John McMillan

Bidding for Contracts: A Principal-Agent Analysis

This article models the process of bidding for government contracts in the presence of moral hazard. Several (possibly risk-averse) potential contractors (agents) submit sealed bids, on the basis of which the government (principal) selects one to perform a task. The optimal linear contract is derived. The bidding process induces the potential agents to reveal their relative expected costs. The optimal contract trades off giving the chosen agent an incentive to limit costs against stimulating bidding competition and sharing risks. The optimal contract is never cost-plus, may be fixed-price, but is usually an incentive contract. Some prescriptions for government contracting emerge.

Auctions with a stochastic number of bidders

Abstract Auction theory is generalized by allowing the number of bidders to be stochastic. In a first-price sealed-bid auction with bidders having constant absolute risk aversion, the expected selling price is higher when the bidders do not know how many other bidders there are than when they do know this. Thus the seller should conceal the number of bidders if he can. Moreover, a bidders ex ante expected utility is the same whether or not there is a policy of concealing the number of bidders: concealment therefore Pareto-dominates announcement. With risk-neutral bidders, the optimal auction is the same whether or not the bidders know who their competitors are.

Auctions with entry

Abstract We consider an auction which bidders can enter upon paying an entry cost. Unlike the case of a fixed number of bidders, the seller should not impose a reserve price higher than his own valuation. When a first-price sealed-bid auction is used, the optimal number of bidders enter.

Competition for Agency Contracts

This article introduces a market for the services of agents into a principal-agent model. The principal and the potential agents are risk neutral. The contract trades off adverse selection against moral hazard. In a broad range of circumstances the optimal contract is linear in the outcome. In an incentive-compatible contract the more able is an agent, the larger is his contractual share of his marginal output; thus, a more able agent is induced to work at a rate closer to the first-best.

OPTIMAL CONTRACTS FOR TEAMS

In a team subject to both adverse selection (each members ability is known only to himself) and moral hazard (effort cannot be observed), optimal contracts are, under certain conditions, linear in the teams output. The outcome is the same whether the principal observes just the total output or each individuals contribution. Thus, monitoring is not needed to prevent shirking by team members; instead, the role of monitoring is to discipline the monitor. Copyright 1991 by Economics Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research Association.

Individual incentives in the supply of public inputs

Abstract The free-rider hypothesis states that, in the decision on public-goods and private-good provision, individual incentives are such that public goods will tend to be undersupplied. This paper examines the free-rider argument as it applies to public intermediate goods. It is shown that, unlike in a static model, in a dynamic world there may exist incentives for firms to act cooperatively in determining the supply of public intermediate goods. In a dynamic context there is a cost to free riding: what one firm does now affects what others will do in the future. Provided future profits are not discounted too heavily, the free-rider problem may disappear when a time dimension is added to the theory.

Price dispersion, price flexibility, and repeated purchasing

This paper presents a search model in which consumers repeatedly purchase some good. The repetition of purchases leads to the formatio n of implicit contracts with sellers that generate a kink in the demand curve facing each seller. These kinks can generate equilibrium price dispersion under conditions where single-purchase search models predict no price dispersion. Despite the kinks, prices are at least partially flexible in that some, and possibly all, sellers change their prices in response to changes in production costs or demand.

The scale effect of public goods on production-possibility sets

Publisher Summary This chapter discusses the shape of the set of production possibilities in the presence of public goods that are inputs to production processes. Examples of such public intermediate goods are the production infrastructure—transport and communications services, irrigation, and flood control, pollution control, and technological knowledge. However, in an economy with non-convexities, existence and Pareto-optimality of competitive equilibrium cannot be assured. The free-rider problem means that in an economy with public goods, Pareto optima may not be attainable. When the public goods are intermediate goods rather than consumption goods, this reason for misallocation of resources loses much of its force because the problem of obtaining the information needed to compute how much public-good supply is optimal and is of lower order of difficulty. If, however, the public intermediate goods give rise to non-convexities in the production-possibility set, then there is a second source of misallocation of resources, separate from the free-rider problem. With non-convexities, even if the public intermediate good is optimally supplied by some mechanism, rational action by individual agents may not lead to an efficient outcome.