Competitive targeted online advertising

This paper examines how an online publisher utilizes its information about consumer preference to target advertising. In our model, two firms first bid for a prominent ad position in a publisher-organized position auction, and then compete on price in the subsequent product marketplace. We consider two dimensions of consumer heterogeneity. First, consumers are heterogeneous in product preference. Based on their tastes, some consumers prefer one product over the other, whereas other consumers may rank the products in an opposite order. Second, consumers differ in search preference, i.e., “nonshoppers” only consider the advertised product, while “shoppers” always search both firms’ products before buying. We show that targeted advertising based on product preference will mitigate price competition in product markets as well as competition in position auctions, the latter to the detriment of the publisher. In contrast, targeted advertising based on search preference always benefits the publisher, as the winning firm can charge monopoly prices to nonshoppers. We show that the publisher’s optimal choice is to utilize only the information about consumer search preference. We also explore the welfare implications of targeted advertising based on different types of consumer preference.     

Strategic segmentation of a market

When competing firms target information towards specific consumers through direct marketing activities, complete segmentation of markets can result. We analyze a two-stage duopoly where, prior to price competition, each firm targets information to specific consumers and only consumers informed by a firm can buy from it. This has the effect of endogenously determining market segments in a model of ‘sales'. In equilibrium, pure local monopoly emerges; firms target and sell to mutually exclusive market segments. When the cost of marketing approaches zero, market shares reflect relative production efficiency (equal shares when firms are symmetric); this may not be the case when marketing cost is high.     

Ordered search in differentiated markets

This paper presents an ordered search model in which consumers search both for price and product fitness. I construct an equilibrium in which there is price dispersion and prices rise in the order of search. The top firms in consumer search process, though charge lower prices, earn higher profits due to their larger market shares. Compared to random search, ordered search can induce all firms to charge higher prices and harm market efficiency.     

The dynamic provision of product diversity under duopoly

This paper builds a dynamic duopoly model to examine the provision of new varieties over time. Consumers experience temporary satiation, and hence higher consumption of the current variety lowers demand for future varieties. The equilibrium can be characterized by a combination of monopolistic pricing and nearly zero profits (competitive timing). In particular, if the cost of producing a new variety is not too low then firms tend to avoid head-to-head competition and set the short-run profit maximizing price. However, firms tend to introduce new varieties as soon as demand has grown sufficiently to cover costs. From a second best perspective, the equilibrium may exhibit excessive product diversity. However, if firms coordinate their frequency of new product introductions, then consumers are likely to be harmed. It is also shown that equilibrium prices are moderated by two factors. First, consumers’ option value of waiting reduces their willingness to pay. Second, competition reduces firms’ incentives to engage in intertemporal price discrimination.     

Search and price dispersion in online grocery markets

Consumers often purchase multiple products at a time from retailers, creating multi-product incentives for search. In this paper we consider how product variety affects consumer search intensity and the dispersion of prices in multi-product retail markets. We employ online grocery pricing data from four large retailers in the UK to estimate search costs and equilibrium price dispersion for food products under circumstances where: (i) consumers search for single products; and (ii) consumers search for multiple products at once. We compare estimates in each case between a model in which utility increases with product variety and a model in which utility is not a function of variety. Relative to our preferred specification with variety effects in utility, we find estimates of both search cost and search frequency to be biased upwards in single product settings when variety effects are ignored; however, we find estimates of search costs are biased upwards while search frequency is biased downwards in multi-product settings when variety effects in utility are ignored.     

Consumer referrals

In many industries, firms reward their customers for making referrals. We analyze a monopoly’s optimal policy mix of price, advertising intensity, and referral fee when buyers choose to what extent to refer other consumers to the firm. When the referral fee can be optimally set by the firm, it will charge the standard monopoly price. The firm always advertises less when it uses referrals. We extend the analysis to the case where consumers can target their referrals. In particular, we show that referral targeting could be detrimental for consumers in a low-valuation group.     

A LOCATION MODEL WITH PREFERENCE FOR VARIETY*

We propose a new location model where consumers are allowed to make multiple purchases (i.e., one unit from each firm). This model may fit many markets (e.g. newspapers, credit cards) better than existing models. A common feature of these markets is that some consumers are loyal to one brand, while others consume more than one product. Our model yields predictions consistent with this observation. If firms are allowed to choose their locations on the interval, then spatial differentiation may not be maximal and in some cases it may even be minimal. Thus, under certain conditions, we restore Hotelling's Principle of Minimum Differentiation.     

Endogenous Mergers of Complements with Mixed Bundling

This paper studies endogenous mergers of complements with mixed bundling, by allowing both for joint and separate consumption. After merger, partner firms decrease the price of the bundled system. In addition, when markets for individual components are sufficiently important, partner firms find it strategically advantageous to raise the prices of stand-alone products, thus making substitute ‘mix-and-match’ composite products less attractive to consumers. Even though these effects favor the profitability of mergers, merging is not always an equilibrium outcome. The reason is that outsiders respond by cutting their prices to retain their market share, and mergers can be unprofitable when competition is intense. From a welfare analysis, we observe that the number of mergers that are observed in equilibrium may be either excessive (when markets for individual components are important) or suboptimal (when markets for individual components are less important).     

Costly Information Disclosure in Oligopoly

We examine the effect of competition on the incentive of firms to disclose quality to consumers before trade when information disclosure is not costless. We demonstrate that no firm will disclose information in the limit, no matter how small the disclosure cost is; that is, the market outcome converges to complete concealment of information as the number of competing firms becomes larger. Nonetheless, it can be shown that under a mild condition, the equilibrium amount of information disclosure is socially excessive for any number of firms, so discouraging information disclosure by levying a tax may increase social welfare.     

Anti‐Competitive Interconnection: the effects of the elasticity of consumers' expectations and the shape of the network effects function

I find that interconnection might cause the market to be less competitive, and might lead to an increase in the price firms charge for their product. Absent interconnection, firms compete for a consumer for two reasons. The first reason is to obtain revenue from selling the product to a consumer (as in the case without network effects). The second reason is that by expanding the network by one more consumer, the product becomes more attractive to all other consumers. Interconnection eliminates the second reason—when firms interconnect, they are no longer concerned with consumers' following the crowd. I show that consumers and society might be worse off from interconnection. I focus on two factors that make the (post‐interconnection) price increase larger: consumer expectations that are highly sensitive to prices and consumers putting a high value on small increases in network size at the equilibrium market shares. Both of these factors make firms highly competitive, but only if the firms' products' networks are not interconnected.     

A more general model of price complexity

This paper analyses a model of competition where the firms set not only prices but also the complexity levels of their prices (which determine how difficult it is for consumers to assess the price offers). Unlike previous work, in this model, the firms’ confusion technology may be non-linear in the aggregate complexity level. The equilibrium probability of using high complexity increases in the number of firms but decreases in the convexity of the confusion technology. In large markets, the firms use high complexity almost surely. However, the industry profit converges to the highest level with concave technologies and to the lowest level with convex technologies. An increase in consumer sophistication, which benefits the consumers, may not reduce market complexity.     

Behavior based price personalization under vertical product differentiation

We study price personalization in a two period duopoly with vertically differentiated products. In the second period, a firm not only knows the purchase history of all customers, as in standard Behavior Based Price Discrimination models, but it also collects detailed information on its old customers, using it to engage in price personalization. The analysis reveals that there exists a natural market for each firm, defined as the set of customers that cannot be poached by the rival in the second period. The equilibrium is unique, except when firms are ex-ante almost identical. In equilibrium, only the firm with the largest natural market poaches customers from the rival. This firm has highest profits but not necessarily the largest market share. Aggregate profits are lower than under uniform pricing. All consumers gain, total welfare is higher herein than under uniform pricing if firms’ natural markets are sufficiently asymmetric. The low quality firm chooses the minimal quality level and a quality differential arises, though the exact choice for the high quality depends upon the cost specification.     

POSTING MULTIPLE PRICES TO REDUCE THE EFFECTIVENESS OF CONSUMER PRICE SEARCH*

In a model of competition with imperfect consumer price information and incomplete price search, some consumers may end up comparing prices originating from the same supplier: either because one firm sets multiple prices or because a group of firms colludes. This leads to added monopoly power for these firms, and average prices in the mixed strategy equilibrium become higher. There is a shift in welfare from consumers to producers, both with exogenous and endogenous consumer search behaviour. However consumers might search more or less with multiple prices. The implications for the price‐setting equilibrium, competition policy and recent judgements are considered.     

The benefits of being small: Duopolistic competition with market segmentation

Consider a duopolistic market in which consumers are not necessarily aware of the firms' existence. The market is characterized by the existence of four segments: a duopolistic segment which consists of consumers who are aware of both firms, a segment of consumers who are unaware of either firm and two captive market segments. We assume that by advertising, firms control the proportion of consumers who are aware of their existence. The relative sizes of the four segments affect the equilibrium of the duopolistic pricing game. We show that being large may be disadvantageous, and that even if gaining awareness is costless firms may wish to remain small.We would like to thank Paul Klemperer and an anonymous referee for valuable comments.     

Supply function equilibria with capacity constraints and pivotal suppliers

The concept of a supply function equilibrium (SFE) has been widely used to model generators' bidding behavior and market power issues in wholesale electricity markets. Observers of electricity markets have noted how generation capacity constraints may contribute to market power of generation firms. If a generation firm's rivals are capacity constrained then the firm may be pivotal; that is, the firm could substantially raise the market price by unilaterally withholding output. However the SFE literature has not fully considered the impact of capacity constraints and pivotal firms on equilibrium predictions. We characterize the set of symmetric supply function equilibria for uniform-price auctions when firms are capacity constrained and show that this set is increasing as capacity per firm rises. We provide conditions under which asymmetric equilibria exist and characterize these equilibria. In addition, we compare results for uniform-price auctions to those for discriminatory auctions, and we compare our SFE predictions to equilibrium predictions of models in which bidders are constrained to bid on discrete units of output.     

Vertical equilibrium in a competitive input market

Imperfect competition, imperfect information, disequilibrium, and rationing all induce vertical integration in relatively straightforward ways. But can vertical integration arise when firms are competitive and markets clear rapidly? This paper demonstrates that a vertical equilibrium can be generated when the intermediate market is subject to external fluctuations and when there are economies of coordination for firms which avoid the market via integration. The vertical equilibrium is one in which some firms will be integrated while some will not, despite the fact that all firms are identical. Thus, one need not conclude that differential degrees of integration within industries are the result of differing circumstances among firms.     

The effects of personal information on competition: Consumer privacy and partial price discrimination

This article studies the effects of consumer information on the intensity of competition. In a two dimensional duopoly model of horizontal product differentiation, firms use consumer information to price discriminate. I contrast a full privacy and a no privacy benchmark with intermediate regimes in which the firms can profile consumers only partially. I show that with partial privacy firms are always better-off with price discrimination: the relationship between information and profits is hump-shaped. In particular, competing firms prefer to target consumers with partial but asymmetric information about preferences. Instead, consumers prefer either no or full privacy in aggregate, but the effects of information on individual surplus are ambiguous: there are always winners and losers. Finally, I study the information acquisition incentives of the firms when there is an external data seller. When this upstream data broker holds partially informative data, an exclusive allocation arises. Instead, when data is fully informative, each competitor acquires consumer data but on a different dimension. These findings are relevant for the strategic decisions of firms active in digital markets and contribute to the policy debate surrounding privacy, exclusive access to data and competition.     

Equilibrium price dispersion with heterogeneous searchers

Firms simultaneously set prices in a homogeneous-product market where uninformed consumers search for price information. Some uninformed consumers are “local” searchers who visit only one seller, whereas others search sequentially with an optimal reservation price. Equilibrium prices may follow a mixture distribution, with clusters of high and low prices separated by a zero-density gap. When the (exogenous) reservation price of local searchers depart from that of the optimizing sequential searchers by a relatively small amount, the presence of local searchers either has no effect on market outcomes or benefits all consumers. A reduction in search cost sometimes leads to higher equilibrium prices.     

How Much Do Employers Learn from Referrals?

This paper tests the hypothesis that referrals from various sources provide employers with more information about job applicants than they would have with‐out a referral. The study uses data that contain information on two workers in the same job, allowing the differences in job and firm characteristics to be canceled out and controlling for the possibility that workers with referrals from different sources (or no referral at all) sort into jobs that put different weights on individual performance. The estimation results are consistent with referrals from current employees providing employers with more information than they would have otherwise. Additionally, it appears as though hiring through friends or relatives of the employer may involve some favoritism that results in employers either collecting less information than they would otherwise or ignoring information when setting wages. The study finds weak evidence consistent with referrals from other firms or labor unions providing useful information, and no evidence that referrals from community organizations or other sources have any effect.     

Asymmetric Price Effects of Competition

When price dispersion is prevalent, a relevant question is what happens to the whole distribution of equilibrium prices when the number of firms changes. Using data from the gasoline market in the Netherlands, we find, first, that markets with N competitors have price distributions that first‐order stochastically dominate the price distributions in markets with N+1 firms. Second, the effect of competition is stronger for the medium to upper percentiles of the price distribution. Finally, consumer gains from competition are larger for relatively well‐informed consumers. To account for these empirical patterns, we extend Varian's [1980] model by allowing for richer heterogeneity in consumer price information.