Excerpted from the HBS working paper "Marketing, Product Differentiation, and Competition in the Market for Antiulcer Drugs"
The pharmaceutical industry commands a significant and growing position in the United States economy. Ethical pharmaceutical sales by research-based pharmaceutical companies in the United States are estimated to be $149.1 billion in 2000, up from $58.3 billion in 1990 [Pharmaceutical Research and Manufacturers of America 2000, p. 124]. In the vast market for pharmaceuticals, marketing constitutes a major competitive force by which firms strive to differentiate their products and soften price competition. Pharmaceutical firms spend as much on marketing as they do on research and development [Hurwitz and Caves 1988].
Three aspects of the pharmaceutical industry make it ideally suited for studying the role of marketing in product differentiation. First, pharmaceutical firms spend an unusually large percentage of their revenues promoting their products. Promotion-to-sales ratios for prescription drugs, including the antiulcer drugs studied here, typically range from 10 to 20 percent of sales, making them among the most heavily promoted of all manufactured goods [Leffler 1981]. This stands in sharp contrast to advertising in the median manufacturing industry, which devotes less than 1 percent of its sales revenues to advertising [Scherer and Ross 1990]. Second, pharmaceutical companies mostly concentrate their marketing efforts in a single form of marketing, visits to doctors' offices, which typically represent some 70 to 80 percent of industry marketing expenditures [Hurwitz and Caves 1988, Berndt, Bui, Lucking-Reiley and Urban 1997], and simplifies the analysis. Finally, drugs characteristics are essentially fixed, determined by the chemical properties of the particular molecule that constitutes the drug. Since changing the dosage, frequency, or other physical characteristics of a drug requires Food and Drug Administration (FDA) approval, this further limits the ability of pharmaceutical companies to compete against each other strategically by changing the characteristics of their products. The FDA also strictly regulates the advertising and promotion of prescription drugs. 1
Despite the importance of marketing, most recent studies of the pharmaceutical industry have focussed on branded versus generic competition and the effects of entry by generic drugs [e.g., Hurwitz and Caves 1988, Caves, Whinston and Hurwitz 1991, Grabowski and Vernon 1992, Frank and Salkever 1997, Hellerstein 1998, Scott Morton 1999]. Relatively few studies have examined the demand for branded pharmaceutical drugs [e.g., Alexander, Flynn and Linkins 1994, Stern 1996, Ellison, Cockburn, Griliches and Hausman 1997, Rizzo 1999], and even fewer have investigated the role of marketing, with the notable exceptions of Berndt, Bui, Lucking-Reiley, and Urban [1995, 1997], King , and Azoulay . Several studies have addressed specific aspects of the antiulcer drug market. Berndt, Pindyck and Azoulay  explored consumption externalities and diffusion; Azoulay  found that marketing had more of an impact on demand than clinical research; and Crawford and Shum  investigated uncertainty and learning in Italian pharmaceutical demand.
Many questions of strong empirical interest remain largely unanswered. How does marketing affect the degree of product differentiation? How important is marketing compared to other drug characteristics? Is marketing socially wasteful? How much does marketing raise the costs of entry to potential competitors? What is the market value of changing a drug characteristic, for example, eliminating one adverse drug interaction?
Economists have long been interested in how product differentiation affects market competition and how product differentiation changes over time within markets, but economists have been hampered by the lack of adequate, tractable models in tackling these empirically difficult issues. To address these and related questions, this paper develops a discrete choice model that incorporates the effects of marketing by the firm and its competitors and that allows marketing to affect product differentiation. The model enables the estimation of hard-to-measure effects of marketing using routine econometric methods while avoiding the sophisticated but computationally intensive numerical techniques of the related random coefficients models of Berry  and Berry, Levinsohn and Pakes . The model is then applied to analyze marketing in the specific context of the market for prescription antiulcer medications, one of the largest pharmaceutical markets. By estimating the model in this market, one can test theories about the role of marketing in expanding markets, differentiating products, deterring entry, and enhancing social welfare.
Based on monthly panel data for four antiulcer drugs from 1977 to 1993, this research yields several findings with important strategic implications. Observed drug characteristics, such as the number of approved uses, side effects, and dosage frequency, became less important over time in determining demand while marketing rose in importance. Marketing by both the firm and its competitors significantly affected the demand for its antiulcer drug. Total marketing by all firms reduced the degree of product differentiation in the antiulcer drug market and dramatically raised the cost of entry to potential competitors. Over the sample period, the role of marketing shifted from market expansion to business stealing (brand switching) as marketing became less effective in attracting new customers and more important as a means of capturing customers from rival firms. Finally, a rough calculation of the value of eliminating one adverse drug interaction from the pioneer drug, Tagamet, suggests that racing to market may entail significant private costs if further research and development could create a better drug.
· · · ·
1. For prescription drugs, the Federal Food, Drug, and Cosmetic Act requires advertisements to disclose certain information about the advertised product's uses and risks and to contain "information in brief summary relating to side effects, contraindications, and effectiveness" (21 U.S.C. 352(n)). Comparative advertising must be supported by clinical trials, and comparisons of side-effect profiles are prohibited [Kessler and Pines 1990].