Download BibTex file | 

As of 2005, 31 US states offered corporate income tax credits on research and development (R&D) expenses in order to encourage more in-state innovation activities. Empirical questions about the efficacy of such tax breaks at the state level persist, in part because the complexity of the tax laws means that simple credit-rate comparisons across states do not fully capture the differential variation in effective after-tax price incentives firms face in choosing where to locate R&D activities. We are unaware of any research analysing and comparing the effective prices of R&D faced by firms, across all US states and utilising micro-level data. Using data extracted from detailed reading of individual firms' 10-K and S-1 filings and of state-level tax credit rules, we estimate the effective after-tax price of basic and qualified research expenditure each firm would have faced in each of the 50 states had they been located there. Our methodology simulates the effective tax price of each firm's marginal dollar of research expenditure, assuming the firm chose to move all of its R&D operations to each of the 49 other states. Through Monte Carlo techniques, we consider the sensitivity of our interstate comparative results to several modelling assumptions. We find significant variation in after-tax R&D prices across states with quite different R&D tax laws. Prices range from $0.176 to $0.520 on a marginal dollar of R&D in Virginia and Washington State, respectively. We also find that the interstate variability is generally more important - indeed, much wider than we had anticipated before investigating state-by-state regulations - than the inter-firm variability within states.

Full version (external link)