Topics in Auctions and Contractual Arrangements in Procurement

Stanford University, Stanford CA

Investigators

Abstract

The goals of this project are twofold. First, a more realistic framework for computa-tion, identification and estimation of bidding models that can be applied to studying the workings of the public sector procurement system in the United States will be developed. A framework that allows for bidders to be heterogeneous in ways that are understood before the auction begins will be used. Bidder heterogeneity is a key empirical problem in many real world applications. For example, in the construction of public works, firms often differ by location, the type of work that they specialize in, their current backlog and their managerial efficiency. As a result, firms will have cost advantages or disadvantages on certain projects that are understood by all agents before bidding begins. A second important source of asymmetries is collusion. A cartel will have a different distribution of costs than a competitive fringe and this fact may help to empirically identify between collusive and competitive bidding. Extensions of this project will include endogenizing the participation decision of firms since endogenizing participation is clearly needed for modeling realism. Also, since one form of collusion is to choose not to bid against another cartel member, modeling participation is needed to understand the difference between competitive and collusive bidding patterns. Another extension is to use the result that the empirical distribution of bids is exchangeable in the underlying model parameters only if the firms are competitive to empirically to distinguish between competition and collusion. Finally, a model of dynamic bidding in procurement will be built. This is an important empirical consideration, since in many applications, firms will be capacity constrained. A model of the costs and benefits of alternative contractual arrangements that are commonly used in procurement will also be developed. In the engineering and construction management literature on the costs and benefits of cost-plus and fixed price contracting in building construction, the flexibility of the contractual form to adjust to changed conditions is a central concern. The centerpiece of the model is a trade-off between providing cost-reducing incentives, and saving on transaction costs of two kinds. The first cost is that of providing a comprehensive design that describes the project. The second cost is due to costly renegotiation of the original design. The model is able to replicate some basic stylized facts from the private sector of the construction industry. When total completion time is to be minimized cost plus contracting is preferred and planning before contracting begins tends to be less complete. If minimizing total project costs is of the essence, then more detailed plans are provided and fixed price contracting is favored. Also, more complex projects are more often governed by cost-plus contracts.