What Is the Social Security Break-Even Age?

Your break-even age is the age at which the larger checks from claiming later have finally made up for the years of checks you gave up by waiting. Before that age, claiming earlier put more total dollars in your hands. After it, claiming later has. It is one number, from one calculation, and it depends entirely on your own benefit amounts.

That is the answer to the question people actually type. The rest of this page is how the arithmetic works, roughly where the number lands, and — honestly — the several things this calculation does not see.

How is the break-even age actually calculated?

Three inputs, one division.

First, the head start: the monthly benefit you would receive by claiming earlier, multiplied by the number of months you would receive it before the later claiming date. That is the money the early claimer banks while the late claimer is waiting.

Second, the monthly gap: the difference between the two monthly benefits once both are running.

Third, divide the head start by the gap. The result is the number of months, after the later claiming date, that it takes the larger check to close the ground it lost. Add that to the later claiming age and you have the break-even age.

Here is the same thing with numbers. This example is illustrative — the amounts are invented, and every figure below is arithmetic, not a rule. Say full retirement age is 67 (20 CFR 404.409(a)) and the benefit at full retirement age is $3,000 a month. Claiming at 62 gives $2,100. Claiming at 70 gives $3,720.

• 62 versus 67. The head start is 60 months × $2,100 = $126,000. The monthly gap after 67 is $900. $126,000 ÷ $900 = 140 months, or 11 years and 8 months. Break-even lands around age 78 and 8 months.
• 67 versus 70. Head start: 36 months × $3,000 = $108,000. Gap: $720. $108,000 ÷ $720 = 150 months, or 12 years 6 months. Break-even lands around age 82 and 6 months.
• 62 versus 70. Head start: 96 months × $2,100 = $201,600. Gap: $1,620. $201,600 ÷ $1,620 = about 124 months, a little over 10 years 4 months. Break-even lands a bit past age 80.

Notice what this arithmetic ignores: cost-of-living adjustments, income taxes on the benefits, and any return on the early money. Add those and the number moves. A COLA is the same percentage on both checks, so it widens the dollar gap between them every year while the head start banked before you claimed stays a fixed number of dollars — which pulls the nominal break-even age a little earlier, not either way. Taxes are the piece that genuinely goes both ways, because the marginal rate depends on the household. A calculator that reports break-even to the month is reporting false precision.

Where does the break-even age usually land?

For most people running this calculation on their own statement, the answer clusters somewhere in the late 70s to early 80s. The comparisons that involve age 70 land later than the ones that stop at full retirement age, because the delayed credits are the last and largest piece of the increase.

But “somewhere in the late 70s to early 80s” is a description of a pattern, not a fact about you. The exact age falls out of a household’s own benefit amounts and its own full retirement age, and those two things vary. A number borrowed from an article — including this one — is not the number that comes off your statement.

What does full retirement age have to do with it?

Everything, because it is the pivot the whole calculation swings on. Benefits are reduced for entitlement beginning before full retirement age and increased for months after it.

Full retirement age is not a single number. It is a ramp set by birth date: 66 for people born January 2, 1943 through January 1, 1955, then rising by two months per birth year, reaching 67 for people born on or after January 2, 1960. The middle of that ramp is where this audience actually sits, so here it is in full (20 CFR 404.409(a)): born 1/2/1955–1/1/1956, 66 and 2 months; 1/2/1956–1/1/1957, 66 and 4 months; 1/2/1957–1/1/1958, 66 and 6 months; 1/2/1958–1/1/1959, 66 and 8 months; and 1/2/1959–1/1/1960, 66 and 10 months. Using a flat 67 will misstate both your reduction and your break-even. (This is the ramp for your own retirement benefit. The survivor benefit a widow or widower draws on a deceased spouse’s record runs on a different ramp — more on that below.)

The reduction itself is not linear. It is steeper for the 36 months immediately before full retirement age — five-ninths of one percent for each of those months, about 6.67% a year — than for any month earlier than that, which costs five-twelfths of one percent, about 5.00% a year (20 CFR 404.410(a)). The months closest to full retirement age are the expensive ones. Both rates are for your own retirement benefit: 42 U.S.C. §402(q)(1)(A) sets the per-month reduction by benefit type, and a widow’s or widower’s benefit reduces at nineteen fortieths of one percent instead, on its own schedule.

When do delayed retirement credits stop?

At 70. Credits accrue for each month from full retirement age up to, but not including, the month you turn 70, and then they stop.

You can still file after 70 — nothing prevents it — the benefit amount simply stops increasing (20 CFR 404.313(a)). That makes 70 the one point in the sequence where the trade-off ends: everywhere else, waiting costs checks and buys a larger one, and past 70 it only costs checks. It is the single unambiguous piece of an otherwise ambiguous decision.

Do spousal benefits work the same way?

No, and this is the most common piece of misinformation on the subject.

A spousal benefit does not earn delayed retirement credits. A wife’s or husband’s benefit is one-half of the worker’s primary insurance amount, and the regulations are explicit that the worker’s delayed credits are not used to increase the benefits of other family members entitled on that record. So a spouse whose own work record produces a smaller benefit gains nothing by delaying past their own full retirement age — the spousal amount is already at its maximum there. A spousal benefit is still reduced for claiming early, but at a different — and steeper — rate than the worker’s own benefit: for a full retirement age of 67, the maximum reduction is roughly 35% at 62, against roughly 30% for the worker’s own early claim.

The same applies to an ex-spouse claiming on a former spouse’s record, where the marriage lasted at least 10 years immediately before the divorce became final (20 CFR 404.331(a)(2)). To claim on an ex-spouse’s record, you also have to be currently unmarried (20 CFR 404.331(c)) and at least 62 (20 CFR 404.310(a)). If the divorce was at least two years ago, you can generally claim independently, on the ex-spouse’s earnings record, even if the ex-spouse has not yet filed.

Where the delayed credits do carry forward is to a surviving spouse. That distinction — dead spouse yes, living spouse no — is the whole ballgame for a married couple, and break-even analysis does not see it at all.

One more wrinkle worth flagging here: the full retirement age used for a survivor benefit is not the same ramp as the one above. It runs two years behind — 66 for people born January 2, 1945 through January 1, 1957, rising two months per cohort, reaching 67 only for people born on or after January 2, 1962. A widow or widower applying the retirement-benefit ramp to their own survivor claim will get the wrong number.

If it helps to have the numbers in one place, I keep them all in the Ultimate Retirement Guide.

What does break-even miss for a married couple?

The survivor benefit. When the first spouse dies, one of the two checks stops. What continues is generally the larger of the two — though if the deceased spouse claimed their own benefit early, the survivor benefit can be capped at what the deceased was actually receiving, rather than the full amount their record would otherwise support. That cap has a floor, and it is the half most people never hear. The benefit is reduced to what the deceased would be receiving if alive, or 82.5 percent of their primary insurance amount, whichever is larger (20 CFR 404.338(c)). So an early claim by the higher earner does not cut the survivor benefit without limit — on a $3,000 full benefit the difference between knowing the floor and not knowing it is roughly $375 a month, for the rest of the survivor’s life. A survivor benefit can also start as early as age 60 (20 CFR 404.409(c)), or 50 for a disabled widow or widower (20 CFR 404.335(c)), at a reduction, well before the survivor’s own full retirement age. The higher earner’s claiming decision — including every delayed credit they earned — still sets the ceiling under whoever lives longer.

So for a married couple the higher earner’s claim is not really a bet on that person’s own lifespan. It is a bet on the second death in the household, which lands considerably later than either individual’s expectancy. A break-even calculation run on one life expectancy answers a question the couple did not ask.

This is where the frame stops describing the thing. Break-even measures one quantity — total dollars received, on one life — and answers it correctly. It does not measure the survivor benefit, the tax cost of the withdrawals taken while waiting, or the difference between an average outcome and a long one. Delaying has the shape of insurance rather than an investment: it costs something up front and it pays most in the outcome that would be hardest to be unprepared for, which is living a long time or leaving a spouse who does. Insurance is not usually evaluated on a payback period, and that is the mismatch — not a verdict on which claiming age is right.

What does claiming early do to a Roth conversion plan?

This is the piece that matters most for households with large pre-tax balances, and break-even is silent on it.

The years between retirement and the start of required minimum distributions are usually the lowest-income years of a person’s life. That is the window in which converting IRA dollars to Roth costs the least, because there is bracket space sitting empty. Claiming Social Security early fills part of that space with benefit income — and because a portion of benefits becomes taxable as other income rises, benefit income and conversion income interact rather than simply stacking.

The trade runs the other direction too. If you delay claiming, you are generally spending IRA dollars in the meantime, and those withdrawals are ordinary income that occupies the same space a conversion would have. Neither path is free. The point is that the claiming decision and the conversion plan compete for the same brackets in the same years, and a break-even calculation cannot show you that because it only counts benefit dollars.

Does the earnings test change the math if I am still working?

Less than people think, because the earnings test is a deferral, not a loss.

If you claim before full retirement age and keep earning above the exempt amount, part or all of your benefit is withheld. The withholding runs at two different rates, and the second one is the one people never hear about. In any year before the year you reach full retirement age, excess earnings are 50 percent of what you earn above the exempt amount — a dollar withheld for every two dollars over. In the calendar year you reach full retirement age, that drops to one dollar withheld for every three dollars over, and it is measured against a much higher exempt amount (20 CFR 404.430(b)). Both figures are indexed and move each year, which is why I have written the structure and not the amounts. The structure does not move.

There is also a rule for the year you actually stop working. In a grace year, the test can be applied month by month against a monthly exempt amount rather than against the year’s total (20 CFR 404.430(a)). Someone who earns a large salary through June and retires can therefore collect from July, even though the annual figure alone would have withheld everything. That is worth knowing before you decide to wait a year for no reason.

What almost nobody is told is what happens next: any month in which the earnings test caused a deduction — whether the whole benefit was withheld or only part of it — is excluded when the permanent reduction for early claiming is recalculated. The adjustment takes effect at full retirement age. The benefit is stepped back up. Over a normal lifespan the withheld money comes back. Social Security’s manual is explicit that a partial deduction is credited exactly like a full one, and that prorating a deduction makes no difference to the adjustment (SSA POMS RS 00615.482 B.1).

Earnings from the month you reach full retirement age onward — including that month itself — are not subject to the test at all.

So the earnings test is a cash-flow issue, not a penalty. It can still be a real problem — a withheld check is a check you cannot spend this year — but describing it as money confiscated is wrong, and I have seen that error steer people into claiming decisions they would not otherwise have made.

What do people get wrong about this?

Treating it as the decision rather than one input. Break-even is a longevity bet expressed in dollars. It is a fair thing to want to know. It is not the same as knowing what to do.

Reading it as precise. The output looks exact — an age, to the month. The inputs are not: future COLAs, future tax law, and your own lifespan are all estimates.

Running it on one life. For a married couple the relevant horizon is the survivor’s, not the claimant’s.

Assuming the early money is invested. Most break-even calculators assume the early checks sit and do nothing. If you would genuinely invest them, break-even moves later; if you would spend them, it does not.

Believing spousal benefits grow to 70. They do not. A spousal benefit is at its maximum at the spouse’s own full retirement age, so months waited past that point produce no increase in it.

When this does not apply to you

Break-even analysis is mostly beside the point if:

• Your health or family history points to a shorter horizon. The calculation still runs, but it is answering a question you have better information about than the calculator does.
• You need the income now. If claiming at 62 is what keeps you from selling assets at a bad moment or taking on debt, the theoretical crossover age is not the operative constraint.
• You are single with no survivor to protect and no large pre-tax balance. With no survivor benefit and no conversion window at stake, break-even is closer to being the whole question — which is a genuine case, just not the common one for this audience.
• Your benefit is a small fraction of your income. For a household drawing most of its spending from a portfolio, the claiming decision moves the tax picture more than it moves the cash flow, and the tax question is not what break-even measures.
• You are affected by rules this page does not cover — a disability benefit, or benefits for a minor or disabled child on your record. Those change the arithmetic in ways a general explainer cannot.
• You have a pension from work that did not pay Social Security tax. Two rules used to reduce or erase your benefit here. Congress repealed both, and the repeal reaches benefits payable from January 2024 — so if you were once told the answer was zero, the arithmetic on this page is live for you again. See the Social Security Fairness Act.

Break-even is a real calculation with a real answer. It is the opening of the analysis rather than the close of it — and for a household with substantial pre-tax assets, it measures the smallest of the moving parts.

If you want the rest of the decision — the survivor benefit, ex-spouse rules, taxation of benefits, and how claiming interacts with the tax years around it — that all lives on the Social Security hub.