Would 401(k) participants use a Social Security "bridge" option?

Abstract

Although annuities would ensure higher levels of lifetime income, reduce the likelihood that people will outlive their resources, and alleviate some of the anxiety associated with post-retirement investing, the market for annuity products is minuscule. Explanations for the low demand include the high cost of private annuities due to adverse selection, a reluctance to hand over a pile of accumulated assets for a stream of future income, and a failure to understand the value of insurance against outliving one's resources. To address these impediments, employers could increase the availability of lifetime income by adopting a Social Security "bridge" strategy within their 401(k) plans. The bridge option would use 401(k) assets to pay retirees an amount equivalent to their Social Security benefits so they can postpone claiming benefits, thereby increasing their monthly payment when they do eventually claim. This paper gauges workers' potential interest in a bridge option, using an online sample representative of the relevant population, and experimentally tests whether framing the bridge as insurance and making it a default affects the outcome. The results indicate that a substantial minority (up to about one-third) of respondents would use the bridge -- even though the concept was totally new. The experiment shows that framing increases the share of assets allocated to the bridge strategy and that defaulting workers into the strategy is even more effective. Further, the two treatments both substantially increase projected Social Security benefits. While the opt-out rate of the default is quite high, it likely reflects the ease of doing so within the experiment. The results do suggest that the default allocation to the bridge -- up to half the participant's assets -- may be too aggressive and that the opt-out would be lower under a default with a smaller share of assets devoted to the bridge.