Essays in the Role of Overseeing Entities in Retirement Plans

Abstract

This dissertation is comprised of three essays that focus on the role of overseeing entities in retirement plans. In the first essay, I study the role of trustee and non-trustee service providers as well as the composition of a firm's board of directors in overseeing 401(k) plans. I ask whether differences in the number and type of these 401(k) plan overseeing entities can explain differences in 401(k) plan performance and structural characteristics. Using a proprietary dataset of 401(k) plans, I find that having more trustee and non-trustee service providers results in better menu performance. However, these findings are not robust when benchmark adjusting performance. Second, I find that having more non-trustee service providers leads to less menu diversification and higher fund level expenses, but lower total plan expense. Last, having more trustee service providers and a greater percentage of insiders on a firm's board of directors results in a more generous company match. My results suggest that 401(k) plans are significantly impacted by oversight decisions, and that improving oversight quality may be a more effective way to mitigate 401(k) plan losses than focusing on increasing financial literacy of plan participants. In the second essay, I examine the nature of compensation for 401(k) plan consultants and ask whether variations in the form of compensation explain variations in 401(k) plan costs and menu performance. Using a proprietary dataset of 401(k) plans, I find that 401(k) plans which hire a consultant experience lower fund level fees and higher after-fee returns if the consultant does not participate in revenue sharing arrangements. In exchange for their services to improve plans, consultants without revenue sharing arrangements charge higher fees to offset their revenue losses from not having opaque arrangements. This results in higher administrative expenses for plans. The net effect is a 9.6 basis point annual gain for the average plan participant or a 24.7 basis point annual gain for a plan participant invested in the default menu choice, assuming that employees pay the higher administrative expense. My findings are robust to a narrower definition of a consultant, additional controls for investment expertise, retirement expertise and bargaining power, falsification tests, and propensity score matching. Overall, my findings suggest that 401(k) plan menu design may be improved through the use of a consultant if the consultant does not suffer from conflicts of interest. In the third essay, I empirically test whether governance mitigates underfunding in US public pension plans. Traditional governance proxies in public sector defined benefit plans focus on plan board of directors. However, plan responsibilities extend beyond the board and are addressed by state or plan policies and by other entities involved in pension oversight. Using unique governance survey data for US public pension plans, I measure governance in an agency theory framework and in a theoretical best practices framework. In the first framework, governance proxies include state and plan policies while in the second, governance proxies include the distribution of oversight responsibilities. I find that the most important governance policies are those that encourage sponsor commitment to paying required annual contributions. I also find that theoretical best practices do not mitigate plan underfunding.