Why Some R&D Organizations Are More Productive Than Others

Abstract

For several years, my colleagues and I have been trying to understand how technology-intensive firms that are highly productive differ from less productive firms in the way they mane their RD namely, the costs of labor, plant, property and equipment, and materials. Economists have developed and applied such a measure to the study of technological change and productivity growth. The methodology is known as the Total Factor Productivity (TFP) approach. The TFP index is the ratio of outputs to inputs. Output is measured by real value of net sales; while input is measured by the combination of (real) production factors used to produce the output, weighted by their respective output shares (L, C and M); i.e., (Equation omitted) TFP growth over time (DeltaTFP) is measured in terms of differences in growth rates of outputs and inputs. Empirically, it has been found that in the past several decades productivity has grown faster in the United States than would be predicted from direct measurement of changes in these three factors of production (after appropriate adjustments for price changes). This in growth of measured outputs relative to measured inputs has been examined carefully and often by economists. They conclude that the excess is caused by technological change, or to be more precise, changes in the stock of technical knowledge. When R&D expenditures are used as a surrogate for the rate of technological change and added to the list of factors of production, it has been found that R&D accounts for impressive national and industry-wide gains in productivity growth after all other factors have been taken into account. Statistically valid estimates ranging from 30 percent to more than 100 percent returns to R&D expenditures are the norm. Thus, the contribution of changes in the stock of technical knowledge and gains in productivity can be understood by empirically estimating the relationship between R&D expenditures and TFP growth, holding other factors constant. This means the productivity impact of R&D can be assessed by the expression: TFP growth = a + b* (R&D intensity), where a is a constant parameter and b is the impact of R&D input on TFP measured as a marginal rate of change. R&D intensity is measured by expressing R&D expenditures as a percentage of sales. Our research goal, then, was to carefully measure all of these variables so that we could isolate the R&D variable and see if we could find a relationship between not only the level of R&D investment but between the composition of that investment and productivity growth. We began the project by selecting 15 drug firms and looking at the relationship between R&D intensity and productivity growth in these firms between 1971 and 1990. Figure 1 shows the average annual total factor productivity growth of these firms over 20 years computed from publicly available data.(Figure 1 omitted) Eli Lilly, for example, improved its productivity an average of 2.9 percent per year over that period and invested 4.5 percent of sales in R&D. Looking at the R&D-to-sales ratio in Figure 1, (i.e., R&D intensity) there is a consistent relationship with annual total factor productivity growth for this sample of companies over this period. That began to make us feel we might be looking at something worthwhile. We then plotted each of those 15 companies against some financial ratios that CEOs or security analysts might be interested in. Return-on-assets measured against total factor productivity growth looks pretty interesting. …