This paper analyzes information reporting by a privately informed expert concerned about being perceived to have accurate information. When the expert’s reputation is updated on the basis of the report as well as the realized state, the expert typically does not wish to truthfully reveal the signal observed. The incentives to deviate from truthtelling are characterized and shown to depend on the information structure. In equilibrium, experts can credibly communicate only part of their information. Our results also hold when experts have private information about their own accuracy and care about their reputation relative to others.

What are the equilibrium features of a dynamic financial market in which traders care about their reputation for ability? We modify a standard sequential trading model to include traders with career concerns. We show that this market cannot be informationally efficient: there is no equilibrium in which prices converge to the true value, even after an infinite sequence of trades. We characterize the most revealing equilibrium of this game and show that an increase in the strength of the traders’reputational concerns has a negative effect on the extent of information that can be revealed in equilibrium but a positive effect on market liquidity.

Transparency is a complex issue, which can be approached from many angles and with various conceptual tools. This contribution provides a brief survey of the economic literature on transparency. The conceptual tool that economists use is the principal-agent model, a game-theoretic setting where a principal (the citizens, shareholders, etc) wants an agent (the government, the CEO, etc.) to perform a certain task. In this setting, transparency corresponds to the ability of the principal to observe what the agent does. We can then ask what happens when transparency improves. The survey begins with the key theoretical result in this area. Holmstrom (1982) provides a powerful and general rationale for full transparency. More information about the agent’s behavior makes the agent more accountable and more likely to work for the common good. The rest of the contribution discusses a long list of objections to this result, from both a conceptual and a practical viewpoint. Given that Holmstrom asserts that transparency improves accountability, possible criticism belongs to two broad classes. First, one can argue that the increase in accountability is not sufficient to offset other drawbacks such as the violation of privacy, the direct cost of

Abstract: Organizational economics predicts that communication patterns within an organization should reflect the relative value of their members to the organization. We propose to measure the impact factor of an agent by applying the Invariant Method–also known as Google’s PageRank algorithm–to electronic communication data. To explore the validity of this measure, we analyze email exchanges among the top executives of a large retail company. We construct their individual impact factors based only on email patterns and we compare them to standard economic measures of organizational importance. We find that: (i) The impact-factor ranking of executives mirrors perfectly their hierarchical ranking; (ii) Impact factor variability is significantly correlated with salary differences; (iii) Subsequent promotions (dismissals) affect executives with unusually high (low) impact factors. We conclude that simple communication-based impact factors may be a useful tool to measure the relative importance of agents in organizations. The great mass of economic activity and much of social activity takes place not in the market but within the internal environments of organizations. As Kenneth J. Arrow (1) emphasized in his classic work on the economics of organizations, “the purpose of organizations is to exploit the fact that many (virtually all) decisions require the

Agents exert effort to improve information that guides actions. observable outcomes. Actions lead to Transparency leads agents to take particular actions and, so, distorts decisions from those which the information would advocate. If these distorted decisions reveal further information, transparency on the actions taken boosts information-gathering incentives. If the distortion is towards decisions that conceal information, transparency dampens incentives. Transparency on the agent’s information directly can also affect incentives, either positively, or negatively.

We provide a multiperiod model of political competition in which voters imperfectly observe the electoral promises made to other voters. Imperfect observability generates an incentive for candidates to offer excessive transfers even if voters are homogeneous and taxation is distortionary. Government spending is larger than in a world of perfect observability. Transfers are partly financed through government debt, and the size of the debt is higher in less transparent political systems. The model provides an explanation of fiscal churning; it also predicts that groups whose transfers are less visible receive higher transfers, and that imperfect transparency of transfers may lead to underprovision of public goods. From the policy perspective, the main novelty of our analysis is a separate evaluation of the transparency of spending and the transparency of revenues. We show that the transparency of the political system does not unambigously improve efficiency: transparency of spending is beneficial, but transparency of revenues can be counterproductive because it endogenously leads to increased

A principal chooses one of n ≥ 2 projects or an outside option. An agent is privately informed about the projects ’ benefits and shares the principal’s preferences except for not internalizing her value from the outside option. We show that strategic communication is characterized by pandering: the agent biases his recommendation toward good-looking projects—those with appealing observable attributes—even when both parties would be better off with some other project. Projects become more acceptable when pitched against a stronger slate of alternatives. We study organizational responses to the pandering distortion, such as delegation and choosing to be less informed.

When central banks are transparent about their decision making, there may be clear benefits in terms of credibility, policy effectiveness, and improved democratic accountability. While recent literature has focused on all of these advantages of transparency, in this paper we consider one potential cost: the possibility that publishing detailed records of deliberations will make members of a monetary policy committee more reluctant to offer dissenting opinions. Drawing on the recent literature on expert advisors with “career concerns”, we construct a model that compares incentives for members of a monetary policy committee to voice dissent when deliberations occur in public, and when they occur in private. We then test the implications of the model using an original dataset based on deliberations of the Federal Reserve’s Federal Open Market Committee, asking whether the FOMC’s decision in 1993 to begin releasing full transcripts of its meetings has altered incentives for participants to voice dissenting opinions. We find this to be the case with regard to both opinions on shortterm interest rates and on the “bias ” for future policy.

We would like to thank Chris Crowe at LSE and Jacob Kirkegaard of the

We study a Downsian model of elections in which two candidates who are primarily officemotivated have policy-relevant private information. A conventional intuition is that electoral competition benefits voters by inducing candidates to choose (constrained-)efficient platforms. A countering perspective is that inefficiencies obtain because candidates distort their platforms toward the voters ’ prior beliefs, i.e. they pander. We find that both intuitions are incorrect for familiar classes of information structures. In our model, office-motivated candidates have an incentive to exaggerate their private information, i.e. to anti-pander. While platforms can still reveal information, equilibrium voter welfare is limited. Our main result is that voter welfare in the Downsian game cannot be any higher than under “dictatorship ” by a single non-ideological politician; furthermore, if both candidates have a positive probability of winning in the Downsian game, voter welfare is strictly lower. Normatively, we also show that pandering would improve efficiency; while this is incompatible with office-motivated candidates, it would be an equilibrium were candidates benevolent.