Academics, managers, and workers often debate the merits of employee ownership and employee involvement in firm decisions. The recent popularity of employee stock ownership plans and the appointment of union and pension fund representatives to several corporate boards has renewed interest in understanding whether employee ownership can make firms more productive and profitable. Few studies, however, have compared the performance of management-run, or ''conventional,'' firms with worker cooperatives. Existing studies usually compare the productivities of firms and workers in different industries. By examining conventional and worker-controlled firms in the same industry, Craig and Pencavel provide a better comparison. The authors analyze data on the inputs and outputs of Northwestern U.S. plywood product mills from 1968 to 1986. Their data contain observations on conventional, union, and worker co-op firms. From these data, the authors try to determine whether cooperatives can produce more output than conventional firms that use the same level of inputs. To determine this, the authors must control for differences in the scale and products of plywood mills. The authors begin by discussing how worker ownership is likely to affect the use and productivity of inputs. They also discuss factors that affect the formation and failure of cooperatives. The authors hypothesize that the availability, cost, and control of capital play key roles in the formation and continuation of cooperatives. The authors next compare output produced per worker, output produced per unit of capital, and output produced per unit of timber input. They find that cooperatives have greater productivity per unit of timber input but lower labor productivity. The cooperatives also operate at larger scales than conventional firms. Craig and Pencavel isolate the effects of scale and differences in factor usage by estimating multifactor production functions. These production functions control for differences in input usage, scale, factor and product prices, and business-cycle effects. Their estimates suggest that cooperatives have slightly better productivities than conventional firms. Output supply function estimates reveal that cooperatives have lower output price elasticities. This latter finding suggests that cooperatives smooth production more than conventional firms when demand unexpectedly changes. In other words, cooperatives are less likely to reduce inputs and output when plywood prices fall or to expand inputs and output when prices rise. Craig and Pencavel conclude that worker ownership does not have ''first-order'' consequences for the factor productivity of plywood product firms. Although they find that plywood cooperatives respond differently to demand and supply shocks, these responses do not seem to have much of an effect on firm productivity. The authors conclude by listing factors that affect the formation and survival of cooperatives in this and other industries.
