Since the fundamental work of Walras (1874), markets have received particular attention by economists because they lead to an efficient allocation of goods and services. However, the proper functioning of markets rests on certain assumptions. For instance, the good or ser vice which is to be traded must be clearly defined. This elementary requirement is often violated in reality, in particular when services are concerned. Consider the example of railway workers who are hired to lay tracks. A labour contract which stipulates a fixed wage and defines the workers' task as "laying tracks" is rather unspecific. Workers may profit from this vagueness by reducing effort to a comfortable amount -as long as tracks are laid, they do not violate contract conditions. Thus, an im precise definition of the service can result in inefficiently low efforts. An obvious solution to this problem is a clearer definition of the ser vice, but often this way is barred: To specify, for instance, all actions which are involved in laying tracks and which may vary with weather, surface and other conditions is far too complicated and too costly. In deed, labour contracts seldom give a detailed account of the task of a worker. Alternatively to a more precise task description, the wage of the worker could be conditioned on information about the worker's performance. For example, the railway workers might be paid by the length of tracks laid so that they are motivated to exert more effort.
Effort Misallocation and the Value of Signals.- 2.1 Problems of Non-contractibility.- 2.1.1 Different Types of Non-contractibility.- 2.1.2 Mechanisms to Enforce Non-contractible Actions.- 2.1.3 Optimal Incentives in the Single Effort Model.- 2.1.4 Misallocation in the Multiple Effort Model.- 2.2 Decomposing Agency Costs.- 2.2.1 Agency Costs.- 2.2.2 Feasibility and Uncertainty Costs.- 2.2.3 Identifying Misallocation Costs.- 2.2.4 The Value of Signals.- 2.2.5 Alternative Measures of Misallocation.- 2.3 Relationship between Misallocation and Uncertainty.- 2.3.1 Uncertainty-induced Misallocation.- 2.3.2 Misallocation Without Uncertainty.- 2.3.3 Bias and Uncertainty in Statistics.- 2.4 Summary.- Firm Performance Measures.- 3.1 Multiple Agents and Free-Riding.- 3.1.1 Principal as Budget Breaker.- 3.1.2 When Costs Are Private Information.- 3.2 The Use of Benefit Signals.- 3.2.1 Lazear's Input versus Output Measure Model.- 3.2.2 A Misallocation-uncertainty Trade-off Model.- 3.3 Implications and Implementation.- 3.3.1 Benefits and Costs of Firm Performance Measures.- 3.3.2 Gauging the Intensity of Performance Measures.- 3.3.3 Proxies for Misallocation Costs.- 3.4 Summary.- Statistical Model and Empirical Evidence.- 4.1 Empirical Studies of the Value of Signals.- 4.1.1 Do Incentives Matter?.- 4.1.2 Choice of Incentive Schemes.- 4.2 The ECMOSS Data Set.- 4.2.1 Survey Design.- 4.2.2 Performance Pay Variables.- 4.2.3 Task Descriptors.- 4.2.4 Expected Relationships Between Tasks and Pay.- 4.2.5 The Bonus-pay Censoring Problem.- 4.3 Estimation Under Multivariate Censoring.- 4.3.1 Survey of Estimation Methods.- 4.3.2 Constructing an Estimator.- 4.3.3 Application to the Bonus Pay Censoring Problem.- 4.4 Results.- 4.4.1 Task Descriptors.- 4.4.2 Control Variables.- 4.5 Summary.- Conclusion.- A Properties of the Multivariate Censoring Estimator.- B Stata Programme.- B Stata Programme.- C Tables.- References.- List of Figures.- List of Tables.- List of Symbols.