The Social Security Statement and Timing of Retirement Benefit Receipt

1. The Social Security Statement and Timing of Retirement Benefit Receipt Barbara A. Smith, Social Security Administration Kenneth A. Couch, University of Connecticut National Tax Association 2016 Annual Conference on Taxation Baltimore, MD, November 10, 2016 The views expressed are those of the authors and do not reflect those of the Social Security Administration.

2. Motivation • Surveys indicate that many Americans know little about their Social Security benefits even though these benefits represent an important component of retirement income. *More than 80% of retirement income for low-income seniors. *66% of retirement income for middle-income seniors. *44% of retirement income for upper-middle-income seniors. • Delaying claiming is one way to increase monthly Social Security benefits. • The Social Security Administration sends out the Social Security Statement, with estimates of future retirement benefits, to American workers on a regular basis. • Does the Statement increase knowledge about Social Security and influence benefit claiming behavior? If so, low-cost informational interventions could be an additional policy tool.

3. Background • This project is a continuation of research we have done on the Social Security Statement (Smith and Couch 2014a, 2014b). Earlier work described the history and implementation of the Statement and examined its impact on younger workers’ knowledge of Social Security. • Our previous research found that the Statement increased younger workers’ knowledge of Social Security. Other research has suggested that the Statement has an effect on knowledge but not on behavior. • Looking at SSA’s administrative data, we noticed a correlation between mailing of the Statement and declines in the claiming of benefits at age 62. • In the current project, we investigate the impact of the Statement on claiming behavior. Preliminary estimates suggest that receipt of the Statement significantly reduces claiming at earlier ages and increases it at older ages. And is related to delayed labor force withdrawal.

4. Statement Background and Implementation (Smith and Couch 2014a) • SSA began rolling out the Statement in FY 1995 • Fiscal Year # of Statements MailedRecipient Age Group • 7.0 million 60 or older • 5.5 million 58-60 • 12.4 million 53-58 • 20.7 million 47-53 • 26.6 million 40-47 • 134.7 million 25 or older • 135.6 million 25 or older • Mailed out annually to all eligible workers age 25 or older beginning October 1, 1999. Largest customized mailing ever undertaken by a Federal agency. • In FY 2010: • *Over 151 million Statements mailed • *Over 12.5 million mailed each month • *About 419,000 delivered each day

5. Effect of Statement on Younger Workers (Smith and Couch 2014b) Percentage of Correct Responses on Social Security Knowledge Questions among Workers Age 46 or Younger Who Did Not Receive a Statement in 1998 and Who Did Receive a Statement in 2001 1998 NS 2001 S Program Knowledge Social Security Pays Survivor Benefits 76.1 87.1* Social Security Pays Disability Benefits 78.5 85.2* Payroll Tax Finances Social Security 88.0 90.8* Benefits Knowledge Benefits Adjusted for Cost of Living 43.8 49.8* Full-Benefit Age is Rising 62.4 71.7* Retirement Benefits Depend on Earnings 81.4 93.4* SOURCE: Data from Gallup Organization surveys commissioned by SSA between 1998 and 2004. Each survey asked about SSA’s programs and benefits. NOTES: NS = No Statement received, S = Statement received; * = statistically significant at p = .01 level.

6. Current Project : Data and Initial Analysis Use data from the Continuous Work History Sample, a 1% sample of the MBR and MEF. Calculated percentage claiming at ages 62 through 70 by birth cohorts born between 1913 and 1949. Also calculated the percentage not in the labor force at those ages Estimated linear probability models for cohorts turning ages 62 through 70 in calendar years 1975 through 2007. Included dummy variables for receipt of one Statement or for two or more Statements, for year effects, and for the increase in the full retirement age (FRA).

7. Exploratory Data Analysis – Trends in Claiming at Age 62

8. Impact of Receipt of at Least One Social Security Statement on Retirement Benefit Claiming, Controlling for the Increase in Full Retirement Age (FRA)

9. Receipt of One versus Multiple Statements: Impact on Retirement Benefit Claiming, Controlling for Year Effects

10. Impact of Receipt of at Least One Social Security Statement on Labor Force Status, Controlling for the Increase in the Full Retirement Age (FRA)

11. Findings and Implications • Findings: Preliminary estimates indicate that receipt of one or more Statements resulted in statistically significant decreases of 1 to 3 percentage points in claiming at ages 62 and 63 and offsetting increases in claiming at later ages. Larger effects for males and whites. Statement receipt is also associated with statistically significant reductions in labor force withdrawals. • Implications: Our results suggest the provision of information regarding availability of expected retirement income might be an effective tool for policymakers interested in encouraging more thoughtful retirement and later life labor force decisions.

12. Previous Papers on the Social Security Statement • Barbara A. Smith and Kenneth A. Couch. 2014a. “The Social Security Statement: Background, Implementation, and Recent Developments.” Social Security Bulletin 74(2): 1-25. • Barbara A. Smith and Kenneth A. Couch. 2014b. “How Effective is the Social Security Statement ? Informing Younger Workers about Social Security.” Social Security Bulletin 74(4):1-19.