Designing Benchmarking and Auction Mechanisms

This thesis considers different designs of benchmarking and auction mechanisms. The two mechanisms are widely used for solving different informational problems. Unlike auctions, the stress in benchmarking is often more on coordination and learning and less on motivation. Besides treating the two mechanisms individually, the thesis aims at combining the two. The thesis is divided into 3 parts, the first part provides an overview of the thesis. The second part consists of 2 papers presenting an interactive benchmarking tool. The third part consists of 4 papers treating different aspects of incentive provision and auction design. In this part a multi-dimensional auction is suggested that combines benchmarking techniques used in part II with the so-called multi-dimensional score auction. Part II: Benchmarking Mechanisms Benchmarking as a management tool adds value by providing information on best practice and relevant peers. Part II focuses on how to tailor a benchmarking system to the individual user and thereby enhancing the value of benchmarking. A design for interactive benchmarking is presented, which eases the matching of the user’s preferences and information about best practice. Best practice is determined using either parametric or non-parametric estimation methods. The user is given two ways of expressing his preferences 1) selecting different subsets of the set of references and 2) determining a direction in which to find more preferred practices. The user may apply the two instruments in various ways. The first paper is a general paper that introduces the idea and presents two applications of Internet based benchmarking: A simple parametric model describing 50 different Danish industries and a non-parametric model analyzing 189 Danish banks. The second paper extends the idea to fully cover the parametric approach, in an up-to-date stochastic frontier model. An efficiency study of Danish dairy farmers is implemented in an Internet based benchmarking tool. Part III: Auction Mechanisms This part focuses on the use of auctions in regulating industries and on improving allocations between economic agents. Besides considering auctions in general, a new auction design that utilizes the benchmarking technique applied in part II is presented. The smallest change in the set of trading rules that make an auction, may change the optimal bidding strategy drastically. In the third paper, we study a trading rule that prevents bidders