by Jeffrey C. Diebold and Susan Camilleri, North Carolina State University
Each year, the Social Security Administration issues a statement to future beneficiaries designed to assist them in understanding how the timing of their chosen claim age influences their monthly benefit level. It also includes an estimate of the survivor benefit the worker’s spouse would be entitled to in the event of the worker’s death. However, the effect of the worker’s claim age on the survivor benefit amount is not clearly explained on the statement. This lack of information is problematic, particularly since the literature suggests that the statement is the primary source of Social Security information for many workers. Furthermore, research also indicates that many workers are unaware of how their claim age affects their spouse’s benefits. Thus, the purpose of this study is to examine the effects of slight modifications of the Social Security Statement’s information regarding the survivor benefit on two outcomes: knowledge of how the primary earner’s claim age impacts the survivor benefit that his or her spouse can be expected to receive and the expected claim age of the respondent. To test the effect of statement modifications on our outcomes of interest, we will use experimental vignettes embedded within a survey. After measuring basic knowledge of Social Security benefit rules, respondents will read an identical vignette describing a hypothetical married couple. Then, participants will be randomly assigned to one of four conditions. Each group will differ with respect to their exposure to types of information they receive about the survivor benefit. Finally, participants will be asked the same set of follow-up questions including what age they think the hypothetical primary earner should claim his retirement benefit and what age they expect to claim their own retirement benefit. We will administer this survey to a sample of Social Security beneficiaries nearing retirement age drawn from the RAND American Life Panel (ALP). The ALP is a nationally representative, Internet-based survey of 5,000 working adults. In June of 2015, we plan to integrate the experimental conditions of our analysis into one of the monthly ALP surveys.