Review of Industrial Organization 17: 357–370, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
The Determinants of R&D Expenditures: A Study
of the Canadian Biotechnology Industry
DOUGLAS J. CUMMING
Faculty of Business, University of Alberta, Edmonton, Alberta, T6G 2R6, Canada
JEFFREY G. MACINTOSH
Faculty of Law, University of Toronto, 78 Queen’s Park, Toronto, M5S 2C5, Canada
Abstract. The relative importance of a multitude of factors for the allocation of expenses towards
R&D are assessed in an empirical study of the Canadian biotechnology industry. The results show
that patent protection and strategic alliances facilitate R&D spending. The results also show that
early-stage ﬁrms spend a greater proportion of the expenditures on R&D, while ﬁrms engaged in
R&D in platform technologies and ﬁrms with high debt-equity ratios spend a lower proportion of
their expenditures on R&D. Demand pull and competition variables are insigniﬁcant factors. Finally,
counter to our expectations, R&D expenditures are more intensive among ﬁrms engaged in R&D in
areas in which consumer controversies are more pronounced.
Key words: Innovation, research and development, technology.
The literature on research and development and the adoption of new technologies is
vast (see, for example, Tirole, 1988, Chap. 10; Scherer, 1989; Acs and Audretsch,
1990; Carlton and Perloff, 1994, Chap. 17; Martin, 1993, Chap. 13). However, there
is a comparative dearth of empirical studies on R&D. This study aims to close this
gap by ascertaining the signiﬁcance of a number of possible factors that play a
role in determining the intensity of research and development in the biotechnology
industry. Biotechnology has tremendous promise to make vast improvements in
agriculture, food products, health care, and the environment, among other things
(see MacIntosh and Cumming, 1999). The success of ﬁrms in the biotechnol-
ogy industry depends on their continued development of new technologies and
applications of existing technologies to new products (Goudey and Nath, 1997).
Previous empirical studies on research and development have focused on par-
ticular testable hypotheses. Scherer (1982) and Griliches (1989) focused on the
importance of patent protection. Farber (1981) considered the effect of the number
of scientists working in an industry. Acs and Audretsch (1990) considered the
The authors owe thanks to Ralph Winter, Frank Mathewson, Corrine Sellars and two anonymous
referees for helpful comments and suggestions.