R&D Investment, Productivity and Rates of Return: A Meta-Analysis of the Evidence on OECD Firms and Industries

Abstract

The volume of work on productivity effects of research and development (R&D) investment has expanded significantly following the contributions of Zvi Griliches and others to microeconometric work in late 1970s and early 1980s. This study aims to meta-analyse the research findings based on OECD firm and industry data, with a view to establish where the balance of the evidence lies and what factors may explain the variation in reported evidence. Drawing on 1,262 estimates from 64 primary studies, we report that the average effect of R&D capital on productivity and the average rate of return on R&D investment are both positive, but smaller than the summary measures reported in previous narrative reviews and meta-analysis studies. We also report that a range of moderating factors have significant effects on the variation among productivity and rates-of-return estimates reported in primary studies. Moderating factors with significant effects include: (i) measurement of inputs and output; (ii) model specifications; (iii) estimation methods; (iv) levels of analysis; (v) countries covered; and (vi) publication type among others. JEL Classification: D24, O30, O32, C49, C80

Innovation and productivity in family firms: evidence from a sample of European firms

“…A comprehensive overview of earlier work can be found in Mairesse and Sassenou (1991), while Hall, Mairesse and Mohnen (2010) cover more recent studies. Finally, three metaanalysis studies by Wieser (2005), Moen and Thorsen (2013) and Ugur et al (2014) provide meta-regression evidence on productivity and rates-of-return estimates. 2 The estimated overall rate of return to R&D is positive, generally statistically significant and with a magnitude according to the econometric methodology, data level, the measurement of output and of the inputs, the production function used, whether or not it controls for spillovers effects and the method of measuring spillovers (among others, see Hall and Mairesse, 1995;Harhoff, 1998;Aiello and Pupo, 2004;Maté-García and Rodríguez-Fernández, 2008;Wakelin, 2001;Rogers, 2010).…”

“…A more recent study by Hall, Mairesse and Mohnen (2010) synthesise in a new Elsevier Handbook a very large literature from the past 50 years on the models used in estimating returns to R&D and on the empirical results that have been achieved using these models. There are also three meta-analysis studies: Wieser (2005) surveys rates-ofreturn estimates at the firm level from 17 primary studies published between 1980-2000, focusing on whether the returns are stable over time and across different countries, industries and econometric specifications; Moen and Thorsen (2013) synthesizes the rate-of-return estimates reported by 41 studies published between 1962-2010 focusing on publication bias, and Ugur et al (2014) considers 64 econometric studies based on OECD firm or industry data published between 1980 and 2013. 3 Hall and Mairesse (1995) apply a Cobb-Douglas production function where productivity was regressed on R&D expenditures, capital and labour of French manufacturing firms (1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987).…”

“…On the other hand, if returns to R&D are found to be high, there are supply of funds problems (the high cost of capital), highlighting regulation on how financial markets work (Hall and Mairesse, 2009). Therefore, a justification for programs intended to stimulate innovative activities rests on the literature that estimates the returns to R&D. Along this line of reasoning, it is not surprising that there is a substantial number of empirical studies assessing the returns to R&D at country, regional, industry and firm-level (for a review of the literature see Hall, Mairesse and Mohnen, 2010; while for meta-analysis studies see Wieser 2005, Moen and Thorsen 2013, and Ugur et al, 2014. Limiting attention to studies at firm level, the evidence reveals a positive relationship between innovation and productivity (among others, see Hall and Mairesse, 1995;Harhoff, 1998;Aiello and Pupo, 2004;Maté-García and Rodríguez-Fernández, 2008;Wakelin, 2001;Rogers, 2010).…”

Meta-analysis, military expenditures and growth

“…Recent discussions of meta-analysis in economics are Ioannidis, Stanley & Doucouliagos (2017), Filges & Hansen (2017), Ugur et al (2014), Ugur (2014), Doucouliagos & Stanley (2009) and Rose & Stanley (2005). 1 One objective of meta-analysis is to combine a large number of estimates to get the overall estimate of the ‘effect size’, in our case a measure of the effect of military expenditure on growth.…”