Absence of conflict of interest.
Citation
Highlights
-
The study’s objective was to examine the impact of the 1983 Social Security reforms on employment status, hours worked, and receipt of disability benefits.
-
The author used a difference-in-differences design to estimate the impact of the Social Security reforms, which gradually increased the normal retirement age (NRA) from 65 to 67 and also gradually increased the penalty for claiming benefits earlier than the NRA. The author used data from 1993-2015 from the Current Population Survey (CPS) to examine employment, retirement status, and receipt of disability benefits as different birth cohorts neared retirement age.
-
The study suggested that the Social Security reforms may have had opposite effects for men ages 55-60 versus those 61-65. Specifically, the reforms may have increased the retirement rate for men ages 55-60 but decreased it for men ages 61-65, and decreased the share of men ages 55-60 who worked part-time while increasing this share for men ages 61-65. The study also suggested that the reforms may have increased the average hours worked per week for men ages 61-65 as well as the share of men ages 66-70 who identified as unemployed.
-
The quality of causal evidence presented in this report is low because the author could not ensure that the groups being compared were similar on the outcome measures before the reforms. This means we are not confident that the estimated effects are attributable to the Social Security reforms; other factors are likely to have contributed.
Intervention Examined
The 1983 Social Security Reform
Features of the Intervention
The Social Security Amendments of 1983, signed into law in April of that year, increased the normal retirement age (NRA) from 65 to 67 while also increasing the penalty for claiming benefits early. Workers born in 1937 or earlier were not affected by the reforms. For workers born between 1938 and 1942, the reforms increased the NRA in two-month increments (e.g., for those born in 1938, the NRA was set at 65 years and two months). For those born between 1943 and 1954, the NRA was set at 66 years old, and, then for those born after 1954, the NRA again increased in two-month increments until reaching 67 for people born in 1960 or later. Similarly, the penalty for claiming benefits early gradually increased from a 20 percent penalty for those born in 1937 or earlier to a 30 percent penalty for those born in 1960 or later.
This study examines outcomes for four birth cohorts, 1933-1937, 1938-1942, 1943-1947, and 1948-1952, examining outcomes for men in the United States as they near retirement age.
Features of the Study
The author used a difference-in-differences model to compare outcomes for those unaffected by the reforms (1933-1937 birth cohort) to those affected by the reforms (1938-1942 birth cohort), and for those in the “flat portion” of the reforms (two birth cohorts: 1943-1947 and 1948-1952) for whom the retirement age was set at 66. This approach allows the author to estimate the policy effect of the reforms while separating out possible birth cohort effects. The study’s models include controls for age, race, education, marital status, census region, the state-level unemployment rate, and more.
The study uses 1993 – 2015 Current Population Survey (CPS) data. CPS is a monthly survey administered to 60,000 households. The outcomes are five indicators of labor force status including retired, working part-time, unemployed, disabled, and number of hours per week usually worked at all jobs. Over 85 percent of men in the sample are White and around 25-35 percent have a college education (the percentage varies by birth cohort). The total number of observations ranges from over 97,000 for those ages 55-60 to 67,000 for those ages 66-70.
Findings
Employment
-
The study suggested that the Social Security reforms may have increased the retirement rate for men ages 55-60 but decreased it for men ages 61-65. For men ages 66-70, the reforms were not associated with a significant change in the retirement rate.
-
The study suggested that the reforms were not associated with a significant change in work hours for men ages 55-60 or 66-70 but may have increased average hours worked per week for men ages 61-65.
-
The study suggested that the reforms may have decreased the share of men ages 55-60 who worked part-time while increasing the share of men ages 61-65 who worked part-time. For men ages 66-70, the reforms were not associated with a significant change in the share who worked part-time.
-
Finally, the study suggested that the reforms were not associated with a significant change in the unemployment rate for men ages 55-60 or 61-65 but may have increased the share of men ages 66-70 who reported they were unemployed.
Considerations for Interpreting the Findings
The author accounted for many possible pre-existing differences between the groups being compared, including possible differences in age, race, education, marital status, census region, the state-level unemployment rate, and more. However, the author only used data that began after the intervention; therefore, the study could not account for possible differences between the groups in employment status or other work patterns prior to the intervention. Therefore, pre-existing differences in employment status or work history between the cohorts—and not the 1983 Social Security reforms—could explain the observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because the author could not ensure that the groups being compared were similar on the outcome measures before the reforms. This means we are not confident that the estimated effects are attributable to the 1983 Social Security reforms; other factors are likely to have contributed.