Essays on Traditional IRAs and Roth IRAs: Effects on Household Saving and the Participation Choice

Abstract

This dissertation studies the effects of traditional IRAs and Roth IRAs on household saving and the participation choice between the two types of IRAs in the U.S. The first chapter summarizes the previous literature on Individual Retirement Accounts (IRAs) and household saving, which is still under debate with two competing hypotheses: the reshuffling hypothesis and the new saving hypothesis. The paper then, derives some implications for further research and suggests a theoretical extension to compare the wealth effects between traditional IRAs and Roth IRAs. The second chapter examines the effect of IRAs on household saving, focusing on the comparison of the saving effects between traditional and Roth IRAs. With the 2004, 2007, and 2010 Survey of Consumer Finances (SCF), I find evidence of reshuffling for traditional IRAs and evidence of new saving for Roth IRAs, which implies that different types of IRAs may have different effects on household saving. The third chapter investigates the participation choice between the two types of IRAs. Using the after-tax dollar approach method, I set up hypotheses on the participation choice and examine the hypotheses with the same SCF data. The result shows that the households that have more savings, discretionary income, and higher risk preferences, and are younger and more educated, are more likely to have Roth IRAs. This finding is consistent with the implication in the second chapter showing that households that can be assumed to have higher saving preferences are more likely to have Roth IRAs. Thus, Roth IRA contributions may be used for accumulating new savings without reducing existing savings.