Knowledge that Transforms

We characterize the optimal editorial positions of the media in a model in which the media influence both voting behaviour and party policies. Political parties are less likely to choose partisan policies when more voters consume informative news. When there are two media outlets, each should be slightly biased relative to its audience in order to attract voters with relatively extreme views. Voter welfare is typically higher under a duopoly than under a monopoly. Two media outlets under joint ownership may provide more diverse viewpoints than two independent ones, but voter welfare is not always higher. © 2008 The Review of Economic Studies Limited.

We study an infinitely repeated first-price auction with common values. We focus on one-sided incomplete information, in which one bidder learns the objects' value, which itself does not change over time. Learning by the uninformed bidder occurs only through observation of the bids. The proprietary information is eventually revealed, and the seller extracts essentially the entire rent (for large discount factors). Both players' pay-offs tend to 0 as the discount factor tends to 1. However, the uninformed bidder does relatively better than the informed bidder. We discuss the case of two-sided incomplete information and argue that, under a Markovian refinement, the outcome is pooling as information is revealed only insofar as it does not affect prices.

In their paper "Information Acquisition in Financial Markets" (this journal, 2000), Barlevy and Veronesi present a model of a one-period financial market, and claim that for an open set of parameter values, the value of information increases with the mass of informed agents. That claim is shown here to be false. The property of strategic substitution is robust in their model.

This paper considers the estimation problem in dynamic games with finite actions. we derive the equation system that characterizes the markovian equilibria. the equilibrium equation system enables us to characterize conditions for identification. we consider a class of asymptotic least squares estimators defined by the equilibrium conditions. this class provides a unified framework for a number of well-known estimators including those by Hotz and Miller (1993) and by Aguirregabiria and Mira (2002). We show that these estimators differ in the weight they assign to individual equilibrium conditions. We derive the efficient weight matrix. A Monte Carlo study illustrates the small sample performance and computational feasibility of alternative estimators.

In criminal organizations, diffusing information widely throughout the organization might lead to greater internal efficiency (in particular, since these organizations are self-sustaining, through enhancing trust). However, this may come at the cost of leaving the organization more vulnerable to external threats such as law enforcement. We consider the implications of this trade-off and characterize the optimal information structure, rationalizing both “hierarchical” structures and organization in cells. Then, we focus on the role of the external authority, characterize optimal detection strategies, and discuss the implications of different forms of enforcement on the internal structure of the organization and policy. Finally, we discuss a number of applications and extensions.

Commitment is typically modelled by assigning to one of the players the ability to take an initial binding action. The weakness of this approach is that the fundamental question of who has the opportunity to commit cannot be addressed, as it is assumed. This paper presents a framework in which commitment power arises endogenously from the fundamentals of the model. We construct a finite dynamic game in which players are given the option to change their minds as often as they wish, but pay a switching cost if they do so. We show that for games with two players and two actions there is a unique subgame-perfect equilibrium with a simple structure. This equilibrium is independent of the order and timing of moves and is robust to other protocol specifications. Moreover, despite the perfect information nature of the model and the costly switches, strategic delays may arise in equilibrium. The flexibility of the model allows us to apply it to various environments. In particular, we study an entry deterrence situation. Its equilibrium is intuitive and illustrative of how commitment power is endogenously determined.

To remain competitive, an organization must both respond to information about its environment and coordinate its activities. We analyse how the allocation of decision rights within an organizational hierarchy influences the organization's ability to solve such problems of coordinated adaptation when information is both soft and distributed inside the organization and the organizational participants behave strategically. The results show that, contrary to the common intuition, the performance differential between centralized and decentralized decision-making is non-monotone in the importance of coordination. Further, both these common structures are dominated by asymmetric structures in sufficiently asymmetric environments (such as a small division developing a new product in the presence of a large division with an established product). Finally, if the incentive conflicts between the participants can be made sufficiently small, centralized decision-making is always dominated by decentralized decision-making.

Do parents have preferences over the gender of their children, and if so, does this have negative consequences for daughters versus sons? In this paper, we show that child gender affects the marital status, family structure, and fertility of a significant number of American families. Overall, a first-born daughter is significantly less likely to be living with her father compared to a first-born son. Three factors are important in explaining this gap. First, women with first-born daughters are less likely to marry. Strikingly, we also find evidence that the gender of a child in utero affects shotgun marriages. Among women who have taken an ultrasound test during pregnancy, mothers who have a girl are less likely to be married at delivery than those who have a boy. Second, parents who have first-born girls are significantly more likely to be divorced. Third, after a divorce, fathers are much more likely to obtain custody of sons compared to daughters. These three factors have serious negative income and educational consequences for affected children. What explains these findings? In the last part of the paper, we turn to the relationship between child gender and fertility to help sort out parental gender bias from competing explanations for our findings. We show that the number of children is significantly higher in families with a first-born girl. Our estimates indicate that first-born daughters caused approximately 5500 more births per year, for a total of 220,000 more births over the past 40 years. Taken individually, each piece of empirical evidence is not sufficient to establish the existence of parental gender bias. But taken together, the weight of the evidence supports the notion that parents in the U.S. favour boys over girls.

The Coase conjecture (1972) is the proposition that a durable-goods monopolist, who sells over time and can quickly reduce prices as sales are made, will price at marginal cost. We show that an arbitrarily small deviation from Coase's assumptions—a deviation that applies in almost any practical application—results in the failure of that conjecture. In particular, we examine that conjecture in a model where there is a vanishingly small cost for production (or sales) capacity, and the seller may augment capacity in every period. In the “gap case”, any positive capacity cost ensures that in the limit, as the size of the gap and the time between sales periods shrink, the monopolist obtains profits identical to those that would prevail when she could commit ex ante to a fixed capacity. Those profits are at least 29.8% of the full static monopoly optimum.

We investigate identification in semi-parametric binary regression models, y = 1(xβ+υ+ε > 0) when υ is either discrete or measured within intervals. The error term ε is assumed to be uncorrelated with a set of instruments z, ε is independent of υ conditionally on x and z, and the support of −(xβ + ε) is finite. We provide a sharp characterization of the set of observationally equivalent parameters β. When there are as many instruments z as variables x, the bounds of the identified intervals of the different scalar components βk of parameter β can be expressed as simple moments of the data. Also, in the case of interval data, we show that additional information on the distribution of υ within intervals shrinks the identified set. Specifically, the closer the conditional distribution of υ given z is to uniformity, the smaller is the identified set. Point identified is achieved if and only if υ is uniform within intervals.