Local and Central Wage Bargaining

Abstract

The literature on wage bargaining usually considers either central or local wage bargaining; an exception is Holmlund (1986). Throughout most of the unionized part of the private sector in Norway, wage bargaining takes place on two levels, first centrally and then locally. In the economic policy debate, it has been argued that this system leads to higher wages and lower employment than bargaining on only one level, regardless of whether it is central or local; see e.g. Skatnland (1983). In this paper, I present a model with wage bargaining on two levels and show that, in this system, the above argument is proved false. We focus on an industry with n identical firms which produce one homogeneous product. For simplicity, the output price and input prices other than wages are considered given. On the central level, there is a utility-maximizing labor union with a quasi-concave utility function increasing in the arguments employment (L) and total wage ( VV). There is a local labor union in each firm. We analyze the system during one period, say a year. Wage setting in this period consists of three stages. In the first stage the central labor union sets the tariff wage q unilaterally. This follows the monopoly model of e.g. Dunlop (1944) and Oswald (1985). In the second stage the firms determine the employment L which prevails until the next period. In the third and final stage the local union negotiates with the management of the firm about local wage drift. The total wage is W= q + wage drift. The assumption that the firms determine employment prior to wage bargaining is used by e.g. Moene (1987) and Horn & Wolinsky (1985). Both the central labor union and the firms know how local wage bargaining functions and can therefore perfectly anticipate what the result-