Retirement Before the Social Security Entitlement Age

Featured in print Bulletin on Aging & Health

In the U.S. and many other developed countries, increasing longevity and long-term fiscal imbalances in the social security system are prompting policy makers to consider increasing the age at which retired workers would be eligible for benefits. While this policy change would undoubtedly save money, it also raises concerns about whether workers would be able to continue working until the new eligibility age and how their well-being might be affected if they were unable to do so.

Some light may be shed on this question by examining the experiences of U.S workers who retire before the Social Security early eligibility age of 62. This is the goal of a new paper How Is Economic Hardship Avoided by Those Retiring Before the Social Security Entitlement Age? (NBER Working Paper 18051) by Kevin Milligan. Specifically, the paper asks three questions. First, who retires early? Second, what are the sources of income before Social Security benefits are available? And third, how do non-workers avoid hardship?

Making use of data from the Health and Retirement Study, the author first conducts a descriptive analysis to determine the characteristics associated with early retirement. The analysis focuses on individuals working at ages 53 to 54, for whom retirement is a salient concept.

Among women, the more highly-educated work longer, as do those with employer-provided pensions or health insurance, though having a defined benefit pension is associated with earlier retirement, likely because of the incentives imbedded in the benefit formulas. Job characteristics such as stress or physical demands do not predict exit. Patterns for men are similar, except that there is no effect of education; there are no patterns with respect to race for either gender. Overall, demographics and work place characteristics (other than pensions) do a relatively poor job of explaining retirement before age 62. The single best predictor of retirement is age, which may reflect changing tastes for work or the effect of eligibility for Social Security and Medicare.

Next, the author explores the composition of income by age and has several key findings. First, the income distribution compresses with age, as the retirement of those in the top half of the income distribution lowers their income, while those at the bottom see their incomes rise when they become eligible for Social Security. Second, government income makes up only a small share of total income until age 62 and a large share thereafter, as individuals become eligible for and claim Social Security benefits. Third, non-labor private income -- including private pensions, annuity income, and capital income -- is widely held, but its distribution is skewed, particularly in the case of capital income.

Finally, the author focuses on those retired before age 62 in order to understand poverty rates in this group and the role that different income sources may play in keeping individuals out of poverty. Poverty rates are about 10 percent for women throughout their late 50s and a bit lower for men, but those who are retired (have zero earnings) have poverty rates twice as high, around 20 percent. Interestingly, the gap in poverty rates narrows considerable after age 62, suggesting an important role for Social Security.

The last part of the analysis is an accounting exercise designed to measure the extent to which different income sources, on their own, would be sufficient to lift each family out of poverty. For women ages 55 to 66, husband's income is the most important source of income, lifting 56 percent of women over the poverty line. While most women (64 percent) have some capital income, fewer than one in five are lifted over the poverty line on this basis. Pension and annuity income is held by less than 20 percent of women and lifts only 5 percent over the poverty line (10 percent after age 62). Own Social Security benefits are (unsurprisingly) more important after age 62, yet even then lift only 11 percent of women out of poverty.

Results for men are a bit different. Spousal income is also the single most important means of poverty avoidance, but spousal government income is much less important, perhaps due in part to the fact that many wives are younger and may not have claimed Social Security benefits yet. Capital income and pension and annuity income are more important, lifting 26 and 17 percent of men out of poverty, respectively.

In concluding, the author notes, "four out of five people not working at ages 55 to 66 are able to avoid poverty through some combination of government and non-labor private income." Even so, poverty rates for early retirees are substantially higher than those for workers. This gap in poverty rates narrows after age 62, suggesting "among those remaining in poverty, Social Security entitlement at age 62 offers some respite."

This research was supported by the U.S. Social Security Administration through a grant to the NBER as part of the SSA Retirement Research Consortium. The author has disclosed a financial relationship of potential relevance for this research; further information is available at