Unionism, Oligopoly and Rigid Wages
利用1947-1976年美国制造业截面数据，估计工会化和市场集中度对平均工资水平的影响，并分析通胀率和失业率如何改变这些影响，检验工资刚性假说。
T HE proposition that imperfections in product and labor markets lead to wage rigidity has been stressed by many authors in their analysis of wage inflation.' The most commonly held view is that wage differentials which result from these imperfections behave countercyclically. That is, periods of rapid inflation and/or low unemployment are believed to be associated with reduction of the wage premium attributed to both unionization and oligopolistic structure. On the labor market side it is argued that wage rigidity is primarily due to lags introduced by collective bargaining.7 In periods of low demand union wage levels are maintained through long term agreements. On the upswing these same long term agreements do not allow the union wage to react as quickly as the nonunion wage to increases in demand. Cross-sectional estimates of economy-wide union/nonunion differentials are often used to support this hypothesis.3 On the product market side it is often argued that concentrated industries have the ability to raise prices in periods of economic slack. This market power is a permissive factor which may allow workers in these industries to demand a share of the excess profits.4 Of course, the work force must have some method of demanding this share. It is usually assumed that union power is relatively great in these concentrated industries. Thus, this argument is really conditioned on imperfections in both the labor and product market, and is therefore much closer to an interaction hypothesis than a pure oligopolistic structure hypothesis. Proponents of this argument further hypothesize that concentrated industries do not increase prices as fast as competitive industries in an economic upturn. Wage differentials between concentrated and unconcentrated industries are therefore expected to narrow in periods of rapidly increasing prices. With one exception,5 the support for this hypothesis comes from studies of wage changes in concentrated versus unconcentrated industries.6 This paper is a reexamination of the wage rigidity hypotheses. The impacts of unionization and concentration on average wage levels are first estimated for a series of cross-sectional observations in manufacturing between 1947 and 1976. The differences in these cross-sectional impacts across time are then analyzed using rates of inflation and unemployment levels as explanatory variables.