Despite the dramatic impact of R&D on firm and economic growth, there currently are no good measures of R&D effectiveness at the firm level. The most prevalent measure, patent counts, suffers from problems of universality (less than 50% of firms conducting R&D have any patents), uniformity (10% of U.S patents account for 81-85% of the economic value of all US patents) and reliability (patents are a poor predictor of firm market value).
Without a good measure of effectiveness, it is difficult to characterize the impact of R&D investment, to understand the mechanisms through which R&D investments generate economic outcomes, or to generate tools guiding future investment. What firms, policy makers and academics need is a universal, uniform and reliable measure of R&D effectiveness. That measure should match constructs in economic theory (so academics can test them) and should provide guidance to firms and policy-makers about the choice of investment levels.
Intellectual Merit: Recent empirical methods have facilitated a new measure of R&D effectiveness using firm accounting data that is based in economic theory. This measure has the potential of predicting both firms' behavior and their market value. The measure is called IQ because it captures firms' technical problem solving capability in much the same way individual IQ captures analytical problem solving capability: those with higher IQ solve more problems per unit of input (dollars for firms, minutes for individuals) than those with lower IQ. Perhaps most importantly, firm IQ (unlike individual IQ) appears to be mutable over long periods of time. Accordingly, understanding the organizational structures and processes driving differences in IQ offers the potential for firms to improve their R&D effectiveness.
Work to date has estimated the IQ for publicly-traded US firms, characterized variation of IQ across industries, and identified high and low IQ firms within each industry.
This pilot study represents the first stage in an effort to characterize the organizational configurations and processes driving IQ. The pilot comprises in-depth interviews with paired firms (one high IQ, one low IQ) in two industries. These interviews form the basis of rich case studies of the four firms, as well as identify candidate factors/organizational configurations to examine in a future quantitative study across the full spectrum of firms engaged in R&D.
Broader Impact: There are both immediate and long term benefits from this study. The immediate benefits of the IQ measure are 1) academics' use of the measure to resolve empirical anomalies in prior studies, 2) firms' use of the measure to compute their optimal R&D investment, and 3) policymakers' use of the measure to allocate funds based on firmer evidence. The longer term benefit from the full study (to be disseminated via conference presentations, journal articles and ultimately a book for practitioners) is advancing understanding about the effectiveness of R&D investments.
R&D was once an engine of economic growth. Since 1980, that no longer seems to be the case: GDP growth has been decreasing despite increasing R&D intensity. One reason for the decline may be that firms are "flying blind". Because there have been no good measures of R&D effectiveness they must use intuition to determine how or how much to invest in R&D. Fortunately recent econometric techniques facilitated a new measure of R&D effectiveness that universal, uniform and reliable: firm IQ. While prior work generated the IQs for all publicly traded firms engaged in R&D, no work explains the differences between high and low IQ. This study was the first stage in a two-stage effort to understand the antecedents of IQ. In this first (qualitative) stage we collected interview data to reveal potential antecedents to IQ. In the second (quantitative) stage we hope to test the statistical significance of those factors across a broad spectrum of firms. The most significant output of this study is the set of candidate antecedents of IQ. These were gathered principally from the ethnographic interviews with two subject firms, but were augmented by case studies using public information on three high IQ firms, and by a review of existing literature on R&D effectiveness. Because the goal of generating factors was to then evaluate them in a second stage, we also conducted a "proof of principle" for stage II using CIS and R&D survey data on Belgian firms matched to their IQs (which we also generated. That effort successfully demonstrated the feasibility of stage 2. More interestingly, it revealed some interesting correlations between IQ and four broad categories of factors: the stimuli for innovation, the knowledge type, form of innovation and cooperative activity.
