Publication Bias in the Returns to R&D Literature
- PDF / 598,142 Bytes
- 27 Pages / 439.642 x 666.49 pts Page_size
- 95 Downloads / 207 Views
Publication Bias in the Returns to R&D Literature Jarle Møen1 · Helge Sandvig Thorsen2
Received: 26 November 2014 / Accepted: 17 September 2015 © Springer Science+Business Media New York 2015
Abstract The returns to R&D literature is large and has been surveyed on several occasions. We complement previous surveys by discussing the scope for publication bias and illustrate how formal meta analytic techniques can be used to analyse the problem. We find evidence consistent with positive publication bias and discuss possible interpretations. The bias appears to be particularly strong in the part of the literature that controls for unobserved firm fixed effects. The reason may be that fixed effects specifications are particularly susceptible to measurement errors and therefore have a high probability of producing implausibly low return estimates. Implausible estimates are likely to be filtered out before being reported, and our analysis suggests that 23 % of a hypothetical complete literature is missing. Future reviews should take into account that the full effect of negative specifications biases may be masked by reporting and publication bias. Keywords Returns to R&D · Meta-analysis · Publication bias · Funnel asymmetry · Trim-and-fill method · FAT – PET – PEESE
The project is financed by the Research Council of Norway. We are grateful to Jonas Andersson, referees and participants at the MAER-Net 2013 colloquium at the University of Greenwich for useful comments. Jarle Møen
[email protected] Helge Sandvig Thorsen [email protected] 1
Department of Business and Management Science, Norwegian School of Economics, Helleveien 30, N-5045 Bergen, Norway
2
Department of Economics, Norwegian School of Economics, Bergen, Norway
J Knowl Econ
Introduction OECD governments spend a substantial amount of public money on programs intended to stimulate innovative activities. The justification for these programs rests on a vast and steadily growing literature that estimates the private and social returns to R&D. Much of the returns to R&D literature build on the R&D capital model formalized in Griliches (1973, 1979), and the literature is reviewed in a number of excellent surveys. Mairesse and Mohnen (1990) is an early example. Hall et al. (2010), in a new Elsevier Handbook, is the most recent authoritative review.1 All surveys conclude that there are large returns to R&D, although no one computes a combined return estimate using a formal meta-analytic technique. Most economists have a prior belief that returns to R&D are positive and possibly large, but the returns to R&D literature are prone to problems related to measurement, selection, choice of functional form, and appropriate lag lengths. This suggests that there is a danger of publication bias. Reporting and publication bias are widely recognized as threats to the validity of empirical research. This bias may be selfimposed by researchers or imposed by editors and referees who consider negative, small, or non-significant coefficients to be suspicious and of little
Data Loading...
