Does Poor Supervisability Undermine Teamwork? Evidence from an Unexpected
基于中国农业改革数据，检验监督困难是否导致团队生产中的搭便车问题，发现家庭承包制比集体制显著提高努力水平，但监督缺陷并非唯一原因。
A central problem with organization of team production is that, if individual effort contributions are difficult to observe, feasible schemes for distributing net product among participants may suffer from poor work incentives. Thus, Armen Alchian and Harold Demsetz (1972) argue that efficiency commonly requires that there be an agent who monitors and disciplines other team members in exchange for right to retain team residual earnings, while Bengt Holmstrom (1982) shows that monitoring may in some cases be dispensed with, provided that there is a budget-breaking principal to perform an analogous disciplinary function. These arguments imply that production in residual-sharing cooperative teams will be beset by free-riding problems occasioned by inadequate reward discrimination due to absence either of monitoring incentives or of an agent with effective disciplinary powers. Such disadvantages are thought to be multiplied in agricultural cooperatives, because farm production is considered especially difficult to monitor (Michael Bradley and M. Gardner Clark, 1972; Joseph Stiglitz, 1974; Hans Binswanger and Mark Rosenzweig, 1986) and because scale economies, especially in labor-intensive systems, appear to be negligible (Putterman, 1989). The fact that restoration of householdbased farming in China during early 1980's led to a pronounced increase in output and productivity has been viewed as evidence in support of this thesis. To be sure, degree to which gains recorded during 1978-1984 reflected changes in farm-level organization as opposed to being a result of policy changes in such areas as crop specialization, pricing, and opening of free markets, remains a matter of some uncertainty (Dwight Perkins, 1988). A more subtle question, however, is that of extent to which any gains due to micro organizational change should be attributed to supervision problems of type mentioned above. A recent contribution by John McMillan, John Whalley, and Lijing Zhu (1989; hereafter MWZ), which attempts to answer first of these questions, gives rise to surprising implications regarding second. MWZ combine a simple model of incentives and utility maximization with standard growth-accounting techniques to reach conclusion that 78 percent of increase in agricultural productivity in China during period in question can be attributed to nonprice factors, principally the incentive effects of new responsibility (p. 782). Their model and estimates also imply that Chinese peasants supplied only a little over half (56 percent) as much effort under communal system as under succeeding household-based farming system and that this is attributable to fact that peasants expected to receive only about 30 percent of value product generated by incremental effort. Questions might be raised regarding study's methodology and, hence, numbers derived. The aim of present note, however, is to point out that, if findings of MWZ are to be believed, then although China's collective farms did indeed suffer from inadequate incentives, it may be wrong to attribute much of this to deficiencies in supervision. The MWZ model is not sufficiently microanalytic to distinguish between monitoring problems and other possible dampeners *Department of Economics, Brown University, Providence, RI 02912. Support from National Science Foundation Grant No. SES-8721382, comments and computational assistance from Si Joong Kim, and helpful suggestions of two referees are gratefully acknowledged.