Does the wage structure depend on the wage contract? A study of public sector wage contracts in Iceland

A Dissertation Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfillment of the Requirements for the Degree of Doctor of PhilosophyIt is widely accepted within the field of labor economics that centralization of collective bargaining leads to lower wage dispersion. But is it possible to change the wage structure through changes in the collective bargaining agreement and decentralization of the bargaining process? A unique opportunity to explore this question presented itself when changes were made to collective bargaining contracts in the public sector in Iceland. In the first chapter I look at the Icelandic labor market. The Icelandic labor force is often described as being flexible. But is it really? Using definitions of labor-market flexibility, I explore whether the Icelandic labor market can be classified as such and find that on most measures of flexibility the Icelandic labor market can be described as flexible. In the second chapter I explore the effects of the changes in the bargaining structure and decentralization in the public sector in Iceland on the wage structure. Did wage dispersion increase with decentralization, as theory would predict? I find that wage levels rose significantly, and that the wage structure for total wages did not change but that the dispersion of daytime wages increased. In the third and final chapter I develop a model of collective bargaining as a two-stage process in the manner of Manning (1987). The resulting two-equation nonlinear structural model is then applied to the central government in Iceland in order to determine whether the collective bargaining structure changed along with the changes in the collective bargaining agreements. The decentralization of bargaining and the change in the collective bargaining agreements has changed the bargaining structure in the public sector in Iceland. I find that the unions have a greater bargaining power over employment than over wages, whereas their bargaining power over wages currently seems to be much greater for daytime wages than for total wages. Based on the means of the estimates of the bargaining power of unions over wages and employment, respectively, I can reject the monopoly union and right-to-manage bargaining models