Essays on asymmetric information in government contracting

TR Number



Journal Title

Journal ISSN

Volume Title


Virginia Tech


The dissertation consists of a set of essays which investigate optimal contracting policies in the presence of asymmetric information and uncertainty. The first essay studies how risk aversion and a sunk investment by the firm influence the contracting outcome. The government contracts with a single, risk-averse supplier for the production of output. Both the government and the firm face uncertainty with respect to the marginal production cost of the item. Prior to full-scale production, the firm performs start-up work, during which it may make a costly investment which lowers the marginal cost of production. This cost-reducing effort is not observable by the government. At the end of the start-up phase, the firm privately learns its production cost. It then reports to the government concerning this cost, and production takes place according to the terms of the contract.

The primary result concerns the effect that the firm’s investment has on the private information problem. Specifically, the investment by the firm in the start-up phase reduces the firm’s incentive to misrepresent (overstate) its cost to the government later on. From this, it follows that the firm provides a strictly smaller investment than the government would prefer under the optimal contract.

The second essay examines the optimal incentive contract to offer to bidders with independent private values when it is costly for the principal to monitor the agent’s cost performance ex post. Cost sharing reduces the winner’s informational rents when the bidders possess heterogeneous private cost information but also discourages the agent from providing effort to reduce cost. In addition, if cost observation is costly for the principal, cost sharing gives the agent an incentive to pad his cost ex post. The essay investigates the consequences of this ex post adverse selection problem for the optimal incentive contract.

The principle results of the analysis are as follows. First, it is demonstrated that when monitoring is costly, a low cost agent will overreport his realized cost with positive probability in equilibrium. Depending upon the cost sharing parameter, the equilibrium cost reporting and monitoring strategies may either involve pooling or a mixed strategy solution.

Second, we show that the optimal contract with costly monitoring generally differs from the contract which is optimal when monitoring is costless. Depending upon the characteristics of the contracting environment, the optimal contract may induce either pooling or a mixed strategy outcome ex post. If the optimal contract involves pooling, the ‘costly monitoring’ cost sharing parameter is weakly smaller than the optimal cost sharing parameter with costless monitoring. If the optimal contract induces a mixed strategy equilibrium, the optimal level of cost sharing is strictly higher than the optimal cost sharing parameter when monitoring is costless. Finally, our model predicts that, other things equal, the level of cost sharing should be higher, the smaller the number of bidders and the more diffuse the bidder’s expected costs.