Division of labor in teams

This paper examines the incentive effects of division of labor on worker effort, in the absence of the scale effects studied by Adam Smith. The game-theoretic model gives two results. (1) Suppose workers are identical and risk-neutral, and there is stochastic observation of group output by the firm offering compensations subject to some worker-participation constraint. Then the firm can arrive at the same first-best outcome with or without division of labor. However, if workers are risk-averse, division of labor can give the firm strictly greater profit. (2) A deepening division of labor magnifies this positive incentive effect; but if workers are heterogeneous, or if there ave certain informational imperfections in the production process, this incentive advantage of division of labor could be impaired or even reversed. The first result may help explain the emergence of division of labor in the early stages of industrialization without relying on the Smithian advantages, which air also present in some labor deployment schemes without division of labor. The second result throws light on some recent anecdotal evidence of a shallowing division of labor in some areas of modern manufacturing. These factors affecting the efficiency of division of labor are then further discussed in the light of recent empirical findings on division of labor and team work, such as those in Katzenbach and Smith (1993)