Preferences and Productivity in Job Matching: Theory and Empirics from Internal Labor Markets

Abstract

A principal often needs to match agents together to perform coordinated tasks. However, agents can quit or slack off if they dislike their match. We study two approaches for matching agents, both widely-used in practice: Centralized assignment by firm leaders, and self-organization through market-like mechanisms. Our model connects the choice of method to the degree of specialization in a firm's workforce, the firm's production technology, worker/CEO information asymmetry, incentive alignment, and firm size. We then study these topics using data from a large organization's internal labor market. Centrally assigned matches are highly productive. Using the organization's preferred metric, the optimal match is 36% more productive than randomly assigned matches within job categories. By contrast, preference-based matches (using deferred acceptance) are only 3% better than random, but are ranked more favorably by the workforce. A key driver of results is the degree of assortative matching: The self-organized match is positively assortative, and the most productive match is negatively assortative.