Comparing predicted prices in auctions for online advertising

Online publishers sell opportunities to show ads. Some advertisers pay only if their ad elicits a user response. Publishers estimate response rates for ads in order to estimate expected revenues from showing the ads. Then publishers select ads that maximize estimated expected revenue.By taking a maximum among estimates, publishers inadvertently select ads based on a combination of actual expected revenue and inaccurate estimation of expected revenue. Publishers can increase actual expected revenue by selecting ads to maximize a combination of estimated expected revenue and estimation accuracy.     

Entry into auctions: An experimental analysis

This paper investigates entry decisions into first and second-price auctions using an experimental design to extract information on willingness-to-pay to enter (WTE). We find that subjects tend to overpay to enter both auction formats. In particular, if the subjects believe they will be bidding against bidders following the risk-neutral Nash strategy, their WTE is greater than the optimal risk-neutral amount 97% of the time for first-price auctions (FPA) and 90% for second-price auctions (SPA). If they believe that they are bidding against subjects who bid as do the other subjects, they submit a WTE that is too high 92% of the time for FPA and 69% of the time for SPA. We also find, in line with previous studies, significant overbidding in both the FPA and SPA. We then investigate whether introducing risk aversion (RA) or “joy of winning” (JOY) can explain the joint observation of over-entry and overbidding. In particular, using bid data alone, we structurally estimate three models, one allowing RA only, one allowing for JOY only and one allowing for both RA and JOY. While a model with JOY alone overestimates WTE, we find that RA alone can explain 38% of WTE but a model with both RA and JOY (where RA is estimated using FPA bids, and JOY is estimated using SPA bids) can explain 65% of WTE. Moreover, JOY appears to explain nearly all of the male WTE but only 44% of the female WTE.     

Fighting collusion in auctions: An experimental investigation

The danger of collusion presents a serious challenge for auctioneers. In this paper, we compare the collusive properties of two standard auctions, the English auction and the first-price sealed-bid auction, and a lesser-known format, the Amsterdam (second-price) auction. In the Amsterdam auction, the highest losing bidder earns a premium for stirring up the price. We study two settings: in one, all bidders can collude, and in another, only a subset is eligible. The experiments show that the Amsterdam auction triggers less collusion than the standard auctions. We compare experimental results to theoretical predictions, and provide an explanation where they differ.     

Auctions with toeholds: An experimental study of company takeovers

We run experiments on English Auctions where the bidders already own a part (toehold) of the good for sale. The theory predicts a very strong (“explosive”) effect of even small toeholds. While asymmetric toeholds do have an effect on bids and revenues in the lab, which gets stronger the larger the asymmetry, it is not nearly as strong as predicted. We explain this by analyzing the flatness of the payoff functions, which leads to large deviations from the equilibrium strategies being relatively costless. This is a general fundamental weakness of this type of explosive equilibria, which makes them fail when human players are involved. Our analysis shows that a levels of reasoning model explains the results better where this equilibrium fails. Moreover, we find that although big toeholds can be effective in a takeover battle, the cost to acquire them might be higher than the strategic benefit they bring.     

Uncertain regulatory timing and market dynamics

Using a dynamic model of capacity accumulation, I examine the relationship between uncertainty about the timing of a new Pigouvian tax and oligopolistic competition. I find that for some market structures uncertainty about the timing of the regulatory change leads firms to increase investment. These results stem from the nature of the uncertainty and its interaction with firms' strategic incentive to engage in capacity races. They dramatize the importance of accounting for initial conditions when forecasting firms' reactions to anticipated regulatory changes. In addition, I find that more protracted uncertainty leads to greater welfare costs.     

Vertical mergers in procurement markets

This paper uses computational methods that reveal ambiguous strategic effects of vertical mergers in a duopoly setting featuring incomplete information about sellers' costs, and differences in sellers' productive capabilities. First, vertical mergers can be jointly unprofitable. Second, the buyer's preferred merger partner is almost always the seller with lower expected costs, and is typically the larger seller. Third, vertical mergers always reduce the unintegrated seller's profits, sometimes dramatically. Finally, vertical mergers can increase total welfare. Some of the results contrast qualitatively with unambiguous findings from models with symmetric sellers, which suggests that caution should be used in drawing general inferences from those models.     

When should a firm expand its business? : The signaling implications of business expansion

We examine an incumbent's trade-off between the improved efficiency that business expansion facilitates and the signaling role that business expansion plays in conveying information to potential entrants about the state of demand. We demonstrate that both separating and pooling equilibria survive the Intuitive Criterion. Essentially, in contrast to models with asymmetric information about unit cost, incumbents' benefits from investing in a signal are not necessarily monotonic in the state of demand. We investigate how the extent of informativeness of the outcome depends on the enhanced efficiency that the incumbent's expansion facilitates and the priors of the entrant.     

An experimental test of automatic mitigation of wholesale electricity prices

In several major deregulated electricity generation markets, the market operator uses an “automatic mitigation procedure” (AMP) to attempt to suppress the exercise of market power. A leading type of AMP compares the offer price from each generation unit with a recent historical average of accepted offer prices from that same unit during periods when there was no transmission-system congestion to impede competition. If one or more units' offer prices exceed the recent historical average by more than a specified margin, and if these offer prices raise the market-clearing price by more than a specified margin, the market operator replaces the offending offer prices with lower ones. In an experiment, we test an AMP of this type. We find that it keeps market prices close to marginal cost if generation owners have low market power in uncongested periods. However, with high market power in uncongested periods, a condition that may apply in many parts of the world, the generation owners are able to gradually raise the market price well above short-run marginal cost in spite of the AMP. We also test the effect of the AMP on the frequency with which high-variable-cost units are used, inefficiently, in place of low-variable-cost units.     

The role of online platforms for media markets — Two-dimensional spatial competition in a two-sided market

We analyze the market for online and offline media in a model of two-dimensional spatial competition where media outlets sell content and advertising space. Consumer preferences are distributed along the style and type of news coverage where the distance costs may vary across dimensions. For integrated provision of online and offline platforms we show that entering the online market reduces average profits and may even constitute a prisoner's dilemma. Specialized provision may yield polarization in the style and type dimensions. This is in contrast to the maximum–minimum differentiation result previously established in the literature on multidimensional horizontal competition. We show that maximal differentiation in both dimensions occurs due to the discrete nature of the type dimension and asymmetric advertising markets.     

Do remedies affect the efficiency defense? An optimal merger-control analysis

This paper examines the optimal use of remedies and the efficiency defense for merger control. We develop a framework in which merger efficiency gains are endogenously determined and are not observed by the Competition Authority. The possibility of an efficiency defense can push firms to design the merger better, leading to greater efficiency gains. Firms can also submit remedies to the Competition Authority: they interact with the efficiency defense, by reducing the firm's incentive to enhance efficiency. At the same time remedies improve the assessment of the merger, in particular by signaling the true level of efficiency gains. We ask to what extent it may be optimal to make use of both instruments simultaneously, and thus discuss the possibility that the Competition Agency refrain from using one of the instruments to try to improve overall merger-control performance.     

A choice experiment model for beef: What US consumer responses tell us about relative preferences for food safety,country-of-origin labeling and traceability

Several food safety issues have prompted questions regarding the role of country-of-origin labeling, traceability, and food safety inspections in consumers’ perceptions of food safety and quality. The importance of origin-labeling and traceability have been discussed in the EU for some time. North American cases of mad cow disease have led to increased discussions of these topics in the US, however, relatively little research has been conducted to examine the value US consumers place on these attributes. Choice experiments were used to analyze US consumers’ relative preferences and willingness-to-pay for these meat attributes in labeled ribeye beef steaks. Relatively speaking, consumers value certification of USDA food safety inspection more than any of the other choice set attributes, including country-of-origin labeling, traceability and tenderness. As a result, indication of origin may only become a signal of enhanced quality if the source-of-origin is associated with higher food safety or quality.     

Patent office in innovation policy: Nobody's perfect

The number of patent applications and “bad” patents issued has been rising rapidly in recent years. Based on this trend, we study the overload problem within the Patent Office and its consequences on the firms' R&D incentives. We assume that the examination process of patent applications is imperfect, and that its quality is poorer under congestion. Depending on policy instruments such as submission fees and the toughness of the non-obviousness requirement, the system may result in a high-R&D equilibrium, in which firms self-select in their patent applications, or in an equilibrium with low R&D, opportunistic patent applications and the issuance of bad patents. Multiple equilibria often co-exist, which deeply undermines the effectiveness of policy instruments. We investigate the robustness of our conclusions as to how the value of patent protection is formalized, taking into consideration the introduction of a penalty system for rejected patent applications, as well as the role of commitment to a given patent protection policy.     

On the endogeneity of Cournot, Bertrand, and Stackelberg competition in oligopolies

In many industries, firms pre-order input and forward sell output prior to the actual production period. It is known that forward buying input induces a “Cournot-Stackelberg endogeneity” (both Cournot and Stackelberg outcomes may result in equilibrium) and forward selling output induces a convergence to the Bertrand solution. I analyze the generalized model where firms pre-order input and forward sell output. First, I consider oligopolists producing homogenous goods, generalize the Cournot-Stackelberg endogeneity to oligopoly, and show that it additionally includes Bertrand in the generalized model. This shows that the “mode of competition” between firms may be entirely endogenous. Second, I consider duopolies producing heterogenous goods. The set of equilibrium outcomes is characterized and shown not to contain the Bertrand solution anymore. Yet, forward sales increase welfare also in this case, notably even when goods are complements.     

Information value and externalities in reputation building

In sequential equilibrium theory, reputation building is independent of whether the reputation builder is matched with one long-run partner or a series of short-run “strangers”. We observe, however, that reputation builders are significantly more challenged by long-run players in both laboratory chain store and buyer-seller games. Reputation builder behavior is more predictable than implied by equilibrium, and so reputation information has more economic value than implied by equilibrium. For short-run players, this reputation information value is an externality. For long-run players, the value of the information is internalized and so they have greater incentive to challenge the reputation builder.     

Rent seeking in groups

Rent-seeking contests between groups involve conflicts of interests at the levels of individuals, groups, and the society as a whole. We conduct a series of economic experiments to study the behavior of groups and individuals competing against each other in rent-seeking contests. We find substantial over-contribution to rent-seeking relative to the equilibrium predictions of standard models regardless whether individuals compete against other individuals, groups compete against other groups, or individuals compete against groups.     

Price dynamics and collusion under short-run price commitments

We consider a dynamic homogeneous oligopoly in which firms set prices repeatedly. Theory predicts that short-run price commitments increase profits and may lead to less price stability. The experiments that we conducted provide support for the first effect and against the second effect when a random ending rule is applied. When a fixed ending rule is applied, we find no significant impact of short-run price commitments on profits and price stability.     

Coordination and the provision of incentives to a common regulated firm

This paper considers the problem faced by two regulators in providing incentives to a common (privately informed) regulated firm under various degrees of coordination. In the model, the firm exerts effort toward cost reduction and self-dealing, and incentives can be input-based (monitoring) and output-based (demanded cost targets). Full coordination between the regulators leads to the second best allocation. A setting in which the regulators do not fully coordinate leads to (i) higher overall monitoring (more aggressive input-based incentives) and (ii) higher demanded cost targets (i.e., more lenience in terms of output-based incentives). As a consequence of (i), in all possible equilibria, the effort toward cost reduction will be smaller when the agent reports to two regulators who do not coordinate. (i) and (ii) imply that the impact on the effort toward self-dealing activities is ambiguous. In our leading example, self-dealing will be larger if the regulators coordinate on monitoring levels but smaller if they choose monitoring levels independently.     

Optimal pricing and quality choice of a monopolist under Knightian uncertainty

This paper analyzes a simple vertical product differentiation model with demand uncertainty and derives a risk neutral monopolist's optimal market entry timing, her optimal pricing and optimal quality choice by incorporating Knightian uncertainty, irreversibility, and flexibility in quality-enhancing investment into a continuous-time stochastic model. It is shown that an increase in Knightian uncertainty induces decreases in the optimal price, the optimal quality, and the value of undertaking the quality-enhancing investment by the monopolist. The social optimal entry timing, pricing and quality are also analyzed.