John Maynard Keynes once observed, "Practical men…are usually the slaves of some defunct economist. Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler of a few years back. I am sure that the power of vested interests is vastly exaggerated compared with the gradual encroachment of ideas. Not, indeed, immediately, but after a certain interval; for in the field of economic…philosophy there are not many who are influenced by new theories after they are twenty-five or thirty years of age, so that the ideas which civil servants and politicians and even agitators apply to current events are not likely to be the newest." 
THE FTC'S OBSOLETE DOCTRINE
The Federal Trade Commission staff exemplifies Keynes' observation. Adopting an obsolete doctrine, it has devised a new policy toward advertising. This policy is expressed in the current FTC case against the cereal industry. In its complaint, the staff argues that "promoting trademarks through intensive advertising results in high barriers to entry into the RTE cereal market."  Members of the FTC staff have added, in various press conferences, that retail cereal prices would fall as much as 25% if they prevail.  The FTC's chief economist does not agree with the 25% figure, but he does believe prices might fall 7-10% if the FTC orders the dismemberment of cereal companies. 
These notions—that the cost of products to consumers is high because of advertising and that advertising is a barrier to entry behind whose shelter firms behave monopolistically—are both old  and obsolete. They are being adopted by the FTC staff as a basis for a complaint just at the time they are in the process of being discarded in economic analysis as erroneous—as inconsistent with economic theory and as unfounded. Behavior observed in the market place does not support these notions.
These ideas emerged among economists in the 1930's from a two-dimensional analysis of an hypothetical static world with costless information and unchanging tastes, technology, and resources (including population). In such a world, advertising performs no useful function. Advertising expenditures in these circumstances waste capital. 
THE MODERN VIEW
But that hypothetical world is not the one in which we live. Just as the mechanical and physical sciences have gone beyond the frictionless models originally used in the development of physical relationships, economic science has gone beyond the static, two dimensional models originally used to analyze economic relationships. Fourteen years ago Professor Harold Demsetz developed a multi-dimensional model of markets. The model allows for the fact that there is geography, that people forget, that information is not costless, that we live in a changing world, that many different qualities may be incorporated in products, and that products may be improved. 
Firms do more than simply decide how much to produce. They also choose the locations at which to offer their products, the expenditure to be made on selling and promotion, the quality to be incorporated in their products, and the budget to invest in improving their products. Demsetz shows that differences among products of a given specie and the advertising of these products in a multi-dimensional world does not result in a monopolistic outcome, contrary to the conclusions reached in the 1930's in the theory of monopolistic competition.  He also shows that the earlier theories contained logical inconsistencies producing spurious results. 
In 1961, Professor George Stigler published a seminal article entitled, "The Economics of Information."  The article's opening is very much worth quoting. It indicates the then current attitude among economic theorists toward advertising as well as indicating the inappropriateness of the standard economic models then employed for analyzing our economy.
One should hardly have to tell academicians that information is a valuable resource: knowledge is power. And yet it occupies a slum dwelling in the town of economics. Mostly it is ignored; the best technology is assumed to be known; the relationship of commodities to consumer preference is a datum. And one of the information- producing industries, advertising, is treated with a hostility that economists normally reserve for tariffs or monopolists. 
The important point made by Professor Stigler, and the theory he developed which has provided the foundation for a great deal of empirical research on the economics of advertising in the past decade, is that the cost of personal search by buyers for sellers in the absence of advertising is high.  Also, he pointed out that the personal search process is inefficient compared to the use of advertising, especially in view of the change in identity of sellers over time and even more because of the turnover of buyers. "In every consumer market there will be a stream of new buyers (resulting from immigration or the attainment of financial maturity) requiring knowledge of sellers, and, in addition, it will be necessary to refresh the knowledge of infrequent buyers." 
Instead of using the static, unchanging world model where buyers never grow old or forget and new buyers never appear—the model previously used by economists which generated hostility to advertising—Professor Stigler introduced the cost of obtaining information into his model. This makes it appropriate for the analysis of advertising. With this dynamic element, we find that advertising is a productive and procompetitive activity, substituting cheaply provided information for expensive search costs, rather than being a wasteful activity producing monopolistic results. In this model, it can be shown that advertising makes firm demand curves more elastic and, therefore, makes markets more competitive.  This conclusion is the direct opposite of the assumption on which Chamberlin and Robinson built their models in the 1930's.
EMPIRICAL SUBSTANTIATION OF THE MODERN VIEW
The obsolete Chamberlinian and Robinsonian models, which have been misapplied to multidimensionally competitive markets, assumed that advertising made demand facing the firm less elastic than it would otherwise be. Advertising, then, was implicitly assumed to make markets monopolistic. No proof was offered for this assumption in either theoretical or empirical terms. As Professor Richard Posner points out
Advertising is said to be an important factor in the diminished rivalry that is thought to characterize many oligopolistic markets. The reasoning is that advertising creates brand loyalties that rival sellers find very difficult to erode and that this is a source of formidable barriers to new entry into concentrated markets.…
…[N]o proof has yet been offered that it is easier for the first advertiser to win a consumer's patronage than it is for a second advertiser to shift it to him. The fact that the soap companies are constantly bringing out new brands suggests a taste for novelty on the part of the consumer that does not square with the theory of the first advertiser's advantage. 
Support for Professor Posner's view is provided by Professor Jules Bachman's examination of a number of markets. He finds behavior that indicates a lack of brand loyalty for many intensively advertised commodities. The decline of Rinso, Super Suds, and Oxydol from a 48 percent occupancy of the household soap and detergent market in 1948 to less than 5 percent in 1953; the fall in Camel, Lucky Strike, and Chesterfield market share from 42 percent in 1956 to 18 percent in 1966; the slide in market share of the four leading cold cereals from 38 percent in 1954 to 30 percent in 1964; Coca-Cola's slide from 53 percent of the soft drink market in 1940 to 37 percent in 1960, and similar experiences for a number of unidentified products all point to the fact that intensive advertising does not necessarily give a brand a lock on its market. 
Perhaps the products examined by Professor Bachman are exceptions. However, another study looking not simply at intensively advertised products but instead relating brand loyalty to levels of advertising finds that brand loyalty is weaker for more intensively advertised than for less advertised products.  The National Commission on Food Marketing also developed data some of which rebuts the presumption that advertising creates brand loyalties. In the case of RTE cereals, for example, it found that only 6% of all buyers showed sufficient brand loyalty to purchase their chief brand 75% of the time. In the case of coffee, despite the fact that the coffee industry advertised at less than half the rate at which cereals advertised, 70% of coffee purchasers showed strong brand loyalty as compared to only 6% of cereal buyers. 
Some would argue that because cereal buyers seek variety, less loyalty would be expected and that advertising makes them more loyal than they otherwise would be. This still assumes that advertising creates loyalty. The examination of advertising undertaken in the past decade, however, points in the opposite direction. Advertising is aimed at destroying loyalty. It is not aimed at the customers a firm already has. Advertising is aimed at informing customers the firm does not yet have. If the firm achieves its aim, it destroys the loyalty of the buyers of other brands and other products. Even Chamberlin allows that, "The effect of advertising is to shift to the right the demand curve for the advertised product by spreading knowledge of its existence, by describing it, and by suggesting the utilities it will provide the purchaser." 
The modern mathematical models successfully used for organizing data on consumer behavior, prices, and advertising are constructed on the basis of the expectation that advertising will have no influence on current customers of an advertiser or on their loyalty. As one of the developers of these models points out
Advertising is more likely to increase the probability of transitions to brand i…than to affect repeat purchases. Repeat purchases…result from consumer satisfaction with the brand. Advertising, however, is…a means for attracting purchases both from other brands and from outside. 
The lack of loyalty for a heavily advertised product is demonstrated by the readiness of beer consumers to desert their customary choice when the price of another brand is cut. When Anheuser-Busch reduced its per case price in St. Louis from $2.93 to $2.68 at the end of 1953, its market share rose by 33 percent in the next six months (from 12.5 percent to 16.55 percent). A further reduction to $2.35 sent its market share zooming to 39.3 percent, a 137 percent increase, in the next eight months.  These data hardly fit the image of the consumer hypnotized by advertising and unwilling to try any alternative to his present brand.
Professor Lester Telser has brought data to bear directly on the presumption that advertising is a barrier to entry and a means for extracting monopoly prices. Professor Telser found that
There is little empirical support for an inverse association between advertising and competition, despite some plausible theorizing to the contrary. I examined four pieces of evidence. First, for three cross-sections of manufacturing industries that make consumer products, concentration and advertising intensity are virtually independent. Second, there is a negative correlation between the change in concentration and the change in advertising intensity from 1947 to 1958. In some industries a rise in advertising accompanied a fall in concentration. Third, brand shares of food products are markedly more stable than brand shares of toiletries, although the latter are more heavily advertised than the former. Fourth, brands of toiletries have a shorter expected life than branded food items. This is shown by the larger changes around the levels reached by a group of leading brands and the lower levels attained by leading brands of toiletries as compared to branded food items. Advertising is frequently a means of entry and a sign of competition. This agrees with the view that advertising is an important source of information. 
Professor Telser's empirical work demonstrates that the theoretical apparatus provided by Professors Demsetz and Stigler makes better sense than the earlier static, two-dimensional models. Also, his work tells us that the Federal Trade Commission staff is wrong in its view that advertising shelters from the vengeance of competition firms engaging in oligopolistic (shared monopoly) behavior. Advertising is a means of competing—not a barrier to entry. Professor Telser has substantiated these latter propositions.
DEFICIENCIE5 IN THE FTC'S SUBSTANTIATION
The Federal Trade Commission staff says there is substantiation for its presumption, however. It points primarily to two studies. One is a paper by Professors Comanor and Wilson. They put together data on the ratio of advertising to sales by industry groups, relate these to accounting rates of return, and arrive at the conclusion that
On the basis of these empirical findings, it is evident that for industries where products are differentiable, investment in advertising is a highly profitable activity. Industries with high advertising outlays earn, on average, at a profit rate which exceeds that of other industries by nearly four percentage points. This differential represents a 50 per cent increase in profit rates. It is likely, moreover, that much of this profit rate differential is accounted for by the entry barriers created by advertising expenditures and by the resulting achievement of market power. 
There is, however, a major methodological flaw in the regressions used by Comanor and Wilson for organizing their data. They treat outlays for advertising as a current expense. They do this despite their explicit recognition that the current position of a firm in a market which demands advertising and product jointly depends upon past advertising outlays. They cite Professor Kristian Palda,  who demonstrated that advertising should be treated as an investment, but then disregard this. They even speak of "investment in advertising," but then expense all such outlays in their statistical treatment.
Their conclusion that, "Industries with high advertising outlays earn, on average, at a profit rate which exceeds that of other industries by nearly four percentage points," is reached by using understated net worths for the firms in such industries. Since firms do not capitalize their investment in advertising for balance sheet purposes, the omission of this asset leads to an understatement of net worth which is large for firms with high advertising sales ratios. As a consequence, the rate of return is overstated for these firms. For example, a firm spending 10% of sales on advertising earning 8% on sales whose stated equity equals 40% of sales will show a 20% accounting rate of return. If this firm's advertising has a ten year life and its sales and advertising have been level during the past ten years, then current outlays on advertising will be equal to depreciation of past advertising. Its current income will be unaffected by capitalizing current outlays and straight line depreciating past outlays. Its equity, however, will be increased by the depreciated value of capitalized past advertising. It will become equal to 100% of sales. An 8% margin on sales, then, will show an 8% return on corrected equity instead of 20% on stated equity.
Comanor and Wilson have discovered only the fact that the upward bias in accounting rates of return for "industries with high advertising outlays" is four percentage points greater than the upward bias "of other industries." Since they did not use true rates of return, or proportionately biased accounting rates of return, they have not substantiated the presumption that advertising is a barrier to entry. 
The second study on which the FTC staff relies to substantiate this presumption is its own Economic Report on the Influence of Market Structure on the Profit Performance of Food Manufacturing Companies (1969). In this study, a positive relationship again is found between advertising intensity and accounting profit rates. Again, however, the same methodological deficiency is present.
To see whether or not advertising truly serves as a barrier to entry, Professor Harry Bloch re-did the FTC study taking account of the fact that advertising has long lived effects.  The fact that it has such effects has been established not only by Professor Palda's work, but also by Professor Telser, Professors Nerlove and Waugh, and many other analysts. 
Professor Bloch finds that the firms used by the FTC in its study do not show a positive relationship between advertising and profit rates after correcting the FTC study for its data and methodological deficiencies. The FTC study did not use actual advertising expenditures of the firms it selected for its work. Rather, it imputed to those firms the same level of advertising as the average level found in their industries. Professor Bloch used their actual advertising expenditures. In addition, he treated advertising as an investment.
With these corrections of the FTC study, it no longer substantiates the staff view that advertising provides a monopoly shelter. With these corrections, the opposite is substantiated. Advertising is a means of entry and an ordinary investment. It is no more a barrier to entry than any other investment, such as that in plant and equipment. Advertising and competition are compatible—not antithetical. Investment in advertising yields the same return as investment in tangible assets—a result to be expected under competitive circumstances (assuming risks are identical) and evidence supporting the view that advertising is a means of entry just as building production facilities is a means of entry.
In addition to this empirical work demolishing the studies the FTC staff offers to substantiate its view that advertising is a barrier to entry, Professor John Gould has developed a dynamic model explaining the time path followed by the advertising for a product during its introduction into the market and its subsequent life cycle.  What is important and interesting about Professor Gould's work is that it demonstrates that outlays on advertising can be explained in exactly the same way as the accumulation of a capital stock in any other form.
Insofar as the model's premises are verified by the behavior we find in the market, it supports the view that advertising is an investment and a cost of entry. It is not, then, a barrier to entry any more than any other cost of entry. 
There is an abundance of evidence supporting Professor Gould's view.  We all know that advertising outlays are very frequently much larger relative to sales in the first year or two after introduction of a product than in subsequent years. Sometimes, they are so large relative to sales early in the life of a product that no rational man could possibly argue that they are a current expense. A drug manufacturer does not spend 80 to 150% of the first year's sales of a drug promoting and advertising the drug  in the expectation that the expenditure will be currently profitable. That is obviously impossible. The first year outlay must be looked upon as an investment upon which a return will be realized in future years—not in the year the outlay is made.
The National Commission on Food Marketing found that the advertising to sales ratio for new products in the cereal industry averages 47% in the first year declining to 22% in the second year. It also found the typical new cereal did not reach a break even point until after it had been on the market at least four years.  Again, it is obvious that the first and second year expenditures on advertising in this instance are investments—not a current expense—which supports Professor Gould's view of outlays on advertising being a capital accumulation process.
Since advertising outlays can be shown to be investments, the econometric studies showing advertising and profitability to be positively related by treating advertising as an expense fail to substantiate the view that advertising is a barrier to entry. Only if it can be shown that where entry is attractive, imperfections in the capital market prevent potentially viable, would-be competitors from raising funds to enter the production and marketing of attractive, advertising intensive products can it be argued that the required investment in advertising is a barrier to entry. There has been no such demonstration. There is much work to the contrary showing that capital markets are efficient. 
A second line of argument for the view that advertising is anti-competitive is that there are substantial economies of scale in advertising. This, presumably, increases the minimum efficient size of a firm within a given market and leads to increased concentration. Professor Telser has demonstrated that there is no empirical positive relationship between advertising intensity and concentration.  Even if there were to be such a relationship, there is no positive relationship between concentration and monopoly. 
Professor James Ferguson has examined the evidence for the proposition that there are economies of scale in advertising. He concludes that, "The scanty available evidence suggest diminishing returns to advertising."  Similarly, Richard Schmalensee concludes that "there is no evidence to suggest that the minimum efficient size of the firm is increased by advertising in general or in any particular industry." 
A third line of argument for the notion that advertising creates a barrier to entry rests on the presumption that advertising creates brand loyalties that rival sellers find difficulty in eroding. Because of this presumed difficulty the investment in advertising required of them is presumed to be greater than that required of their forerunners. What little evidence is available on brand loyalty, however, seems to indicate a negative relationship between advertising and brand loyalty.
It may even be plausibly argued that insofar as advertising by any one firm provides information to "free rider" consumers who use the information in deciding to purchase a product, but then purchase a non-advertised brand, advertising may reduce the cost of entry to some producers or marketers of products. The existence of little-advertised private label products in many markets alongside intensively advertised brands  is evidence consistent with the view that high advertising industries are easy to enter. Most marketers of vodka do little advertising, relying on one major marketer to advertise recipes for vodka drinks and suggest the circumstances appropriate for serving such drinks. Recently, each of the prospective marketers of white whiskey was hanging back waiting for one of the others to invest in developing the white whiskey market. 
These examples suggests that we are getting less than the optimal amount of advertising because entry is eased by the advertising of any one marketer. Others hope for an opportunity to grab the coattails of an advertiser and avoid any expense on their part in providing information. If that is the case, then advertising invites and creates entry. It facilitates entry rather than being a barrier to entry. 
Yale Brozen is Professor of Business Economics at the Graduate School of Business, University of Chicago and Adjunct Scholar, American Enterprise Institute for Public Policy Research. Professor Brozen's article is based on the highly acclaimed paper which he presented before the Annual Meeting of the American Bar Association Section of Antitrust Law in San Francisco last August.
NOTES AND REFERENCES
 GENERAL THEORY OF EMPLOYMENT, INTEREST AND MONEY 383-84 (1936).
 Docket No. 8883.
 DETROIT FREE PRESS, 21 June 1971; WALL STREET JOURNAL, 25 January 1972.
 "Does advertising in practice promote industrial concentration, restrict the entry of newcomers and unnecessarily raise prices above the level they would be at in its absence? The question is a matter of fact, not only of theory. Nevertheless, since the work of Professor E.H. Chamberlin and Mrs. Joan Robinson in the 1930's most economists have inclined to the belief that advertising does have these adverse effects." H. Gray, "Economics of Advertising: A Summary of Sources," 11 THE ADVERTISING QUARTERLY 11 (Spring 1967).
 "The very nature of a purely competitive market precludes a selling problem.…Advertising would be without purpose under conditions of pure competition, where any producer can sell as much as he pleases without it." E.H. Chamberlin, THE THEORY OF MONOPOLISTIC COMPETITION 10, 72 (1933).
 Harold Demsetz, "The Nature of Equilibrium in Monopolistic Competition," 67 JOURNAL OF POLITICAL ECONOMY 21 (1959). The model used by Professor Demsetz was anticipated by George J. Stigler, THE THEORY OF PRICE 260-63 (1946).
 E.H. Chamberlin, supra note  and Joan Robinson, THE ECONOMICS OF IMPERFECT COMPETITION (1933).
 Harold Demsetz, "Do Competition and Monopolistic Competition Differ?" 76 JOURNAL OF POLITICAL ECONOMY 146 (1968).
 69 JOURNAL OF POLITICAL ECONOMY 213 (1961).
 Id. at 213.
 Id. at 216.
 Id. at 220.
 "[F]or experience goods we would expect an increase in the elasticity of demand as a result of advertising.…[F]or search goods advertising increases the consumer's knowledge of the utility distribution of brands. It can be shown…that the elasticity of demand is highly sensitive to the consumer's knowledge of the mean of the utility distribution of brands. Hence, one would expect advertising to increase the elasticity of demand for search goods as well." Philip Nelson, "Advertising As Information" (unpublished paper, 1972) at 32.
 U.S. Congress Subcommittee on Monopoly of the Senate Select Committee on Small Business, ROLE OF THE GIANT CORPORATIONS, Part 1-A, July 1969, p. 923.
 Jules Bachman, ADVERTISING AND COMPETITION 63-78 (1967).
 Lester Telser, "Advertising and Competition," 72 JOURNAL OF POLITICAL ECONOMY 537 (1964).
 National Commission on Food Marketing, STUDIES OF ORGANIZATION AND COMPETITION IN GROCERY MANUFACTURING, Technical Study No. 6, p. 198.
 Supra note  at 119 (7th ed. 1956).
 Lester Telser, "The Demand for Branded Goods as Estimated from Consumer Panel Data," 44 REVIEW OF ECONOMICS AND STATISTICS 300 (1962).
 Cited in id. at 324, n.23 from FTC v. Anheuser-Busch, Inc., 80 S. Ct. 1267 (1960).
 Lester Telser, "Advertising and Competition," 72 JOURNAL OF POLITICAL ECONOMY 537, 558 (1964).
 W.S. Comanor & T.A. Wilson, "Advertising Market Structure and Performance," 49 REVIEW OF ECONOMICS AND STATISTICS at 437 (1967).
 THE MEASUREMENT OF CUMULATIVE ADVERTISING EFFECTS (1964).
 Leonard Weiss, "Advertising, Profits, and Corporate Taxes," 51 REVIEW OF ECONOMICS AND STATISTICS 421 (1969), attempts to correct the methodological deficiency of the Comanor and Wilson study. After correction by capitalizing advertising expenditures over a six year period, he finds that there is still a positive relationship between profits and advertising intensity. As he points out, however, "It is possible that the net relationship between advertising and profit rates would fall to nonsignificance if ads could be depreciated over more realistic lives (perhaps ten years instead of six)." Id. at 428. Professor Weiss uses only six years of advertising expenditures for lack of more complete data from the IRS Source Book.
Professor Harry Bloch finds that both Weiss and Comanor and Wilson commit a fundamental error in using industry aggregates. Advertising intensity effects can be appropriately measured only on a firm basis. "Advertising and Profitability: A Reappraisal," DISCUSSION PAPER 81, Department of Economics, University of British Columbia (1972).
 ADVERTISING, COMPETITION, AND MARKET PERFORMANCE (Ph.D. dissertation, University of Chicago, 1971).
 L. Telser, "Advertising and Cigarettes," 70 JOURNAL OF POLITICAL ECONOMY 47 (1962), M. Nerlove & F. Waugh, "Advertising Without Supply Control," 3 JOURNAL OF FARM ECONOMICS 813 (1961); Yoram Peles, "Rates of Amortization of Advertising Expenditures," 79 JOURNAL OF POLITICAL ECONOMY 1032 (1971); Jacques Lambin, "Advertising and Competitive Behavior," Louvain University, Belgium, Center of Socio-Economic Studies in Advertising and Marketing (CESAM), Working Paper #2, 1969; M.L. Vidale & H.B. Wolfe, "An Operations Research Study of Sales Response to Advertising," 5 OPERATIONS RESEARCH 370 (1957).
 John Gould, "Diffusion Processes and Optimal Advertising Polity," in MICROECONOMIC FOUNDATIONS OF EMPLOYMENT AND INFLATION THEORY (1970).
 The distinction between a barrier to entry and the cost of entry is discussed by George J. Stigler, ORGANIZATION OF INDUSTRY at 87 (1968).
 See footnote , supra, and the following discussion.
 David Schwartzman, "Size of Firm, Research, and Promotion in the Drug Industry" (unpublished paper, 1972).
 STUDIES OF ORGANIZATION AND COMPETITION IN GROCERY MANUFACTURING, Study No. 6 (1969), pp. 37-38.
 Eugene F. Fama, "Efficient Capital Markets: A Review of Theory and Empirical Work," 25 JOURNAL OF FINANCE 383 (1970).
 Lester Telser, "Advertising and Competition," supra note .
 Yale Brozen, "The Antitrust Task Force Deconcentration Recommendation," 13 JOURNAL OF LAW & ECONOMICS 279 (1970); Y. Brozen, "Bain's Concentration and Rates of Return Revisited," 14 JOURNAL OF LAW & ECONOMICS 351 (1971); Y. Brozen, "The Persistence of 'High Rates of Return' in High-Stable Concentration Industries," 14 JOURNAL OF LAW & ECONOMICS 501 (1971); Harold Demsetz; "Why Regulate Utilities," 11 JOURNAL OF LAW & ECONOMICS 55 (1968).
 James M. Ferguson, "Empirical Studies of Advertising and Competition" at 2 (unpublished, April 1972).
 Richard Schmalensee, ON THE ECONOMICS OF ADVERTISING (1972).
 Jules Bachman, supra note  at 58-59.
 "Advertising: New Light Whiskey," NEW YORK TIMES 13 June 1972.
 Harold Demsetz, "The Effect of Consumer Experience on Brand Loyalty and the Structure of Market Demand," 30 ECONOMETRICA 33 (1962); Yale Brozen, "Barriers Facilitate Entry," 14 ANTITRUST BULLETIN 851 (1969).