Essays on Advertising and Price Collusion

Abstract

Chapter I: Persuasive advertising alters consumers' tastes and creates brand loyalty. The established view in the economics literature is that such advertising is anti-competitive and results in higher prices. This paper demonstrates that this is not necessarily true. This is shown in a model of a duopoly with horizontal product differentiation, where firms interact repeatedly over an infinite horizon. Firms in such a market try to increase their profits by avoiding price competition. They do this by colluding on price while making independent decisions on advertising. This practice is called price semicollusion. However, collusion on price leads to intensified advertising, which may lower firm profits to below the competitive level when advertising is market-stealing (rather than market-expanding). In such a case the collusion on price would break down and firms would revert to price competition. Thus, persuasive advertising can induce price competition. Moreover, the paper shows that the equilibrium price in a market with persuasive advertising can be lower than the price without it. This contradicts the prevalent view on the effect of persuasive advertising.

Chapter II: This paper examines the effect of advertising on price collusion using data on price fixing across manufacturing industries in the United States. I construct an original dataset from summaries of price fixing cases initiated by the Department of Justice between 1960 and 2003. In determining if advertising hinders or facilitates price collusion, the paper makes a distinction between market-expanding and market-stealing advertising. The need for the distinction between the two kinds of advertising is driven by the theoretical model outlined in the paper. The model shows that price collusion results in intensified advertising. This could undermine the gains from collusion if advertising is market-stealing rather than market-expanding. The paper identifies two types of industries where advertising is more market-stealing: (1) Industries with low market growth and (2) Industries with high product differentiation. The econometric results from a probit model provide evidence that supports the theory. The incidence of collusion is found to be lower in low-growth industries with high advertising. Collusion is also found to be less likely in product differentiated industries with high advertising.