Pairwise Matching Models in Economics

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This award funds three projects in the economic theory of matching. This kind of economic theory has many applications ? such as how workers are matched to jobs, how students are sort into schools, or how borrowers are assigned to loans. In each case, a 'match' creates a productive relationship. The goal of the first project is to develop general results about how socially optimal matches change as the nature of the productive relationship changes. For example, if a technological shift disproportionately raises the productivity of highly skilled workers, how does that affect optimal worker-firm matches? The second project explores matching that occurs when there is incomplete information about one side, like students applying to colleges, and considers whether or not higher quality means a higher rejection rate. The third project considers a strategic pairwise matching theory of accidents, and asks when regulations designed to reduce risk will in fact reduce the number of accidents, after taking into account individual re-optimization (so called ?risk compensation?). This project develops methods broadly useful in many related contexts, in order to help shed light on the design of markets and policies that to enhance societal goals. The first project makes predictions for non-assortative matching, using monotone methods to produce equilibrium general comparative statics in sorting and sharing rules. The results will show how sorting, sharing rules, and inequality covary as monotone technological changes occur. The second project introduces spatial competition models into directed search. The third project develops a new strategic pairwise-matching theory model of motor vehicle (and similar) accidents, and derives conditions under which risk compensation is incomplete.