Do Non-Compete Agreements Reduce Innovation?

Reinmuth and Rockall (2025) examine whether stronger enforcement of non-compete agreements reduces innovation. They analyze U.S. state-level changes in enforceability from 1991–2016, using patent data from the USPTO, firm data from Compustat, and business formation data. They find that increases in enforceability significantly reduce innovation, with patenting declining by about 14% after five years for an average policy change. The negative effects are larger for more novel patents and venture-backed firms. They show this decline is driven primarily by reduced inventor mobility and weaker knowledge diffusion, rather than changes in firm entry or shifts away from patenting.

Why This Article Was Selected for The Policy Scientist

This article addresses a central policy question: whether restricting worker mobility alters the production and diffusion of innovation across an economy. The topic is especially salient given ongoing federal and state debates over banning or limiting non-compete agreements, placing the study directly within current policy design discussions. Reinmuth and Rockall have contributed to a growing empirical literature on labor mobility and innovation, and this paper extends that work by offering economy-wide evidence rather than industry-specific findings. Using comprehensive patent and firm-level datasets, the analysis is well-grounded, and the multi-source data improves credibility. The quasi-experimental design based on policy variation represents a meaningful causal inference approach.

Full Citation and Link to Article

Reinmuth, K., & Rockall, E. (forthcoming). Innovation through inventor mobility: Evidence from non-compete agreements. American Economic Journal: Applied Economics. https://www.aeaweb.org/articles?id=10.1257/app.20240804 

Central Research Question

This study examines whether stronger enforcement of non-compete agreements (NCAs) affects innovation and, if so, through what mechanisms. The central question is twofold: what is the net causal effect of changes in NCA enforceability on innovation, and which channels—incumbent incentives, firm entry, or labor mobility—drive that effect. The authors explicitly position the inquiry within a broader theoretical ambiguity. On one hand, NCAs may increase innovation by protecting firm-specific knowledge and strengthening incentives for investment in research and development. On the other hand, they may reduce innovation by restricting worker mobility and limiting the diffusion of knowledge across firms. The study seeks to resolve this ambiguity using causal empirical evidence drawn from policy variation across U.S. states.

Previous Literature

The article engages with several strands of literature, including research on labor mobility, innovation spillovers, and intellectual property protections. Prior studies have examined specific channels—such as firm entry, wages, or industry-level innovation—but have produced mixed and sometimes contradictory findings. For example, some work suggests that stronger NCAs increase firm investment in physical capital or certain types of innovation, while other studies find reductions in mobility, spinouts, or exploratory research. A longstanding hypothesis in the legal and economics literature posits that restricting labor mobility enhances innovation by allowing firms to better appropriate returns. However, competing work emphasizes the role of knowledge spillovers, particularly in regions like Silicon Valley, where weak enforcement of NCAs is associated with high innovation. This study contributes by providing comprehensive, economy-wide causal estimates and by explicitly testing the mechanisms underlying observed effects, rather than focusing on isolated channels or industries.

Data

The analysis draws on multiple large-scale datasets to capture different dimensions of innovation and economic activity. The primary measure of innovation is patenting activity from the U.S. Patent and Trademark Office (USPTO) PatentsView database, covering granted patents with detailed information on inventors, assignees, citations, and technological fields from 1991 to 2016. These data allow for the construction of state-year panels of corporate patenting and enable analysis of inventor mobility across firms. To measure firm-level performance, the authors use Compustat data on publicly traded U.S. firms, which provide information on revenues, inputs, and productivity. Entry dynamics are captured using the Census Bureau’s Business Formation Statistics, which track new business applications at the state level. The key independent variable is an index of NCA enforceability, derived from legal sources and extended over time to capture state-level policy changes. This index is normalized to facilitate interpretation and reflects both judicial rulings and legislative actions affecting enforceability.

Methods

The empirical strategy relies on quasi-experimental variation in state-level NCA enforceability over time. The authors implement a local projections difference-in-differences (LP-DiD) framework that accommodates staggered and continuous treatment effects. This approach compares changes in innovation outcomes before and after policy shifts, relative to states that do not experience contemporaneous changes, while incorporating time fixed effects and restricting the sample to “clean” treatment and control observations. The identification strategy assumes that changes in enforceability—driven by court decisions and statutory reforms—are plausibly exogenous to underlying innovation trends. The authors supplement the main analysis with a case study of a major judicial change in Ohio, using a synthetic control approach to illustrate the immediate effects of increased enforceability. Additional analyses examine heterogeneity across patent types, technological fields, and firm characteristics, as well as alternative outcomes such as productivity. Robustness checks address potential confounders, including substitution between patents and trade secrets and cross-state migration of inventors.

Findings/Size Effects

The results indicate a consistent and economically meaningful negative effect of increased NCA enforceability on innovation. In the baseline specification, an average increase in enforceability reduces patenting by approximately 14% within five years. Larger changes in enforceability produce even more substantial declines, with estimates suggesting reductions of up to 28% when moving from median to high levels of enforcement. These effects are not uniform across all types of innovation. The decline is significantly more pronounced for high-value and novel patents, including those with fewer backward citations, those in younger technological fields, and those produced by venture capital-backed firms. In some cases, the magnitude of the effect on these subsets approaches or exceeds 20–30%. Complementary analysis of firm-level productivity shows a decline of roughly 6% in total factor productivity after five years, with peak effects reaching approximately 9%, indicating that the impact extends beyond patenting to broader economic performance.

Mechanism analysis reveals that the negative effect is not driven by reductions in firm entry or by firms substituting away from patenting toward trade secrecy. Instead, the evidence points to reduced labor mobility as the primary channel. Increases in NCA enforceability lead to fewer inventor moves between firms and greater reliance on internal knowledge, as reflected in citation patterns. These findings support the interpretation that restricted mobility limits knowledge diffusion and spillovers, which in turn reduces innovation. The absence of significant cross-state migration effects further suggests that the observed declines are not offset by relocation of inventive activity.

Conclusion

The study provides robust causal evidence that stronger enforcement of non-compete agreements reduces innovation, primarily by constraining labor mobility and limiting knowledge diffusion across firms. While NCAs may enhance firms’ ability to appropriate returns from innovation, these benefits appear to be outweighed by the broader costs associated with reduced spillovers. The findings extend beyond patent counts to include declines in firm productivity, reinforcing the conclusion that the effects are economically substantive. By leveraging policy variation across states and employing a rigorous quasi-experimental framework, the authors contribute to a more comprehensive understanding of how labor market institutions shape innovation. The results have direct implications for ongoing policy debates, suggesting that restrictions on worker mobility may carry significant trade-offs for long-run technological progress and economic growth.