• R&D/SALES = research and development expenditures divided by sales; obtained from Standard and Poor's Compustat Annual Data Tape. • CFM (cash flow margin) = profits after taxes plus depreciation divided by sales, all lagged two years; obtained from the Compustat Tape. • NR = sum of NPS(1) for ! 2; where NPS(1) = = 0, −1, −2, −3, −4, divided by R&D expenditures in sales of new chemical entities introduced in year in + 1. + 2, and + 3. Units are normalized for descriptive purposes (see figure 2). List of new chemical entities each year obtained from Paul de Haen, Nonproprietary Name Index, and special reports by de Haen. All data on sales of new chemical entities obtained from Intercontinental Medical Statistics. • DVR • PC = index of diversification, which equals 1 - Es; where s = share of prescriptions in ith class. This index was constructed using data from a marketing research firm. Lea Associates. In a special report for the year 1968, an analysis of pharmaceutical manufacturers was prepared in which each firm's total prescriptions were distributed among twenty-two classes (for example, infective and parasitic diseases, neoplasms, allergic disorders). Hence, DVR values for 1968 were assumed to hold for the entire period. = percentage of total firm sales accounted for by pharmaceutical sales. The value of PC for 1970 was obtained from NEDO Chemicals, E. D.C. Focus on Pharmaceuticals, September 1972. Values of PC for 1975 were obtained from Scrip, January 8, 1977. Hence, the 1970 PC values were applied to all prior years, and succeeding years were found by linearly interpolating between 1970 and 1975 values. greater than private rates of return on R&D activity for a number of well-known reasons. 32 Furthermore, Mansfield et al. recently investigated this question empirically for a sample of seventeen (nonpharmaceutical) innovations and found the social rate of return on average 32 A classic article in this regard is Kenneth Arrow's paper, in the National Bureau of Economic Research conference volume, "Economic Welfare and the Allocation of Resources to Invention," in R. Nelson, ed., The Rate and Direction of Inventive Activity (Princeton: Princeton University Press, 1962). R&D PROCESS: ECONOMIC FACTORS to be roughly double the private rate." The case study analysis by Weisbrod and Geweke to be presented later in this volume also indicates relatively high social returns to R&D. Although considerable research remains to be done in this area, these initial results suggest that policy makers should at least examine the basic factors underlying the declining levels of drug firm research intensities and innovations and should also consider possible policy options for dealing with this situation. Data Appendix The data used in this study were obtained from various sources, as will be described. Table 2 contains values for the variables for the year 1970 to provide the reader with an understanding of the relative magnitudes. The ten firms selected were all firms for which complete data were available for the 1962-1975 period. Generally, the-unavailability of data on R&D expenditures was the primary reason that most other firms failed to be included. 33 Edwin Mansfield, J. Rapoport, A. Romeo, S. Wagner, and G. Beardsley, “Social and Private Rates of Return from Industrial Innovation," Quarterly Journal of Economics, vol. 91 (May 1977), pp. 221-40. John F. Geweke and Burton A. Weisbrod, “Some Economic Consequences of Technological Advance in Medical Care: The Case of a New Drug," herein. Reprinted from Robert Helms, ed., Drugs and Health, American Enterprise Appendix A4 A Sensitivity Analysis of Expected Profitability of Pharmaceutical R & D Preliminary Draft Henry Grabowski and John Vernon April 1981 This research has been supported by National Science Foundation Grant PRA-7917524 made to Duke University. The pharamaceutical industry has been one of the most innovative industries in the U.S. over the past thirty years. However, the rate of new drug introductions in the past decade has been significantly lower than it was in the earlier post World War II period. As a result, the reasons for and social significance of this decline have been the subject of · considerable attention by both policymakers and academicians. The decline in new drug introductions has been accompanied by strong upward trends in costs, time, and risks associated with discovering and developing new drugs. As one would expect, studies of the rate of return to drug innovation have found relatively low returns (Schwartzman (1975) and Weston and Virts (1981)). It is also the case that U.S. firms are increasing their R & D expenditures in foreign countries at a faster rate than in the U.S. In fact, in real terms, U.S. R & D expenditures may be declining. One important explanation for these trends has been the increased regulatory controls of the Food and Drug Administration (FDA) which resulted from the 1962 Kefauver-Harris amendments to the Food, Drug and Cosmetic Act (Grabowski, Vernon, and Thomas (1978)). These amendments required a new drug's efficacy, as well as safety, to be demonstrated on the basis of well controlled scientific tests prior to marketing approval by the FDA. An indirect effect of regulation has been a reduction in the effective patent life for a new drug. The reason is that the average time to develop a new chemical entity (NCE) and gain regulatory approval far exceeds the time necessary to obtain a patent. While the length of patent protection has been of secondary import historically in the drug industry, this situation appears to be changing with the repeal of antisubstitution laws (Grabowski and Vernon (1979)). That is, the antisubstitution laws made it possible for innovating firms, through strong brand loyalties, to maintain dominant market positions for their products even after patent expiration. Now, in many states, lower cost generic products that become available upon patent expiration can be substituted by pharmacists even though the physician prescribes the original brand name products. The period of patent protection now averages only ten years or so as compared to the legal life of seventeen years. For this reason, legislative proposals have been made to restore part or all of the patent life lost during the chemical testing and FDA review period. The objective, of course, is to stimulate innovation by increasing the expected return to pharma ceutical R & D. Given the current interest in patent policy and its impact on the expected return to pharmaceutical R & D, we have performed a preliminary sensitivity analysis which sheds some light on the relationship between product life and profitability. Of course, the results are inadequate to support any particular product life as being the "socially optimal" patent life. Rather, the work here is intended as a first step in understanding the quantitative effects of various product lives on profitability as well as other related issues. Based upon a number of important assumptions, we show, for example, that at a 10 percent real interest rate the average 1970-1976 new drug required 19 years to break even. At an 8 percent interest rate, 12 years would permit the firm to break even. "Breaking even" means to cover all R & D discovery and development costs in addition to production and marketing costs. While the above paragraph refers to the average investment in drug innovation, we also show that the variance in payoffs is great and highly skewed. For example, of the 37 NCE's discovered and introduced in the U.S. in the 1970-1976 period, only 13 were able to at least cover their costs (over a 20 year life). This is true despite the fact that the average payoff to the 37 NCE's was slightly in excess of the average cost. An interesting finding for R & D strategic decisions is the variation of profitability across therapeutic classes. Although the small numbers of NCE's in certain classes makes it dangerous to generalize, it appears that for the 1970-1976 period the anti-infective category was clearly the most profitable. The cardiovascular and anti-inflammatory drugs were apparently next in order of profitability, while the remaining classes failed, on average, to break even. Another interesting result is the impact of reducing FDA approval time on profitability. Suppose there is no change in the amount of clinical 81-860 0-81--15 |