|
|
| Home | Current Issue | Archives | Topic Collections | Search | Blog | Subscribe | Contact Us | Help |
Health Tracking |
TRENDS
The Link Between Gross Profitability And Pharmaceutical R&D Spending
Since the late 1950s, when the Kefauver Committee investigated the business practices of U.S. pharmaceutical companies, representatives of that industry have argued that its profits are an important stimulus to, and source of funding for, research and development (R&D)—which in turn leads to a stream of health-enhancing new products. Although the argument is plausible on its face, quantitative evidence on the robustness of the linkage has been scarce. This paper reports the results of a simple data analysis yielding surprising new insights.
Profitability and investments in R&D can, in principle, be linked in three rather different ways. First, successful R&D leads, with long and variable lags, to new products, which, depending upon their reception in the market, can add greatly to company profits. The distribution of profit outcomes, as research by Henry Grabowski and John Vernon has shown, is highly skewed.1 A minority of new products confer blockbuster profits, while the majority return less than the capitalized cost of R&D, including the cost of failed projects.
Second, the profits earned by a company serve as a source of funds to support R&D investments, and some managers are known to set R&D budgets using rules of thumb emphasizing an indicator of current cash flow or sales. To be sure, as recent experience in biotechnology shows, funds for R&D can also be raised through new capital issues. Prior tests of the hypothesis of internally generated funds have yielded mixed results. For most well-established corporations, R&D spending is not greatly dependent upon internal cash flow, but small high-tech enterprises—before the 1990s venture capital boom—and the research-intensive pharmaceutical industry were probable exceptions.2 Third, managers’ expectations of future profit opportunities, which are tempered, inter alia, by contemporary market conditions, can exert a demand-pull influence on R&D investments.3
Testing how well these relationships hold for investments in pharmaceutical R&D is rendered difficult by the complex structure of the leading pharmaceutical companies. They operate within a wide variety of fields in addition to ethical drugs—for example, pharmacy benefit management, herbicides and pesticides, medical instruments and supplies, prosthetics, hair care products, dental products, and nutritional products. The more diversified companies almost never publish R&D outlay breakdowns subdivided among these fields, and they seldom report their operating margin results in enough detail to relate R&D indices with any precision to measures of profitability.
An alternative approach, and the one used here, is to analyze data on R&D investment and profits at the aggregated industry level. The principal industry trade association, Pharmaceutical Research and Manufacturers of America (PhRMA), conducts an annual survey, from which one can obtain a continuous data series on PhRMA members’ ethical drug R&D outlays extending back to the early 1960s (Exhibit 1
).
|
| Time-series analysis. |
|---|
Examination of Exhibit 1
reveals why the usual techniques of time-series analysis used by economists work poorly in ascertaining the links between R&D and profitability. Those methods focus on year-to-year changes in the variable of interest. But there are few sharp changes in the spending path from year to year and, accordingly, few true "degrees of freedom" necessary for a standard time series analysis.4 Rather, one sees gradual swings in actual R&D spending around a best-fitting long-term R&D time trend, assuming steady exponential growth. The trend line implies growth in inflation-adjusted spending at an average rate of 7.5 percent per year. The most closely comparable aggregate time-series measure of industry profitability is derived from Census of Manufactures and Annual Survey of Manufactures data from the U.S. Census Bureau. It is computed as sales less outside materials purchases, payroll outlays, and employee fringe benefits—including those mandated by law along with voluntary benefits. As such, it is best described as a measure of pharmaceutical manufacturing plants’ gross margins—that is, the surplus of revenues over in-plant production costs available to cover R&D costs along with depreciation, marketing costs, central office costs, debt service costs, income taxes, and net profits. The coverage match between ethical drug R&D outlays and this gross margin measure is not perfect, as the census universe under the Standard Industrial Classification (SIC) code 2834—Pharmaceutical Preparations— also includes less research-intensive over-the counter drugs, generic drugs, and some vitamin formulations.5 Fringe benefit outlays, amounting to 3.68 percent of gross margins in 1967, had to be estimated by extrapolation for 1962–1966, imparting possible inaccuracies in the gross margin measure too small to affect the results reported here.
| Gross margins versus R&D outlays. |
|---|
 TopTime-series analysis. Gross margins versus R&D... A virtuous rent-seeking model.NOTES |
|---|
The growth rate of deflated gross margins was 4.23 percent per year—much lower than the 7.51 percent growth rate found for R&D outlays (Exhibit 2
). The disparity of growth rates implies a likely slackening of R&D growth rates in the future. If R&D were covered solely by domestic gross margins, continuation of growth trends experienced since 1962 would mean that R&D outlays would exceed gross margins in the year 2025. To be sure, profits from overseas sales also help to repay R&D costs, but since the United States is the largest single market for U.S. drug companies’ products, retardation of R&D growth rates seems likely in the long run.
|
As in the R&D time series, pharmaceutical industry gross margins exhibit long swings around their exponential time trend. To some extent, coincidence in the timing of the swings can be seen by comparing Exhibits 1
and 2. However, the relationships are brought into sharper focus by computing the percentage deviations of actual R&D outlays and gross margins from their exponential time trend values.6 The resulting trend deviation series are juxtaposed in Exhibit 3.
|
The degree of coincidence was, at least to this investigator, surprisingly close. The simple Pearsonian correlation between the two time series is +0.92. Deviations from trend values rise and fall in tandem. The swings are so closely correlated that it would be implausible to infer a chain of causation running from R&D to profits, since lags of ten to fifteen years from peak R&D spending to peak profitability for new products are typical.7 At two of the three clear turning points, reversals in the R&D spending series precede reversals in the gross margin series by a year or two. This is superficially inconsistent with a hypothesis that changes in gross margins drive changes in R&D spending. However, the paradox diminishes if decision makers are able to foresee changes in general industry conditions two or more years into the future—for example, recognizing that the "rational drug design" approaches, demonstrated by the introduction of Tagamet in 1977, presaged increasingly rich opportunities for profitable new product development.
Sensitivity tests revealed that the patterns observed in Exhibit 3 persist when domestic R&D outlays, a time series available only beginning in 1970, are substituted for worldwide R&D outlays, and when fringe benefit outlays, reported by the Census Bureau only beginning in 1967, are not deducted in calculating gross margins.
It is conceivable, as one referee suggested, that the cycles observed here reflect spuriously correlated changes in industry aggregates, for example, as a result of differences in sample coverage between the trade association and Census Bureau universes. To test this possibility, a further analysis correlated trend deviations in variables defined as ratios, with no intermingling of trade association and census universe data for a given ratio. For R&D, the relevant ratio was worldwide R&D outlays, divided by worldwide sales of trade association members in any given year. The PhRMA sales variable was not used in the previous analysis. For gross margins, the relevant ratio was the gross margin, as defined for Exhibit 2, divided by the total value of plant shipments, a rough Census Bureau surrogate for sales, which can double-count interplant shipments. For the time series of each variable, a best-fitting linear trend was estimated. A linear trend was used because, as variables with a distinct upper bound, the two ratios could not plausibly sustain exponential growth for any extended period. Percentage deviations from these best-fitting linear trends were computed (Exhibit 4). Except during the early 1960s, their movements over time are similar to and consistent with those of Exhibit 3. Their simple correlation is 0.863. Given the internally consistent industry sample frames but different trend measurement assumptions used for Exhibit 4 compared with Exhibit 3, the similarity of trend deviation patterns suggests that there was indeed cyclical comovement in pharmaceutical industry gross margins and R&D outlays.
|
| A virtuous rent-seeking model. |
|---|
| TopTime-series analysis.Gross margins versus R&D...A virtuous rent-seeking model.NOTES |
|---|
Thus, a robust pattern persists. Combined with evidence that profit rates of return on pharmaceutical industry R&D investments tend to exceed risk-adjusted capital costs by only modest amounts, the pattern suggests that pharmaceutical industry R&D is best described by a virtuous rent-seeking model.8 That is, as profit opportunities expand, firms compete to exploit them by increasing R&D investments, and perhaps also promotional costs, until the increases in costs dissipate most, if not all, supranormal profit returns. If this is a correct interpretation of the industry’s behavior, it has self-evident implications for policy interventions aimed at reducing industry prices and profits.
| Editor's Notes |
|---|
F.M. Scherer is Aetna Professor of Public Policy Emeritus at Harvard University’s Kennedy School of Government and a lecturer in public affairs at Princeton University.
| NOTES |
|---|
| TopTime-series analysis.Gross margins versus R&D...A virtuous rent-seeking model.NOTES |
|---|
- H.G. Grabowski and J.M. Vernon, "A New Look at the Returns and Risks to Pharmaceutical R&D," Management Science (July 1990): 804–821.
- See C.P. Himmelberg and B.C. Petersen, "R&D and Internal Finance: A Panel Study of Small Firms in High-Tech Industries," Review of Economics and Statistics (February 1994): 38–51
; W.W. McCutchen Jr., "Estimating the Impact of the R&D Tax Credit on Strategic Groups in the Pharmaceutical Industry," Research Policy (August 1993): 337–351 ; and H.G. Grabowski, "The Determinants of Industrial Research and Development," Journal of Political Economy (March/April 1968): 292–306. - See J. Schmookler, Innovation and Economic Growth (Cambridge: Harvard University Press, 1966)
; and F.M. Scherer, "Demand-Pull and Technological Innovation: Schmookler Revisited," Journal of Industrial Economics (March 1982): 225–238. - Despite that difficulty, the use of standard year-to-year time series analysis yields results consistent with those reported here. See F.M. Scherer, Industry Structure, Strategy, and Public Policy (New York: HarperCollins, 1996), 388.
- Extending the analysis into 1997 and especially beyond is problematic, since a new industry classification, the North American Standard Industrial Classification, was adopted by the U.S. Census Bureau in that year. Using splice data provided by the Census Bureau for 1997, one finds that constant-dollar gross margins for the industry, as defined in 1996, probably rose by approximately 11.9 percent relative to 1996. Constant-dollar R&D outlays rose by 10.6 percent. The census universe excludes production in Puerto Rico, which is an important source of U.S. pharmaceutical products.
- For example, for R&D, the calculation is [(R&D–Trend)/Trend] x 100.
- See U.S. Congress, Office of Technology Assessment, Pharmaceutical R&D: Costs, Risks, and Rewards, Pub. no. OTA-H-522 (Washington: U.S. Government Printing Office, February 1993), 20–22.
- Ibid., chap. 1.
For the pioneering theoretical analysis of such behavior, see Y. Barzel, "Optimal Timing of Innovations," Review of Economics and Statistics (August 1968): 348–355. For an application to pharmaceuticals, which at the time of writing was considered one of two plausible theoretical alternatives, see Scherer, Industry Structure, Strategy, and Public Policy, 365–366. - See C.P. Himmelberg and B.C. Petersen, "R&D and Internal Finance: A Panel Study of Small Firms in High-Tech Industries," Review of Economics and Statistics (February 1994): 38–51
This article has been cited by other articles:
 |
|
 |
|
  J. A. DiMasi, R. W. Hansen, and H. G. Grabowski Misleading Congress about Drug Development: Reply Journal of Health Politics Policy and Law, January 1, 2008; 33(2): 319 - 324. [Abstract] [PDF]
|
|
|
|
|
|
D. W. Light Misleading Congress about Drug Development Journal of Health Politics Policy and Law, October 1, 2007; 32(5): 895 - 913. [PDF]
|
|
|
|
 |
|
 S. D. Reed, R. M. Califf, and K. A. Schulman How Changes In Drug-Safety Regulations Affect The Way Drug And Biotech Companies Invest In Innovation Health Aff., September 1, 2006; 25(5): 1309 - 1317. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. K. Choudhry and A. S. Detsky A Perspective on US Drug Reimportation JAMA, January 19, 2005; 293(3): 358 - 362. [Full Text] [PDF]
|
|
|
|
 |
|
 F.M. Scherer The Pharmaceutical Industry -- Prices and Progress N. Engl. J. Med., August 26, 2004; 351(9): 927 - 932. [Full Text] [PDF]
|
|
| Home | | | Current Issue | | | Archives | | | Topic Collections | | | Search | | | Blog | | | Subscribe | | | Contact Us | | | Help |
|
© 2001-2001 Project
HOPE–The People-to-People Organization |
||||||||||||||||