Research

This paper examines whether state minimum wage increases change how technological change affects employment and wages in the United States. Using occupation-level automation exposure measures from Webb (2020) and exploiting cross-state variation in minimum wages from 1980 to 2010, we find that higher minimum wages magnify both the employment and wage losses associated with software exposure: software-exposed occupations fared worse in states where minimum wages increased the most. The pattern differs for industrial robots. We find no comparable employment effect, but wages decline more in robot-exposed occupations where minimum wages rose more, with a wage interaction about twice the magnitude of the software estimate. Taken together, the results indicate that minimum wages do not interact uniformly with automation: wage floors amplify broad employment and wage adjustment under software exposure but mainly wage adjustment under robots. Minimum wage policy alone therefore appears insufficient to shield workers from technological change, and its labor-market effects depend on the form of automation to which workers are exposed.

This paper examines whether state corporate tax regimes moderate the labor market consequences of automation exposure. Using the occupation-level exposure measures developed by Webb (2020) for robots and software and interacting them with cross-state variation in state corporate taxes, our estimates show a modest displacement effect: higher state corporate tax burdens are associated with larger employment losses in occupations exposed to automation from 1980–2010. The tax interaction is statistically significant but economically modest: a one percentage point increase in the corporate tax rate is associated with an additional employment-share decline of about 0.02 points for each percentile increase in automation exposure for occupations more exposed to robots and software. The robot–wage interaction is small and positive, consistent with a composition channel in which low-earning workers exit high-tax, robot-exposed occupations, raising average wages among remaining workers rather than implying broad wage gains for all workers.

This paper examines whether robot and software automation exposure generates heterogeneous employment and wage effects by race and gender in the United States. Using the occupation-level automation exposure measures developed by Webb (2020) and IPUMS USA microdata, we estimate employment and wage outcomes separately for African American, White, female, and male workers. For software, we find no employment effects for any group but symmetric wage losses across race and gender. Industrial robots, by contrast, displace African American workers from employment while White workers experience no employment loss, and they generate a simultaneous contraction of female employment and expansion of male employment, with female workers facing approximately double the robot-induced wage loss of male workers. The type of automation—not automation per se—determines which dimension of labor market inequality is amplified.

States must tax something, but what they tax varies widely. This paper asks whether the composition of state taxation—capital versus labor—affects business dynamism: firm entry and the job creation it generates. We link a forty-five-year panel of state tax rates to the Census Bureau's Business Dynamics Statistics and use three complementary designs: a two-way fixed-effects state-year panel, a saturated state-sector-year design interacting capital-tax pressure with sector capital intensity, and a county panel isolating the local property tax. Point estimates carry the sign theory predicts—capital-tax pressure lowers entry and job creation in capital-intensive sectors while labor-tax pressure does not—but they are economically small and statistically fragile under wild-cluster bootstrap inference and event-study designs, implying that the effect of tax composition on aggregate dynamism is too small to detect with public, design-based data.