Strike Activity and Centralisation in Wage Setting

Abstract

AbstractThe moderating effects of centralisation in wage bargaining on strike activity are considered as a stylised fact by many economists (Hoel, Moene, & Wallerstein, 1993, p. 103. Berthold & Fehn, 1996, p. 82). The theoretical rationale is based on Hicks' (1962) observation that strikes cannot occur in bargaining models based on perfect symmetric information, since rational agents can agree ex ante on the outcome resulting after a strike and thus avoid a harmful/costly strike. If strikes are caused by uncertainty and asymmetric information, centralisation of wage setting should have a moderating effect since wage setting at higher levels is based on aggregate and consequently more 'objective' data (e. g. productivity growth data provided by statistical offices) and these data are (symmetrically) available to all bargaining parties. Modern theoretical models of strikes (see e. g. Kennan & Wilson, 1989, 1990, 1993) are in line with this reasoning: Strikes serve as screening devices to extract private information (regarding profits, strike funds, willingness to strike of the work force etc.) of the opponent.