Delaying Social Security is the cheapest longevity insurance available
The claim
For someone in average health reaching their mid-60s, deferring Social Security to age 70 raises expected lifetime benefits, and the gain is largest for the higher earner in a married couple because the survivor inherits the larger benefit.
Why we rate it moderate
The delayed retirement credit is statutory rather than estimated, so the mechanism is certain. Whether deferral pays depends on mortality and on real interest rates, both of which vary by household, so the conclusion is conditional rather than universal.
Claiming early permanently reduces the monthly benefit; claiming late permanently raises it, by roughly 8% a year between full retirement age and 70. Because the benefit is indexed to inflation and paid until death, deferral is functionally the purchase of a real, life-contingent annuity — and it is priced far better than anything an insurer sells, because there is no load, no credit risk, and no adverse-selection markup.
Shoven and Slavov worked through when it actually pays, and the answer depends on two things: how long you live and what real interest rates are. At the low real rates that prevailed for most of the last two decades, deferral was advantageous for a wide range of people. At high real rates the case weakens, because money in hand today can be invested instead.
The married-couple case is the one most often got wrong. When one spouse dies, the household keeps the larger of the two benefits, so deferring the higher earner's claim buys protection for whichever of the two lives longest. That makes deferral by the higher earner roughly a joint-life bet rather than a single-life one, and it survives far worse mortality assumptions.
The framing that ruins the decision is 'break-even age'. Presented that way it looks like a wager on living past about 80, which invites people to think about whether they feel lucky. It is better read as insurance: you are not betting on a long life, you are protecting against the financial consequence of one. The bad outcome being insured against — running out of money at 92 — is much worse than the bad outcome of insuring unnecessarily.
Practically, deferral usually has to be funded by drawing harder on the portfolio between 65 and 70. That is the trade being made, and it is the reason the decision belongs next to a withdrawal plan rather than on its own.
Where this breaks down
- Poor health or a family history of early mortality genuinely flips the calculation, and the person deciding usually knows more about this than any table does.
- Deferral has to be funded from somewhere. Someone with no bridge assets between retiring and 70 may have no realistic choice.
- Interactions with means-tested programmes, spousal benefits and the taxation of benefits can change the answer, and they are specific enough that this is exactly where individual advice earns its fee.
- Programme rules are politically determined and have been changed before. A benefit formula is not a contract in the way a bond is.
Sources
Follow these rather than taking our word for the summary.
John B. Shoven and Sita Nataraj Slavov (2014). Does it pay to delay social security?
Journal of Pension Economics and Finance, 13(2), 121-144
Finding: Delaying benefits raises expected present value for a wide range of individuals, with the gain increasing as real interest rates fall and concentrated in the primary earner of a couple.
Social Security Administration (2026). Delayed retirement credits
ssa.gov
Finding: Benefits increase by a statutory percentage for each month of deferral after full retirement age, up to age 70, after which no further credit accrues.