| This dissertation is comprised of three chapters on compensation issues: executive compensation, cash balance pension plans, and employer provided health insurance. In the first chapter we (co-authored) examine executive compensation in light of traditional agency theoretic assumptions, but also integrate institutional and environmental arguments in an attempt to understand why CEO compensation varies across industries. We find that the effect a CEO has on organizational outcomes varies considerably across industries. We also find that some industries have considerable within-industry differences in CEO discretion on organizational outcomes, such as ROA, annual stock returns, and Tobin's Q. In industries where there is a large variance in CEO impact, total CEO compensation tends to be significantly higher. However, we do not find that industry-wide CEO discretion affects contingent pay.;In the second chapter, I endeavor to explain the likelihood that a given firm would have converted its traditional pension plan to a more controversial cash balance plan during the 1990s. Drawing upon the theory of implicit contracts, I argue that firms trying to reduce their stocks of firm specific human capital, with younger and more mobile workforces in rapidly changing industries would have been most likely to utilize such a plan. Using data primarily from the Department of Labor, the Current Population Survey, and Compustat, I find plans with more educated and older workers were more likely to convert, and plans covering unionized workers were less likely to convert. I find no support that over-funded plans were more likely to convert.;The chapter on employer provided health insurance examines the role of spousal coverage in determining labor market outcomes. Specifically, I utilize an instrumental variable model to control for the endogeneity of a husband's insurance coverage to his wife's decision to enter the labor force or sort into part-time jobs. Contrary to previous research in this area, I find that husband's coverage has a positive effect on the number of hours the wife will work. However, after netting out the likely assortative mating bias, I find that wives with spousal coverage are more likely to work part-time (and less likely to work full-time). |
