Tax Incentives for Research and Development: Theory and Evidence
- Date: 04/18/2006
Ken McKenzie (University of Calgary)
University of Calgary
R&D differs from other types of investments for several reasons. One is that there are several market failures thought to be associated with R&D – knowledge spillovers, business stealing effects, asymmetric information in financial markets, etc. Various aspects of these market failures have been well investigated in the literature and the upshot is that government subsidies for R&D are typically thought to be justified in order to overcome these market failures and encourage more innovation in the economy. A typical way of delivering these subsidies is via the tax system. Another aspect of R&D that has not received much attention is that unlike physical capital R&D is intangible in nature (knowledge) and created within the firm using other inputs: labour (scientists), materials (test tubes), machinery (microscopes), and buildings (labs). Each of these inputs differ not only in their characteristics (current vs. capital, but in their tax treatment.
In this presentation I will outline a methodology for measuring the tax adjusted user cost of intangible R&D capital which takes these characteristics into account. Unlike the standard Jorgenson-King-Fullerton (JKF) approach to measuring the user cost of capital, and the related notion of the marginal effective tax rate on capital, the methodology takes explicit account of the microeconomic foundations, and flow-input--flow-output characteristic, of R&D. Illustrative calculations are presented for Canadian provinces which show that relative to the methodology developed here the standard JKF approach substantially overstates the tax subsidy offered to R&D. Moreover, the variation in the effective tax/subsidy rates for R&D in Canada’s ten provinces is substantial. It is shown that while a sizable tax subsidy for R&D exists in every province, the variation across provinces is significant, ranging from an effective subsidy rate of about 40% in Alberta to over 200% in Quebec.
I will also discuss how the extent to which direct tax subsidies that lower the user cost of undertaking R&D (the ''push'' effect of the tax system) and the overall competitiveness of the general production tax system (the ''pull'' effect) independently impact upon aggregate R&D intensity across countries. The ''push'' effect of direct tax subsidies is measured by the after-tax user cost of R&D capital, and the ''pull'' effect of the production tax regime is measured by the effective tax rate on marginal production costs (ETRMC), which aggregates the marginal effective tax rates on production inputs (labour and capital) into an effective excise tax rate. Following the presentation of a simple mathematical model motivating the empirical investigation, a panel data set of nine countries over nineteen years is used to estimate a dynamic fixed effects model of aggregate R&D intensity. The short-run elasticity of the ratio of R&D to output with respect to the ''push'' effect of direct tax subsidies is significant, ranging from -0.15 to -0.22, while the long-run elasticity ranges from -0.46 to -0.77, depending upon the specification. The ''pull'' effect of the overall production tax system, as measured by the ETRMC, is significant as well, with the short-run elasticity ranging from -0.19 to -0.31 and the long-run elasticity from -0.58 to -0.93. This partly explains why R&D intensity in Canada is so low relative to other OECD countries despite the fact that we offer very generous ''push'' subsidies for R&D – our high production tax rates tend to discourage investment in R&D.
PIMS is presenting a series of lectures at the Calgary Place Tower 1 in downtown Calgary. These lectures, given by experts from the PIMS Universities, will focus on mathematical techniques and applications relevant to the oil and gas industry and will demonstrate the utility and beauty of applied mathematics. The talks are aimed at a general audience. Attendance may qualify for APEGGA Professional Development Hours.