Tax incentives, R&D and productivity

This thesis explores the causal relationships between tax incentives, research and development (R&D) and productivity. Using R&D survey data from the United Kingdom (UK) Office for National Statistics and administrative data on corporation tax returns from HM Revenue and Customs, I first conduct empirical analyses of tax incentive policies for R&D, and then estimate the elasticity of output with respect to firms' own R&D efforts as well as external R&D performed by neighboring firms in technology and product space. In the first two chapters which focus on tax incentive policies and their evaluation, I am able to identify the policy effect of interest by exploiting two significant reforms in the UK in 2002 and 2008. I find that tax incentives had a positive and significant stimulating effect on businesses' R&D spending. I argue that the availability of a quasi-experimental set up helps in better identifying the policy impact. The production function estimation exercise in the third chapter shows that double counting of R&D human resources and materials in the production function causes the elasticity of output with respect to the firms' own R&D to be substantially underestimated. I also find that the R&D done in multi-unit enterprise groups is productive for the production facilities which themselves do not perform R&D. The Jaffe (1986) and Bloom et al. (2013) measures of external R&D, which account for closeness of firms in technology and product space can be constructed and included in the production function in the spirit of Griliches (1979). I find that the point estimate for the elasticity of output with respect to firms' own R&D is around 3 percent and statistically significant. Evidence is mixed regarding the productivity effects of R&D carried out by competitors in the product market or neighboring firms in technology space. The detailed data sets used in this study offer valuable resources for empirical work on R&D and productivity.