Content area
Abstract
This dissertation is divided into two unrelated sections. The first section is a collection of five essays which addresses issues in the theory and practice of collective bargaining and collective bargaining institutions. The first essay describes the recently passed New Jersey arbitration law and presents a descriptive analysis of the parties' reliance on the arbitration procedures in the first year after the law took effect. The second essay analyzes first-year salary settlements under the New Jersey arbitration law, and concludes that arbitrated settlements were not significantly different from nonarbitrated settlements. The results also contain evidence of asymmetric behavior by the parties under final-offer arbitration with the unions appearing more risk averse than the employers. The third essay presents and analyzes a model of collective bargaining under arbitration and derives the conditions under which arbitration provides bargainers with incentives to settle voluntarily. This essay also demonstrates the importance of modelling collective bargaining and arbitration processes, and not just outcomes. The fourth essay presents a simple model of arbitrator decision-making under final-offer arbitration and tests this model using data on New Jersey's first two years of experience with final-offer arbitration. The empirical results are supportive of the model developed and they provide important evidence on arbitrator preferences and on the factors arbitrators consider in rendering an award. The fifth essay presents an analysis of Hicks' theory of collective bargaining, deriving the model from the first principles of microeconomic theory, investigating its properties, and showing that much of the criticism to which the model has historically been subjected is misguided.
The second section of this dissertation investigates the age patterns of women at first birth using the marriage model developed primarily by Ansley J. Coale. Briefly, the Coale formulation is a three parameter model which, in its application to first birth data, relates an observed distribution of first birth frequencies to a standard Swedish schedule of first marriage frequencies. There are good theoretical reasons for believing the marriage model will provide a good fit to age distributions of first birth frequencies and the results of this study confirm that belief through detailed analyses of Finnish and Italian cohort and period first birth data. In addition, this study investigates several applications of this finding. The first application involves using the model in descriptive analyses of first birth fertility. This is an important application since the model summarizes and entire first birth distribution in terms of three easily interpretable parameters. The second application involves exploiting the parametric nature of the model by using the fit of the model to incomplete first birth data to forecast the unobserved portion of the first birth distribution. After investigating the conditions under which the model is likely to be useful in these applications, the model is fit to first birth data for recent cohorts of women in eleven national populations and for recent cohorts of white and nonwhite women in the United States. The estimated parameters are analyzed in terms of their implications for the completed first birth fertility of these cohorts.
