How can we know in advance whether simplifying assumptions about beliefs will make a difference in the conclusions of game-theoretic models? We define critical types to be types whose rationalizable correspondence is sensitive to assumptions about arbitrarily high-order beliefs. We show that a type is critical if and only if it exhibits common belief in some non-trivial event. We use this characterization to show that all types in commonly used type spaces are critical. On the other hand, we show that regular types (types that are not critical) are generic, although perhaps inconvenient to use in applications.

We derive a closed-form solution for Tobin's Q in a stochastic dynamic framework. We show analytically that investment is positively related to Tobin's Q and cash flow, even in the absence of adjustment costs or financing frictions. Both Q and investment move in the same direction as expected revenue growth, so changes in expected revenue growth induce Q and investment to comove positively. Similarly, shocks to current cash flow, arising from shocks to the user cost of capital in our model, cause investment and cash flow per unit of capital to comove positively. Furthermore, we show that this alternative mechanism for the relationship among investment, Q, and cash flow delivers larger cash flow effects for smaller- and faster-growing firms, as observed in the data. Moreover, the empirically small sensitivity of investment to Tobin's Q does not imply implausibly large adjustment costs in our model (since there are no adjustment costs). Calibrating the model generates values of Q similar to those in the data; investment is more sensitive to cash flow than it is to Q, and both responses are of empirically plausible magnitudes.

The paper studies the effects and the determinants of interregional redistribution in a model of residential and political choice. We find that paradoxical consequences of interjurisdictional transfers arise if people are mobile: while self-sufficient regions are necessarily identical with respect to policies and average incomes in our model, interregional redistribution always leads to the divergence of regional policies and per capita incomes. Thus, interregional redistribution prevents inter regional equality. At the same time, however, transfers may allow for more inter personal equality among the inhabitants of each region. The voting population may therefore in a decision over the fiscal constitution deliberately implement such a transfer scheme to foster regional divergence. Empirical evidence from panel data from OECD countries and Canadian provinces is consistent with the theory.

The incentive and project selection effects of agent anonymity are investigated in a setting where an evaluator observes a subjective signal of project quality. Although the evaluator cannot commit ex ante to an acceptance criterion, she decides up front between informed review , where the agent's ability is directly observable, or blind review , where it is not. An ideal acceptance criterion balances the goals of incentive provision and project selection. Relative to this, informed review results in an excessively steep equilibrium acceptance policy: the standard applied to low-ability agents is too stringent and the standard applied to high-ability agents is too lenient. Blind review, in which all types face the same standard, often provides better incentives, but it ignores valuable information for selecting projects. The evaluator prefers a policy of blind (respectively informed) review when the ability distribution puts more weight on high (respectively low) types, the agent's pay-off from acceptance is high (respectively low), or the quality signal is precise (respectively imprecise). Applications discussed include the admissibility of character evidence in criminal trials and academic refereeing.

We develop a theory of media slant as a systematic filtering of political news that reduces multidimensional politics to the one-dimensional space perceived by voters. Economic and political choices are interdependent in our theory: expected electoral results influence economic choices, and economic choices in turn influence voting behaviour. In a two-candidate election, we show that media favouring the front-runner will focus on issues unlikely to deliver a surprise, while media favouring the underdog will gamble for resurrection. We characterize the socially optimal slant and show that it coincides with the one favoured by the underdog under a variety of circumstances. Balanced media, giving each issue equal coverage, may be worse for voters than partisan media.

This paper investigates the relationship between media bias and the influence of the media on voting in the context of newspaper endorsements. We first develop a simple econometric model in which voters choose candidates under uncertainty and rely on endorsements from better informed sources. Newspapers are potentially biased in favour of one of the candidates and voters thus rationally account for the credibility of any endorsements. Our primary empirical finding is that endorsements are influential in the sense that voters are more likely to support the recommended candidate after publication of the endorsement. The degree of this influence, however, depends upon the credibility of the endorsement. In this way, endorsements for the Democratic candidate from left-leaning newspapers are less influential than are endorsements from neutral or right-leaning newspapers and likewise for endorsements for the Republican. We also find that endorsements are more influential among moderate voters and those more likely to be exposed to the endorsement. In sum, these findings suggest that voters do rely on the media for information during campaigns but that the extent of this reliance depends upon the degree and direction of any bias.

We offer an explanation for the forward premium puzzle in foreign exchange markets based upon investor overconfidence. In the model, overconfident individuals overreact to their information about future inflation, which causes greater overshooting in the forward rate than in the spot rate. Thus, when agents observe a signal of higher future inflation, the consequent rise in the forward premium predicts a subsequent downward correction of the spot rate. The model can explain the magnitude of the forward premium bias and several other stylized facts related to the joint behaviour of forward and spot exchange rates. Our approach is also consistent with the availability of profitable carry trade strategies.