The economics of wage fairness: Three essays examining the theoretical and empirical impact of fairness norms on the United States wage structure

In recent years, it is increasingly acknowledged among labor economists that perceptions of wage fairness affect wage outcomes, productivity, and labor turnover. The premise of the dissertation is that norms of wage fairness can only be understood by examining the institutional framework behind the wage determination process. Specifically, the dissertation's three essays focus on the role of collective bargaining in determining the set of fairness norms which prevails. In the first essay, it is argued that the collective nature of wage determination results in unions giving more emphasis to wage comparison across workers resulting in the predominance of standard pay in unionized firms. It is examined whether the wage structure in the union sector differs from that in the nonunion sector consistent with the hypothesis that the dispersion of union wages is less than the nonunion wage dispersion. Using Current Population Survey data, 1984 and 1994, the dispersions are compared by adopting the method of co-variation decomposition to control for variation in individual characteristics across worker groups. The findings show that union wages are more equal than nonunion wages, and thus supporting my hypothesis. The second essay extends the efficiency wage model of wage fairness by examining wage determination under collective bargaining. Outcomes of wage and effort are derived by applying the game theoretic concept of a "fairness equilibrium." A fairness equilibrium differs from the traditional Nash equilibrium by not only material payoffs but also the reciprocity of the outcome itself enter the objective functions due to concerns for fairness. The model shows that a Pareto superior outcome, characterized by higher wages and higher effort, is more likely to be obtained and sustained under collective bargaining, since unions increase benefits from cooperation and costs of conflict. Finally, the third essay examines the role of secular wage comparison. The essay adopts the argument made by Marshall (1887) that it is unfair for the employers and the workers to take advantage of improvement in bargaining power due to economic fluctuation. Using Current Population Survey data, 1973-81 and 1983-94, the union wage gap is estimated. Then a cointegration test is applied to test the hypothesis, that union wages are less sensitive to variation in economic conditions than nonunion wages. The empirical findings show that: (1) Union wages relative to nonunion wages were less sensitive to inflation in the first period. (2) Union wages have become sensitive to economic growth in the second period.