Tax Subsidies vs. Auto Enrollment: What’s Best?

A new study released this week examines a critical policy design question: what is the most effective way to increase retirement savings, including among lower-income workers? Most notably, the study finds that while tax subsidies to retirement savings result in a negligible increase in overall savings, “nudge” policies like automatic enrollment and default contributions boost retirement savings without corresponding decreases in other accounts. As noted in the N.Y. Times Economix blog, this research may have serious policy implications for tax reform in the U.S. – particularly as Congress seeks to raise revenue through lowering tax expenditures.

The study, by researchers from Harvard, the University of Copenhagen, and the Danish National Center for Economic Research, fills a gap in previous research by analyzing the actual impacts of changes to tax policy on retirement savings behavior. The requisite data is unavailable in the U.S., but in Denmark, which has a similar retirement savings system, the government collects more detailed information. Interestingly, the study finds that tax subsidies do little to increase overall savings; in Denmark, for every dollar in retirement tax breaks, total saving increased by only about a penny. This is partly attributable to the fact that only “active” savers – who are typically higher-income and account for around 15% of savers overall – are prone to adjusting their retirement savings behavior in response to incentives and changes to tax policy. This group, which typically has the most to gain from subsidies, increases their retirement contributions in response to tax breaks – but offsets these increases by reducing saving elsewhere. By contrast the remaining 85%, “passive” savers, are less responsive to tax reforms and benefit more from policies that promote default saving.

As we reported earlier this year, the current budget’s support for retirement savings disproportionately favors higher-income households, with 99% of the federal government’s spending on retirement delivered through tax subsidies that primarily benefit the rich. This new report adds a new dimension to this data, revealing that tax subsidies to retirement savings not only favor the wealthy, but also do little to increase savings overall. Consequently, the study’s authors (including “genius grant” recipient Raj Chetty) recommend that policies like auto enrollment may be a more effective method of increasing retirement savings, particularly for lower-income workers. Previous research has found that auto enrollment markedly increases participation in retirement savings plans and may also increase contributions, and the number of American employers adopting this practice is growing. The new study bolsters these findings by revealing that, unlike with tax subsidies, savers who benefit from auto enrollment aren’t reducing their contributions to other accounts, and are thus experiencing a net savings gain.

Auto enrollment has also been found to significantly diminish racial disparities in retirement savings, one manifestation of the large (and widening) racial wealth gap. In a study from earlier this year, researchers found that the racial gap in participation was practically eliminated among employers who implemented an auto enrollment policy. For example, in 2010, African Americans who were not subject to auto enrollment participated in plans at a rate of 64%, while Hispanics participated at 59%; among whites, the participation rate was much higher, at 77%. However, employers who offered auto enrollment reported both higher participation and a much narrower racial gap: their African American employees participated at 82%, Hispanics at 83%, and whites at 85%.

This is not to say that auto enrollment is a perfect solution. While the policy has been found to increase participation, contribution levels often remain lower across racial groups compared to employees who are not subject to auto-enrollment, since the default contribution rates are frequently set at as little as one or two percent. Nevertheless, both these studies suggest that with the right design, auto enrollment policies have the potential to facilitate a more equitable and effective retirement savings system, especially when paired with an effective mechanism, such as the Financial Security Credit, geared to increasing the adequacy of savings contributions from low- and moderate-income Americans.