Does Framing Affect Social Security Claiming?

The decision of when to claim Social Security benefits is one of the most economically significant choices facing older Americans. Eligible individuals are entitled to claim benefits as early as age 62 but can defer claiming to as late as age 70. Monthly benefit levels are adjusted depending on claiming age-for example, an individual who stops working at age 62 but waits to claim until age 70 will receive a monthly benefit that is 76 percent higher (in real terms) than what she would have received if she had claimed at 62.

About half of workers eligible for Social Security benefits claim at age 62 and roughly two-thirds claim before age 66, the current Full Retirement Age. Does the substantial amount of early claiming represent rational, utility-maximizing behavior on the part of workers? Or is it possible that other factors, such as how information about Social Security benefits is presented, also influence workers' decisions?

This question motivates a new working paper by researchers Jeffrey Brown, Arie Kapteyn, and Olivia S. Mitchell, Framing Effects and Expected Social Security Claiming Behavior (NBER Working Paper 17018). The researchers use an experimental design to explore whether the manner in which Social Security claiming information is framed influences expected claiming behavior.

The authors first explain the "frames" that are shown to survey participants. The first frame is designed to present the information as neutrally as possible. This is similar to the approach used by the Social Security Administration (SSA) since 2008 and serves as a baseline against which other frames may be compared. The second frame emphasizes a "breakeven" concept, stressing the minimum number of years one would need to live in order for the incremental benefits resulting from delayed claiming to exceed the benefits "forfeited" by claiming later. This frame is similar to the approach used by the SSA for decades, prior to the adoption of more neutral language in 2008. It is also an approach frequently used by financial advisers.

The other frames test workers' sensitivity to framing the claiming decision in terms of consumption vs. investment, gains vs. losses, and older vs. younger reference ages. The motivation for exploring each of these dimensions comes from previous studies in economics and psychology. For example, prior studies have shown that consumers are more interested in purchasing an annuity when it is described as protecting one's ability to consume throughout life, than when it is described in terms of its investment return. Past literature has also shown that individuals are often more sensitive to losses than to gains with an equivalent value, and that "anchoring bias" affects decision-making in a wide variety of contexts.

To test the effect of these frames on expected claiming behavior, the authors fielded a survey through the RAND American Life Panel, a sample of roughly 3,000 households who are regularly interviewed over the Internet. Survey respondents were asked about their expected claiming age in one wave of the survey, and then in subsequent survey waves were presented with different frames and asked to provide their expected claiming age again in view of the new information. This approach allows the authors to test how different frames affected expected claiming behavior, controlling for any individual-specific factors (e.g., poor health) that might also affect it.

Turning to the results, the authors find that presenting individuals with the breakeven frame leads them to plan to claim 15 months earlier than they would if presented with the neutral frame - a very large effect. Framing the decision in terms of gains rather than losses or using an older anchoring age (66 or 70 vs. 62) is associated with later claiming ages, though the effects are not as large as that seen with the breakeven frame. The authors find no significant difference in framing the decision in terms of consumption vs. investment.

The authors derive two conclusions from their study. First, the results "cast doubt on a simple economic model of fully rational decision-making by showing that individual decisions are influenced by factors other than ultimate consumptions outcomes." Second, on a more practical level, the findings suggest "the manner in which information is provided to plan participants can shape behavior." The authors note that their findings are particularly relevant for an agency such as the SSA, which prides itself on providing information without offering advice and has the authority to determine how information is presented to future Social Security beneficiaries.


National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us