none, and similar differences among states exist for personal taxes. Whether and how much this uneven fiscal terrain affects the migration of workers and firms, however, is still disputed.
In a recent article titled “The effects of state taxes on the geographical location of top earners: evidence from star scientists” (National Bureau of Economic Research working paper no. 21120, April 2015), economists Enrico Moretti and Daniel Wilson take on this question by examining the effects of state tax differentials on the location choices of “star scientists,” or prolific inventors whose patenting activity places them in the top 5 percent of peers. Focusing on these highly skilled (and highly paid) workers is important, the authors explain, because their presence in a state is linked to gains in labor productivity and to clustering of industries that drive innovation.
The intuition behind Moretti and Wilson’s investigation is simple: all else equal, favorable tax environments should lure star talent either directly, through lower personal taxes, or indirectly, by attracting firms which then boost demand for skilled workers. To see how this theory stacks up against the evidence, the authors compile data on all U.S. patents filed during the 1976–2010 period, along with state-level information on corporate taxes, tax credits for R&D and investment, and income taxes for the top few percent of earners (the group that presumably boasts star scientists among its ranks).
The data reveal a strong attraction effect of low taxes on top talent—a pattern that grows over time. For a given pair of states, Moretti and Wilson estimate a tax-related relocation elasticity of about 1.7; in simple terms, that means that a 1-percent relative decline in the after-tax income of the top 1 percent of earners in one state creates a 1.7-percent increase in outmigration of star scientists to the state with more favorable tax conditions. Impacts of similar sizes also surface for corporate taxes and investment tax credits, but not for R&D credits, whose effects are smaller.
In support of a causal interpretation of their results, the authors show that changes in migration patterns lag, rather than precede, changes in taxes. Also telling is that star scientists fail to react to changes in tax rates for wages at the median, an indifference Moretti and Wilson do expect for workers in top-income brackets. As far as corporate taxes are concerned, and again in line with the authors’ causal expectations, effects operating through demand for skilled labor are observed for private sector workers but not for those in government or academia.