Accumulation and Decumulation in Retirement Savings Plans

National Bureau Of Economic Research, Cambridge MA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Over the last 30 years, defined contribution (DC) savings plans, such as 401(k)s, have gradually replaced defined benefit (DB) pension plans as the primary privately sponsored vehicle for providing retirement income. Employers now sponsor over 650,000 401(k) plans with more than 87 million participants and $2 trillion in assets.1 401(k) plan participants outnumber DB plan participants by 5 to 2. This continuation application proposes projects that analyze how institutional, economic, and psychological mechanisms interact to shape participation, savings, asset allocation, savings leakage, annuitization, and decumulation outcomes in defined contribution savings plans. Key mechanisms that our ongoing research has identified include financial incentives, institutional constraints, procrastination, anchoring, status quo bias, framing, salience, peer influence, heuristics, mental accounting, and bounded rationality. Our goal is to better understand these mechanisms so that institutions can be optimized to improve retirement savings outcomes. Building on our current R01, this continuation proposal has four aims that complement our previous and ongoing work. First, we plan to study the features of savings plans that affect participation, savings rates, and retirement wealth accumulation. Specifically, we will study, in the context of employer adoption of Roth 401(k) plans, the effects of changing the menu of retirement accounts and the role of savings heuristics. We will also measure the impact of numerical psychological anchors, salient thresholds, and goal-setting prompts that are included in communications to employees. Second, we plan to identify institutional features of employer-sponsored plans that affect savings crowd-out and savings leakage. We will measure the link between the automatic enrollment of unmotivated savers and resulting increases in 401(k) cash-outs that occur when workers separate from firms. We will also measure the magnitude of crowd-out across different types of savings accounts, including health savings accounts (HSA) and retirement savings accounts. Third, we plan to study institutional features of retirement plans that affect decumulation decisions. We will study the effect of changing the set of annuitization options in DB and DC accounts, the effect of making partial annuitization salient as an alternative to total annuitization, and the impact of new decision tools that make longevity risk and annuitization easier to understand. Fourth, we will expand the time horizon and number of firms in our 12-year longitudinal database on defined contribution savings plan participants. We will also create four entirely new databases on annuitization decisions in both DB and DC retirement plans. These new databases will cover several million individuals. 1 See EBRI Databook on Employee Benefits (2010).