PERSUASION OR INFORMATION?
`THE ECONOMICS OF PRESCRIPTION
`DRUG ADVERTISING"
`
`KEITH 3. LEFFLER
`University of Washington
`
`T H E economic and welfare effects of advertising have been extensively
`studied over the last half-century. However, these efforts have not led
`to a consensus either a s to the effects or the value of advertising.
`Economists continue to reach polar conclusions that appear to be derived
`mainly from preconceptions of the social desirability of advertising. On
`the one hand is research that emphasizes promotion's ability to create
`"artificial" product differentiation and thereby produce informational
`c o n f u s i o n . ~ h i s line of research stresses the empirical association of
`extensive advertising with high concentration and high accounting
`profits-evidence
`judged to support increased market power and entry
`barrier e f f e ~ t s . ~ A contrary body of research emphasizes the value of
`advertising in providing information and, hence, in promoting compe-
`t i t i ~ n . ~ The empirical findings most consistent with this view are that
`prices paid by consumers in (selected) markets are lower with advertising
`than without it.4
`
`* Yoram BarzeI provided heIpful comments
`This "Harvard View" of advertising was developed by Edward Chamberlin, The
`Theory of Monopolistic Competition (1933). Joe Bain, Baniers to New Competition: Their
`Character and Consequences in Manufacturing Industries (1956); and William S. Comanor &
`Thomas A. Wilson, Advertising and Market Power (I974), pursue some of the issues raised
`by Chamberlin. The Comanor and Wilson book provides the most-cited work supporting the
`product-differentiation, entry-barrier view of advertising.
`* James M. Ferguson, Advertising and Competition: Theory, Measurement, and Fact
`(19741, reviews much of the empirical Literature on advertising's effect. Papers in Part IV of
`Issues in Advertising: The Economics of Persuasion (David G. Tuerck ed.) (Am. Enterprise
`Inst. 1978), also discuss the relationship between advertising, concentration, and profits.
`This "Chicago" view is represented by Lester G. TeLser, Advertising and Competition,
`72 J. Pol. Econ. 537 (19t3); Philip Nelson, Advertising as hformation, 82 J. Pol. Econ. 729
`(1974); and Yale Brozen, Entry Baniers: Advertising and Product Differentiation, in Indus-
`trial Concentration: The New Learning 115 (Harvey J. G o l d s c h i d , H. Michael Mann, & J.
`Fred Weston eds. 1974).
`' See, for example, Lee Benham, The Effect of Advertising on the Price of Eyeglasses, 15
`J. Law & Econ. 337 (1972); and Robert L. Steiner, Does Advertising Lower Consumer
`Pnces? 37 J. Marketing 19 (1973).
`
`ALCON 2028
`Apotex Corp. v. Alcon Research, Ltd.
`Case IPR2013-00428
`
`

`
`46
`
`THE JOURNAL OF L A W A N D ECONOMICS
`
`Economic analysis generally treats advertising a s a homogeneous ac-
`tivity that is evaluated independently of why it might increase demand.5
`Yet advertising's effects need not be the same in different markets or in
`different settings within a market. For example, price comparison ads of
`standardized products (for example, ground beef) may lower both entry
`costs and average price paid, while "image advertising" of heterogeneous
`products (for example, perfumes) may increase prices and the cost of new
`entrants gaining consumer trials. Both positive and normative analysis
`should therefore be prefaced by the particulars of the products advertised,
`the message delivered, and the buyers addressed.Vf advertising is a
`multifaceted, heterogeneous activity, general statements as to the effects
`and efficiency of advertising may not be possible, and empirical studies
`using individual industries as cross-sectional observations may be eco-
`nomically ~ninterpretable.~ However, the study of advertising within a
`single industry can provide a piece in the montage required for economic
`understanding.
`This paper analyses one market characterized by very large promo-
`tional expenditures-the market for prescription drugs. This market is
`especially appropriate for detailed analysis since the polar positions on
`the desirability of advertising are weU represented in policy discussions of
`the prescription drug market. The continual introduction of new, poten-
`tially life-saving products makes the potential gains from the rapid dis-
`
`Comanor & Wilson, supra note 1, briefly discuss advertising that is designed to produce
`"bandwagon" effects and artificial product differentiation. Even though their nonnative
`analysis seems predicated on such advertising effects, they fail to operationally define such
`notions. Indeed, they model only advertising that provides correct information about a
`product's characteristics. K. Boyer, Informative and Goodwill Advenising, 56 Rev. Econ.
`& Stat. 541 (1974), does explicitly recognize that advertising is not homogeneous either in its
`purpose or its effects. However, his definition of goodwill (versus informative) advertising as
`that "which has the effect o f encouraging buyer inertia and loyalty," id. at 541, fails to
`distinguish advertising types.
`a I n a working paper, An Analysis of the Functions of Advertising (Mhrch 1980) (unpub-
`lished paper at Univ. of Washington), 1 define and distinguish five reasons for advertising
`affecting product sales. These are: (L) supply of (correct or incorrect) information on a
`product's characteristics (price, availab'ility, use, color, odor, and so on), (2) supply of
`information to nonusers of a product about the tastes, preferences, and self-image of the
`product's users, (3) reduction in the perception or recall costs required to identie products
`in product classes, (4) information on price-marginal cost differences and, hence, on the
`incentive to maintain quality, and (5) persuasion designed to substitute emotional decisions
`for rational, evaluative decisions. 1 argue that the competitive effects of advertising depend
`upon what function advertising plays.
`This problem of heterogeneous relationships within the data is confirmed in studies by
`Boyer, supra note 5; M. Porter, Consumer Behavior, Retailer Power, and Market Perfor-
`mance in Consumer Goods Industries, 56 Rev. Econ. & Stat. 419 (1974); and Frank Bass,
`Phillippe Cattin, & Dick Wittink, Market Structure and Industry Influence on Profitability,
`in Strategy + Structure = Performance 181 (Hans B. Thorelli ed. 1977).
`
`

`
`PRESCRIPTION DRUG ADVERTISING
`
`47
`
`semination of product information via advertising substantial. Nonethe-
`less, government investigations of the pharmaceutical industry stress that
`intensive advertising of drugs results in excessive use of high-priced,
`heavily promoted brand-name products even though equivalent low-
`priced products are available. Those viewing pharmaceutical advertising
`with disfavor insist that these ads are frequently uninformative and seem
`simply to harp the products' names in order to persuade doctors to select
`products out of habit rather than by evaluative c h o i ~ e . ~
`The advertising of medicines is closely monitored by government au-
`t h o r i t i e ~ . ~ To understand constraints on pharmaceutical advertising, Sec-
`tion I of this paper briefly considers the history and the regulation of
`pharmaceutical advertising in the United States. Section I1 empirically
`examines drug advertising that focuses on the informative versus the
`"habit formation" roles of product promotion. Hypotheses concerning
`the variance in advertising intensities across drug submarkets and among
`individual drug products are developed and tested for these two alterna-
`tive advertising theories. Section 111 explores the welfare effects of phar-
`maceutical advertising. The empirical analysis concentrates on the re-
`lationships between product innovation, product entry, product price,
`and the promotional strategies of both established and new products.
`The empirical results developed here indicate a dual role of phar-
`maceutical advertising: advertising appears to inform physicians about
`the existence and characteristics of new products while also producing
`"brand-name recall" effects that favor established products facing new
`competition. Pharmaceutical advertising thus serves to speed the entry of
`
`I n the Kefauver hearings Leading to the 1962 amendments to the FDA act a witness from
`Premo Pharmaceutical testified that "the only real competition we have in our field is the
`tremendous competition for the eye and ear of the physician-how many pages of advertis-
`ing we can put out, how many samples we can distribute, how many detailmen we can put in
`the field. These . . . alone govern the ultimate acceptance of the product." Cited in Richard
`Harris, The Real Voice 90 (1964). Senator Kefauver concluded that, "the promotional
`efforts . . . had essentially one purpose-to plant trade names firmly in the minds ofphysi-
`cians ." Id. Kefauver repeatedly noted the large price differentials between brand-name and
`generic drugs even when they were produced in the same plant by the same manufacturer.
`This was attributed to the continuing barrage of promotional material addressed to physi-
`cians. Advertising also was considered the prime cause of alleged excess profits.
`* This feature of the pharmaceutical market serves to limit the roles of advertising such
`that the analysis is relatively tractable compared to other markets. The combination of FDA
`testing and quality-control requirements and FDA and ITC monitoring of advertising con-
`tent suggests that advertising of prescription drugs will not provide fraudulent or incorrect
`persuasive messages and also will not be crucial in guaranteeing product quality. I n addition,
`the private nature of drug consumption and the reliance on third-party experts should l h i t
`the value of "image" advertising for prescription drugs. This contrasts, in my view, to much
`of the advertising of "social" drugs (tobacco and alcohol) through which consumers can
`indicate their tastes and preferences to their associates via the (public) consumption of
`advertised products.
`
`

`
`48
`
`THE JOURNAL OF LAW A N D ECONOM[CS
`
`superior new products while likely retarding the entry of later, low-priced
`close substitutes.
`
`Medicines are claimed to be the first products advertised in printed
`form.'* Regulation of medical advertising was not long in following. The
`colony of Virginia was first to pass such regulatory legislation in the
`United States. Setting a precedent followed until 1962, Virginia's 1736
`legislation required only that the "label" of medicines specify the ingre-
`dients." Claims about the effects of the drugs were not addressed. The
`first federal legislation relating to the promotional material accompanying
`medicinals was passed in 1848." This legislation applied only to imported
`drugs and again required only a correct listing of the ingredients of the
`drugs.
`Until the late 1800s, medicines were mixed by pharmacists from stan-
`dardized generic ingredients. Pharmacists served both as advisors and the
`assurers of quality. However, by about 1880, advances in the technology
`of large-scale mixing, forming, and bottling of tablets led to more cen-
`tralized production of medicines.13 Pharmacists were no longer able to
`directly monitor the ingredient mixes of the centrally produced drugs they
`dispensed. This provided the manufacturers of pharmaceuticals with the
`opportunity to compete by providing homogeneous, high-quality prod-
`ucts. However, such competition requires the identification and knowl-
`edge of individual manufacturers. Thus, manufacturer trademarks and
`brand-name promotion became important means of internalizing the gains
`from producing high-quality , unpatented drugs.14
`
`Frank Presbrey, The History and Development of Advertising 289 (1929), repons that
`"The 'first puff, which appeared in a German news book in 1591, announced the discovery
`oCa mysterious and wonderful curative herb. I n France and England the quacks, who have a
`much longer history than advertising, were the quickest to appreciate the printed word as an
`aid to selling."
`IL Parts of this legislation are reproduced in Edward Kremers, Georg Urdang, & Glenn A.
`Sonnedecker, Kremers and Urdang's History of Pharmacy 158 (4th ed. 1976).
`See Stephen Wilson, History of Pure Food and Drug Legislation I0 (Am. Council of
`Public Affairs 1942).
`L3 See Frank 0. Taylor, Forty-five Years of Manufacturing Pharmacy, 4 J. Am. Phar.
`Ass'n 468 (1915). The centrally manufactured products were ooopatented proprietary prod-
`ucts sold by generic name with formulae and dosage published in the Association of Pharma-
`cists Guide, the Pharrnacopeia.
`l4 See R. George Kedersha, Brand Name Prescription Products and Their Impact: A
`Historical Survey, Medical Marketing and Media, May 1978, at 32-38. By 1877, Parke, Davis
`and Company was publishing a "house journal" mailed largely to physicians. The magazine
`documented the therapeutic uses and quality of Parke, Davis products. Kedersha, supra at
`33. The centralid drug manufacturers also developed patented "specialties" and prior to
`
`

`
`PRESCRIPTION DRUG ADVERTISING
`
`49
`
`With increased drug promotion came increased agitation over exagger-
`ated product claims. In 1906 the first food and drug act was passed, and
`the Bureau of Chemistry of the Department of Agriculture was designated
`as the agency to monitor compliance with federal "labeling require-
`ments" for drug products." Among other endeavors, the bureau at-
`tempted to act against false therapeutic claims. However, in 1911 the
`Supreme Court ruled in United States v. Johnson that the labeling provi-
`sion can "by no possible construction be extended to an inquiry as to
`whether or not the prescription be efficacious or worthless to effect the
`remedy claimed for it."'"
`Three years later, the federal government attempted to regulate the
`therapeutic claims of drug producers via the 1914 Federal Trade Commis-
`sion Act. l7 However, in 1931 the Supreme Court again intervened, ruling
`that the FTC act was applicable only to those false claims which injure
`present o r potential competitor^.^"^ a result, the act was powerless to
`prevent injuries to consumers from falsely advertised products.
`The loopholes in both the FTC and food and drug regulations were not
`rectified until 1938. In that year, Congress passed the Wheeler-Lea
`amendment to the FTC act, which made false advertising for the purpose
`of inducing the sale of an article injurious to health a violation of the FTC
`act regardless of the competitive effects. Like the original act, advertise-
`
`1900 used salesmen to market these products and extoll their quality control to both physi-
`cians and pharmacists.
`Is On June 29, 1906, the federal government passed a Food, Drug, and Cosmetic Act
`designed to regulate interstate commerce in food, drink, cosmetics, and drugs. The bill as
`first introduced into Congress prohibited any false or misleading statements in or on drug
`packages. Under pressure from patent medicine interests, the bill as finally passed prohib-
`ited only false claims about the ingredients of a drug.
`lE Johnson was shipping medicine interstate with Iabels claiming the medicine would cure
`cancer and that the product was "guaranteed under the Pure Food and Drug Act." See H.
`Leo Fusilier & Jerome C. Darnen, Competition and Public Policy 377 (1971).
`The Shedey Amendment of I912 to the Food and Drug Law, ch. 352, 37 Stat. 416-17
`(1912), was passed shortly after this judicial decision. This amendment designated a drug as
`misbranded "if its package or label shaU bear or contain any statement, design or device
`regarding the curative or therapeutic effects of such article . . . which is faIse and fraudu-
`lent." Id. In addition to being restricted to labeling, the requirement for the government to
`prove fraudulent intent rendered the amendment impotent.
`L7 Federal Trade Commission Act of 1914, ch. 3 11, 38 Stat. 717.
`kB FTCv. Raladarn Co., 283 U.S. 643 (1931). Raladam Company was the business m e o f
`Edward Hayes. Hayes was in frequent trouble with both the Post Office and the l T C for his
`patent medicine merchandising techniques and had been convicted of mail fraud prior to the
`Raladarn Case. The present action resulted from his promotion of Maramola, an anti-
`obesity drug containing thyroid. The JTC issued a complaint against Raladam in 1928 ob-
`jecting to Hayes's claim that Maramola was "safe, effective, dependable and without danger
`of harmful results." Even though the ETC demonstrated that Maramola was highly danger-
`ous, the Court ruled that since Raladam's rivals were engaged in similar hyperbole, the R C
`was precluded from taking any action against Hayes.
`
`

`
`50
`
`THE JOURNAL OF LAW AND ECONOM[CS
`
`ments to physicians were e ~ e m p t e d . ' ~ The Wheeler-Lea amendment re-
`sulted from an extended congressional debate about food and drug regu-
`lation. At the same time the FTC amendment was passed, a new food,
`drug, and cosmetic act became law. The 1938 Food, Drug, and Cosmetic
`Act was explicitly intended not to give the FDA authority over drug
`promotion because such authority belonged to the FTC.24 However, the
`FDA drug-labeling requirements were extended to include directions for
`use and "adequate warnings" on aU drug labels.
`The apparent1 y innocuous labeling requirements were given a broad
`interpretation by the FDA. Temin reports that in the FDA's view
`adequate directions for laymen could not be written for many drugs. It
`followed that any directions written for laymen were misleading. Ef-
`fectively, this made many over-the-counter products automatically
`mislabeled. However, since directions and warnings written for physi-
`cians were not judged so harshly, the FDA's regulatory interpretation led
`firms to restrict most drugs to prescriptions use only .21
`The shift to prescriptive drugs enabled the FDA to usurp FTC power
`since the FTC authority over drug promotion specifically exempted
`promotion directed to physicians. While the FDA initially interpreted
`labeling to include all written, printed, or graphic matter accompanying a
`medicine at any time, by 1948 they had extended their interpretation to aU
`promotional material whether or not it accompanied the
`The re-
`striction of the labeling provision to "material" did, however, place the
`
`l9 "No advertising of a drug shall be deemed to be false i f it is disseminated only to
`members of the medical profession." Federal Trade Act of 1938, 5 15(a)(l), ch. 49, 52 Stat.
`1 1 1, 116. See Morton J. Simon, The Law of Advertising and Marketing 527 (1956).
`Hearings on a revised food and drug law were first held in 1933 before the Senate
`Commerce Committee. See Food, Drugs and Cosmetics: Hearings before a Subcomm. of
`the S. Commerce Comm. on S. 1944, 73rd Cong., 2d Sess. (1933). The biJl introduced
`included expanded controls over advertising: "An advertisement of a food, drug or cosmetic
`shall be deemed to be false i f in any particular it i s untrue, or by ambiguity or inference
`creates a misleading impression regarding such food, drug or cosmetic." S. 1944, 73rd
`Cong., Zd Sess. (1933). This vague section of the proposed Iaw was vehemently opposed by
`both patent medicine producers and newspaper and magazine interests. A revised bill, S. 5,
`74th Cong., 1st Sess., incIuding a misleading advertising section but explicitly reIieving
`publishers of responsibility, passed the Senate in 1935, but did not pass the House due to a
`jurisdictional dispute over FTC versus FDA regulation of advertising. I n 1938, the "elixir
`sulfanilamide" tragedy spurred the compromise that the Wheeler-Lea Amendment and the
`revised Food, Drug, and Cosmetic Act represented. See Peter Temin, The Origin of Com-
`pulsory Drug Prescriptions, 22 b. Law & Econ. 91 (1979), for discussion.
`21 Temin, supra note 20, at 97-98. The promotion of both prescription drugs and some
`designated over-the-counter drugs (together called the ethical drugs) can, by FDA regulatory
`degree, be directed only toward physicians. The ethical drugs currently comprise over 80 per
`cent of total drug sales. The major category of non-prescription ethical drugs is cough and
`cold preparations which comprise about l.5 per cent of total drug sales.
`This interpretationof the Food, Drug, and Cosmetic Act was upheld by Supreme Court
`in Kordel v. United States, 335 U.S. 345 (1948).
`
`

`
`PRESCRIPTION DRUG ADVERTISING
`
`51
`
`verbiage of the detail men (drug salesmen) beyond the direct control of the
`FDA.
`Beginning in 1958, the Senate Subcommittee on Monopoly, chaired by
`Senator Kefauver, held hearings on the pharmaceutical industry. Among
`other allegations, pharmaceutical promotion directed to physicians was
`claimed to be misleading, uninformative, and responsible for high drug
`The hearings culminated in the 1962 amendments to the Food,
`Drug, and Cosmetic Act. In addition to codifying extant FDA policy
`towards pharmaceutical promotion, the 1962 amendments required firms
`to submit to the FDA all promotional material relating to ethical drugs .24
`The amendments imposed other specific requirements for pharmaceutical
`advertising. The chemical name and quantitative composition of a drug
`were to be included in all advertisements, "printed prominently and in
`type at least half as large as that used for any trade or brand name.'12A
`statement "in brief summary" of side effects, contraindications, and
`effectiveness was also required. The effectiveness statement was inter-
`preted by the FDA to mean that "advertising was to avoid conveying a
`false, incomplete, or otherwise misleading impression that the drug is a
`panacea, or more effective or more safe than is medically justified." 26 The
`FDA does not provide explicit guidelines on what constitutes a misleading
`include "non-
`impression, though "commonly encountered defects"
`sequiturs, such as drawing conclusions or projections by inference; the use
`of anecdotal testimony; vague, open-ended claims o r suggestions of
`superiority; lack of appropriate contextual and physical fair balance; and
`casting the "brief summary in small black print contrasted to vividly
`colored, bold-lettered pr~motion."~'
`Since the 1962 amendments, the FDA has considered "all oral state-
`ments, in fact any form of communication, calling attention of medical
`and other audiences to a drug product"2R to be subject to their regulatory
`jurisdiction under the labeling provisions of the Food, Drug, and Cos-
`metic Act. Hence, the activity of detail men is in principle regulated by
`the FDA.
`
`23 See Hanis, supra note 8, for a very illuminating account of the legislative process,
`gaming behavior, and compromise that preceded the 1962 amendments to the Food, Drug,
`and Cosmetic Act.
`See Food and Drugs, 21 C.P.R. $ 3 I0.300,67b (1980). Labeling is takento includejournal
`ads, brochures, house magazines, price lists, literature reprints, and audio-visual materials.
`See FD Form 2253 (5f773, "Transmittal of Advenisements and Promotional Labeling."
`2s Drug Amendments of 1962, 76 Stat. 791.
`26 H. W. Chadduck, Division of Drug Advertising, FDA, I n Brief Summary: Prescription
`Drug Advertising, 1962- I97I, 6 FDA Papers, February 1972, at 13.
`27 Chadduck, supra note 26, at 2425.
`Id.
`
`

`
`52
`
`THE JOURNAL OF LAW AND ECONOMICS
`
`The controls over promotion by the FDA seem sufficiently stringent to
`make fraudulent advertising by pharmaceutical producers unprofitable.
`Drugs may only be promoted for approved uses, and the promotional
`material, including both journal advertisements and detail "handouts,"
`is scrutinized by the FDA. Companies are frequently required to rewrite
`their promotional literature and occasionally "asked"
`to run corrective
`advertisements .2g Of course, the compliance of detail men to FDA rules is
`conjectural. Detection of inappropriate verbal claims depends essentially
`on physician complaints.
`
`Prescription drugs are one of the most heavily promoted products in the
`American economy .30 Table 1 reports the expenditures on pharmaceutical
`promotion, the sales of these products through drug stores, and the pro-
`motion to sales ratios for the years 196 1 and 1968 to 1977.3' The promotion
`averaged about 13 per cent of sales over this period.32 As mentioned in
`Section I, drug advertising consists mainly of two types: advertising in
`professional journals and direct contact of physicians by drug salesmen
`(detail men). Table 1 breaks down promotion into these two types. In
`dollar terms, detail advertising dominates and is about three times that of
`journal advertising.
`The sales figures reported in Table 1 include only a segment of the
`
`29 Id, at 13, reports that by I97 I, th~rty-three "Dear Doctor" remedial letters and nineteen
`remedial advertisements had been issued. I n addition, six criminal and five civil cases had
`been brought on charges of false and misleading advertising. I n the civil cases, the govern-
`ment seized shipment of the drugs. Chadduck also reports that "regulatory actions" were
`taken against 5% advertisements and promotional pieces. Id. at 22. These actions include
`warning letters, advisory conferences, cancelled ads and mass letters of warning to the
`industry. See also, Medical Advertising: State of the Craft and of Regulation, FDA Papers,
`February l967; or A Primer of Medical Advertising (FDA 1970).
`30 Without citation, the Kefauver hearings (incorrectly) estimated a pharmaceutical promo-
`tion-to-sales ratio of 25 per cent. See Harris, supra note 8, at 89. F. M. Scherer, Industrial
`Market Structure and Economic Performance 329 (Ist ed. 19701, reports (also without docu-
`mentation) a promotion-to-sales ratio of 29 percent.
`3L The sales figures are calculated at wholesale prices and exclude the drug purchases of
`hospitals and the government and "over-the-counter" drugs at drug stores. The promotional
`expenditures exclude promotional "frtls" (pens, tablets, stethoscopes, and so forth), lob-
`bying expenditures and advertising d~rected towards consumers for over-the-counter drugs.
`See David Schartzman, Innovation in the Pharmaceutical Industry (1976), lor a more com-
`plete discussion of IMS data used in all the tables and regressions.
`32 Comawr & Wilson, supra note I, tab. 6-A.2, at l34-35, estimate that phmaceutical
`promotion was 9.9 per cent of sales from 1954-1957. Their sales figure includes hospital
`sales, over-the-counter sales, and saies of subsidiaries (lor example, medical equipment,
`chemicals). For the period examined by Cornanor-Wilson, the pharmaceutical advertising
`intensity was third highest of 41 industries (9.9 for pharmaceuticals, 10.3 for cereals, and
`15.3 for perfumes).
`
`

`
`PRESCRIPTION DRUG ADVERTISING
`
`TABLE 1
`DETA~L A N D JOURNAL PROMOTIONAL EXPENDITURES, DRUGSTORE SALES, AND
`
`PROMOT~ON-SALES RATIOS FOR THE ETHICAL DRUG INDUSTRY
`
`Journal
`Promotion
`Expenditures
`(000s $)
`
`Detail
`Promotion/
`Total
`F'romotion
`Sales
`Sales
`Promotiont
`Expenditures
`(ooo* $1
`(o@)s $1
`(000s $1
`Year
`(Percent)
`11.8
`1,720,562
`202,955
`40,813
`123,581*
`1961
`I %8
`86,779
`292,982
`14.9
`2,910,135
`433,900
`14.3
`3,054,956
`436,960
`100,810
`281,571
`1969
`102,285
`280,933
`I970
`12.7
`3,392,804
`431,127
`12.0
`3,640,798
`438,028
`10 1,850
`295,643
`1971
`101,550
`3I3,862
`I972
`12.4
`3,724,294
`456,097
`12.3
`4,025,844
`494,082
`111,455
`339,786
`1973
`115,246
`347,840
`1974
`4,411,746
`499,618
`11.3
`116,890
`352,191
`1975
`5 14,834
`10.4
`4,941,373
`425,293
`1976
`11.2
`5,422,025
`604,900
`132,38 1
`43 1,895
`1977
`630,870
`145,485
`10.4
`6,055,381
`52 I,726
`1978
`736,285
`163,634
`6,606,956
`II.1
`* Esnmated assumhg the ratla ofdetail tojournaladvenising esprnditurcs in I%I equalled the average ratta (3.03) from
`1968- 1977.
`t D~rect mad advert~stng plus detail and journal promatian
`Sourccss. Promotian dataare 6 o m Intercont~nental Medical Statistics (IMS), A M U ~ Intercaetruentd Medical Statistics,
`Mail. Sournal and Detailing Audits, and the dataansales from yearend arc from I M S Drug Slort and Haspital Putchasts.
`
`ethical pharmaceutical market, sales through retail drug sellers. Total
`ethical pharmaceutical sales to nongovernment buyers are about 20 per
`cent larger, the additional sales being accounted for by hospital pur-
`chases. There is probably less reliance on manufacturer advertising for
`hospital drug purchasers (formulary boards) since the gains from inde-
`pendent information sources (for example, research reports and price
`lists) are relatively concentrated as compared to drug store purchasers.
`Table 2 presents data confirming the relative unimportance of phar-
`maceutical promotion in 1978 for products sold to hospitals. Nine pairs of
`drugs (or drug classes) are compared that treat "similar" symptoms. One
`of each pair is purchased mainly by hospitals and the other by drug stores
`for resale to consumers. As shown in the table, the average promotion-to-
`sales ratio for the hospital-dominated products was 3.1 per cent, whereas
`those products directed towards drug store purchasers had a promotion to
`sales ratio nearly four times as large (12.2 per cent). Because of the
`unimportance of drug advertising for hospital drugs, this paper focuses
`exclusively on the promotion of drugs purchased by consumers through a
`physician's prescription.
`Treatment of disease via drug therapy is a n ever-changing technology
`characterized by high costs of information as to the most efficacious
`product. Schwartzman reports that about 22,000 prescription drugs were
`
`

`
`54
`
`THE JOURNAL OF LAW AND ECONOMICS
`
`TABLE 2
`COMPARIWN OF ADVERTISING TO SALES RATIOS IN 1979 FOR PRODUCTS PURCHASED BY
`(000s $1
`HOSP~TALS AND BY DRUGSFORES
`
`Hospital Use Dominant
`
`Product
`
`Inject. Anticoagulants
`Inject. Antihistamines
`Ampicillin
`Aminoglycosides
`PeniciUin G
`Keflin Neutral
`Xylocaine
`Cardiovascular
`Silvadene
`Vistaril
`
`Total
`ADV.ISALES RATIO
`
`Adv.
`Exp.
`
`120
`0
`494
`5,647
`606
`1,385
`11
`
`248
`2,533
`
`l 1,044
`
`Drug Store Use Dominant
`Adv.
`Exp.
`
`Sales
`
`Product
`
`21,911 Oral Anticoagulants
`2,405 Oral Antihistamines
`75,950
`Amoxicillin
`107,447 Tetracycline Congeners
`27,041 Penicillin V & VR
`82,096 Keflex
`2,716 Quinicane
`
`8,274 Neospnrin
`32,238 Atarax
`
`360,078
`,031
`
`374
`9,303
`l0,016
`13,656
`1,270
`7,024
`6
`
`1,778
`2,243
`
`Sales
`
`10,765
`50,124
`67,778
`68,450
`44,8l5
`104,746
`2,920
`
`10,675
`14,110
`
`available in the United States in the early 1 9 7 0 ~ ~ ~ The characteristics
`
`relevent for determination of the most efficient drug therapy in any par-
`ticular circumstance are manifold. Drug effects vary by patient age,
`weight, sex, diet, cleanliness, rest patterns, and work habits. Some pa-
`tients have difficulty in swallowing pills, in following dosage directions, or
`in tolerating nausea and other side effects. In addition to patient-specific
`factors, the relative efficacy of drugs treating, for example, infections, will
`vary by infection site, bacterial or viral form, other medication used con-
`currently, and past medications used.
`The cost to the inhvidual physician of information on a particular drug's
`effects (and, of course, the potentially high value of such information)
`provides the economic incentive for pharmaceutical firms to inform pre-
`scribers of the advantages (demand-increasing effects) of their products.
`Obviously, firms also have economic incentives to inform physicians of
`the adverse effects or inferiority of competing brands.34 Less obviously,
`the demand for a firm's product might increase by the firm's publicizing its
`product's adverse effects-even
`if these adverse effects are more serious
`than "average." This potential results, in part, from the value of known
`rather than unknown side effects. Knowledge (and anticipation) of ad-
`verse side effects can drastically lower the expected cost to the patient of
`
`a Supra note 3 1.
`34 Though, as indicated in Section I, "sumestions of superiority" are viewed with dis-
`favor by the FDA.
`
`

`
`PRESCMPTION DRUG ADVERTISING
`
`55
`
`drug therapy. For example, rapid recognition and treatment of an allergic
`reaction to penicillin can prevent death.
`In addition, when a drug firm produces (or expects to produce in the
`future) more than one product, provision of unfavorable information
`about one product can increase the demand for its other products. Detail
`men repeatedly sell physicians information-paid
`for by the opportunity
`cost of physicians' time. If patients frequently experience less favorable
`effects than the physician ha