CAREER: Taxes and Innovation: Optimal Taxation and the Effects of Taxes on Entrepreneurs, Inventors, and Firms' Innovation

Project Outcomes Statement

Innovation and entrepreneurship are essential engines of economic growth, yet public policy plays a central role in shaping their trajectory. This NSF-funded project advances our understanding of how tax systems can best encourage innovation by entrepreneurs, inventors, and firms. Across its outputs, the project demonstrates that tax policy is not merely a tool for revenue generation but a key determinant of technological progress, business formation, and the diffusion of new ideas. By clarifying the mechanisms through which personal, corporate, and R&D-specific taxes affect innovation incentives, this work provides policymakers with empirically grounded guidance for designing fiscal systems that sustain long-term growth.

The broader social contribution of this research lies in showing how well-designed tax policies can foster entrepreneurship and invention, while poorly calibrated ones can slow innovation or drive talent and investment elsewhere. These findings inform ongoing debates about competitiveness, productivity, and the role of government in promoting scientific and technological advancement.

This project produced three major peer-reviewed papers published in leading economics journals, each illuminating a different dimension of how taxes influence innovation and growth.

In "Optimal Taxation and R&D Policies" (Econometrica, 2022) we develop a new dynamic mechanism-design framework to determine how governments should structure corporate taxes and R&D subsidies when firms differ in their research productivity, and when that productivity is private information. The paper identifies two critical constraints for policy: technological spillovers between firms and informational asymmetries that prevent governments from perfectly targeting the most innovative firms. Using firm-level and patent data, we quantify the optimal joint design of corporate taxes and R&D subsidies and show that relatively simple policy structures - such as linear corporate taxes combined with nonlinear R&D subsidies that decline at higher R&D levels - can closely approximate the theoretical optimum. This result is both analytically novel and policy-relevant, providing a tractable framework for governments seeking to encourage innovation without excessive complexity or cost.

In "Tax Simplicity or Simplicity of Evasion? Evidence from Self-Employment Taxes in France" (2024), we investigate how simplifying tax systems affects entrepreneurial behavior. Using comprehensive administrative data covering the universe of French self-employed taxpayers, we explot the introduction of simplified tax regimes to disentangle two motives for adopting them: the desire for administrative simplicity and opportunities for tax evasion. It finds strong evidence of "bunching" at regime eligibility thresholds - indicating that individuals value simplicity but that evasion incentives also play a role. We estimate a large value of simplicity at 162-942 Euros per self-employed individual per year, while also estimating a substantial evasion elasticity. This paper contributes a rigorous empirical foundation for evaluating the trade-offs between simplicity, compliance, and equity in entrepreneurship-oriented tax policy.

In "Taxation and Innovation in the Twentieth Century" (Quarterly Journal of Economics, 2022), we offer the first comprehensive empirical analysis of how personal and corporate income taxes have shaped innovation in the United States over the past century. Using a newly assembled panel of over 2.9 million inventors linked to patent data and historical state-level tax records, we find that higher personal and corporate taxes substantially reduce the quantity and location of innovation, primarily by discouraging inventor mobility and firm-level inventive activity. However, average innovation quality is largely unaffected. The estimated elasticities - around 0.8 for personal and 1.3-2.8 for corporate net-of-tax rates - reveal that tax policy significantly influences where and how innovation occurs. This long-run historical evidence powerfully connects tax structures to patterns of technological progress and regional growth.

Together, these studies provide a unified picture of how tax design affects innovation: optimal policy must balance incentives for innovation, information constraints, and administrative feasibility. The combination of theoretical modeling, structural estimation, and rich historical data advances our fundamental understanding of tax policy and innovation.

This project has also had substantial educational and mentoring impacts. Undergraduate and graduate research assistants funded by the grant contributed to model development, empirical coding, and historical data construction, gaining advanced skills in econometric methods, computational modeling, and data analysis. Several students have since entered Ph.D. programs or policy research positions, reflecting the project's contribution to building a diverse and capable research workforce.

The research outputs have directly informed new teaching materials that have been integrated into graduate and undergraduate courses and shared with other instructors, extending the project's reach to a broad audience of students. By linking frontier research to pedagogy, the project advances NSF's goals of improving STEM education and enhancing public understanding of science and economics.