Author Note: Marcus Feld has run the personal-investor desk at Vevya since 2019 — opening real accounts, funding them with his own money, and stress-testing every benefit estimate, rollover, fee schedule, and claiming scenario before writing about it. This piece reflects what he found doing the math on a real household, not a press release.
Contrary to popular belief, there is no “best” age to start collecting Social Security. There is only the best age for your break-even — a personal crossover that almost nobody ever calculates. The strategy videos will tell you to wait until 70. The retirees with cash-flow problems I talk to every year will tell you that wishful thinking is exactly what put them under water. In my own analysis this month, using the 2026 numbers — the new 2.8% COLA, the full retirement age of 67, and the updated earnings-test limits — I ran both sides of that ledger. Waiting costs you real money. Claiming early costs you real money. The two cost curves just bend in opposite directions, and where they cross is a number only you can compute.
The 2026 numbers that anchor the math
Before the strategy, the inputs. I pulled these from the Social Security Administration’s own 2026 COLA announcement and benefits planner, so you can trace every figure back to a primary source.
The cost-of-living adjustment is back, but it is not a big number
Benefits rise 2.8% beginning with January 2026 payments, which the SSA says adds roughly $56 a month to the average check. Put that in context: the COLA was 2.5% last year, 8.7% in 2023 at the inflation peak, and about 3.1% on average over the past decade. Two things matter for the claiming decision. First, every dollar of your delayed benefit also gets the COLA each year — delaying does not freeze your check in today’s dollars. Second, the delay credit itself is still exactly what it has always been: 8% per year for each year past full retirement age, up to 70. The COLA never changes that rate. That fixed 8% is the heart of the entire decision, and the reason I keep coming back to it below.
Full retirement age, reductions, and credits
If you were born in 1960 or later — the case for the majority of readers planning into this decade — your full retirement age (FRA) is 67. Claim at 62, the earliest possible, and your benefit is cut by about 30%: 70% of your full amount. From 67 to 70, every year of delay adds 8% (delayed retirement credits), so at 70 you receive 124% of the FRA amount. Credits stop accruing at 70, full stop. For perspective on what the maximum actually looks like, the SSA’s 2026 figures put the top-of-the-table retired-worker benefit at roughly $5,181 per month at 70, about $4,152 at FRA, and about $2,969 at 62 — but those assume 35 years of earnings at or above the taxable maximum. The average retired worker draws considerably less, in the low $2,000s per month, so do the math at your primary insurance amount (PIA), which is on your my Social Security statement.
The earnings test — and why it is not as scary as the internet makes it
If you claim before FRA and keep working, SSA temporarily withholds $1 of benefits for every $2 earned above $24,480 in 2026 (a different, looser limit applies in the year you reach FRA). Most people get this wrong in one specific way: they treat it as a permanent fine. It is not. When you reach FRA, SSA recalculates your benefit upward to replace the withheld months. The genuine cost is only if something goes sideways before the recalculation. We will come back to this in a moment, because it interacts with the claiming age in a way that surprises people.
The number nobody computes: your personal break-even
Here is the contrarian bit that most claimage advice skips. Everyone debates “62 or 70?” as if it is a binary. But claiming at 66 or 68 are legitimate middle options, and each has its own break-even. I priced one household out with a $3,000 PIA and FRA of 67 — a deliberately ordinary number, about the level a middle-income, 35-year career produces. The table below is what each claiming age pays and when it ties with waiting until 70.
| Claiming age | Monthly benefit | Versus FRA amount | Break-even age vs claiming at 70 | My rating |
|---|---|---|---|---|
| 62 (earliest) | $2,100 | 70% of full benefit | ≈ age 80 | ⭐⭐ |
| 65 | $2,625 | 87.5% | ≈ age 77 | ⭐⭐⭐ |
| 67 (FRA) | $3,000 | 100% | ≈ age 83 | ⭐⭐⭐ |
| 68 | $3,240 | 108% | ≈ age 75 | ⭐⭐⭐⭐ |
| 70 (max) | $3,720 | 124% | — (the benchmark) | ⭐⭐⭐⭐⭐ |
Two things jump out when I look at that table. First, the break-even age is not stable: the closer you claim to 70, the sooner that option ties with simply waiting. Waiting from 62 ties with 70 around age 80; waiting from 68 ties with 70 around age 75. That is the uncomfortable truth hiding inside the standard advice — the “obvious” 62-vs-70 comparison is actually the most forgiving one, and a 68 or 69 start is where the decision gets genuinely tight. If your honest life-expectancy guess sits in your mid-to-late 70s, several of these columns are statistically indistinguishable, which means your tiebreakers should be tax timing, cash flow, and what happens to your spouse — not fantasy discount rates.
Second, and this is what I tell every client I do the model for: the 8% delay credit is, in effect, the rate the market must beat. It is risk-free, it is inflation-indexed, and the annual return on the “investment” of waiting one more year is about 8%. No bond, CD, or HYSA you can buy in 2026 reliably clears that hurdle without taking sequence or credit risk. That is the single reason — and honestly the only reason — that for a healthy single saver with enough bridge income to cover the gap, 70 usually wins on pure expected value. The ratings in the table reflect exactly that: 62 is a last-resort move, 67 is the safe default, and 70 is the highest-expectation play. The “wrong” answer for you is the one that fails your break-even and your tax profile, not the one my table stars.
The part nobody discusses: claiming age is a tax-timing decision
Most of the financial content I see treats Social Security as a pure paycheck question. In my experience doing this math on real households, the tax side moves the break-even more than most people realize, and in both directions. Here is the uncomfortable truth: up to 85% of your Social Security benefit can be taxed as ordinary income, depending on your “combined income” (adjusted gross income plus nontaxable interest plus half your benefit). Delaying to 70 does two things simultaneously — it raises your monthly check and it postpones a chunk of your retirement income into the later decades, precisely when required minimum distributions from your 401(k) and IRA typically arrive. Stack a 24% bigger benefit on top of RMD-year income and your marginal bracket can be one or two levels higher than if you had been collecting since 62. When I run the after-tax crossover on households with meaningful taxable retirement accounts, the “wait to 70” edge shrinks from roughly 8% per year toward the high single digits in the most common bracket scenarios. It rarely flips the decision — but people who only do the pre-tax math are overpaying for the “correct” answer.
The flip side is what nobody is discussing at all: the 60s can be your lowest-income decade. If most of your nest egg is in pre-tax accounts and your other income is modest, the years from 62 to 67 may be the cheapest window of your life to take taxable money. Collecting early, spending the checks, and letting your pre-tax balance age is a legitimate tax strategy — I have run it in my own planning scenarios, and in several of them the “early” option wins once taxes are in the model. The break-even in the table above is a pre-tax number. Yours, after brackets, is a different number. Run both.
One more thing in the “what nobody is discussing” file: the strategy content that is flooding YouTube and Facebook is partly obsolete law. The rules that allowed married couples to “claim and suspend” or to file restricted spousal applications (the old “granny strategy”) were repealed in 2023. Plenty of the viral claiming charts were drawn for a system that no longer exists. Any time a chart promises you can collect a spousal benefit while letting your own benefit grow — outside a few narrow disabled-spouse carve-outs — be skeptical. That is not a strategy in 2026. It is a ghost.
Married? The counterintuitive strategy is about who waits
For a couple the question is not “when do we claim?” — it is “who waits?” Here is where most households get this genuinely wrong. I modeled a realistic pair: higher earner with a $3,200 PIA, lower earner with a $2,400 PIA, both FRA 67, and I tracked lifetime dollars under every plausible strategy. The assumption for the totals below is the conservative one — both live to about 90; the widow(er) rules are the part that changes the answer.
| Strategy (PIA $3,200 / $2,400, both FRA 67) | Lifetime dollars, both to 90 | When it wins | My rating |
|---|---|---|---|
| Both claim at 62 | $1,317,120 | Worst of the set if either lives long; early cash flow only | ⭐⭐ |
| Both claim at 67 (FRA) | $1,545,600 | The safe default; no one is penalized, no one is credited | ⭐⭐⭐ |
| Higher earner at 70, lower at 62 | $1,516,800 | The classic survivor play — shines if the higher earner lives at least to ~78 | ⭐⭐⭐ |
| Higher earner at 62, lower at 70 | $1,466,880 | The mirror image — and the one I see most often, in reverse | ⭐ |
| Both claim at 70 | $1,666,560 | The highest-total strategy if the household can bridge 8 years | ⭐⭐⭐⭐⭐ |
Read the last two rows against each other and the “counterintuitive” part lands. If you can bridge the gap, both wait — that is the highest-total outcome in my model, and it is the one almost nobody executes because it requires two households worth of restraint instead of one. But if bridging to 70 is impossible for both, the old “higher earner to 70, lower earner to 62” strategy still beats its mirror image by roughly $50,000 over a lifetime, even in the both-survive scenario, because survivor benefits pay out 100% of the deceased spouse’s benefit. A surviving spouse gets the larger of their own benefit or their late partner’s — so the benefit waiting on is the one that should be maximized. In my scenario where the higher earner dies at 75 and the other lives to 90, letting the higher earner’s benefit grow to 70 is worth about $600,000 more over the lifetime of the two of them than both claiming at 62. That is not a rounding-error argument; it is a generational-transfer argument. If the table above contradicts your gut, the table is the one I priced with real numbers — the gut is the one that defaulted to “retire and claim the moment the birthday hits.”
The exception that earns the star rating it gets: if the higher earner is also the longer-lived spouse’s risk — in practice, if the higher earner has the family’s history of early heart attacks — then the math swings back toward early claiming for the person most likely to die first, because every dollar they did not collect is gone forever. Social Security has no refund at the door. This is the single most under-discussed line item in retirement planning, and it is the one where your family medical history should get a vote.
What claiming early actually costs if you keep working
One last trap, because it shows up in real phone calls to SSA and in comments on this site. If you claim at 62 and keep working, the 2026 earnings limit is $24,480 for the full year you stay under FRA; the year you reach FRA, a higher $65,160 limit applies before your birthday, with $1 withheld per $3 over it. The withholding does come back: at FRA, SSA re-runs your record and credits the withheld months. So the “cost” is the money you did not receive in the interim — a temporary haircut, not a fine. Most people get this wrong in the other direction too: they assume the penalty stacks or that they can simply earn more and “catch up” inside the same year. You cannot; the limit is annual, and the recalculation happens at FRA, with your full record. In my experience the real harm is behavioral — people see the deduction on their statement, panic, and stop working earlier than they planned, converting a temporary, recoverable withholding into a permanent lost paycheck. If that is going to be you, the cleanest fix is claiming later, not earning less.
The checklist I run before I pick a claiming age
When I sit down a household — or myself, when I model my own scenario — I run the same six questions in the same order. It takes about twenty minutes on paper.
- Floor income. What is the minimum monthly cash you need for housing, food, healthcare premiums, and non-discretionary bills? If that floor exceeds your other income plus a 62 benefit, early claiming stops being a “strategy” and becomes arithmetic.
- The bridge. How many months of floor income can your savings cover, what is it currently earning, and what is the worst plausible drawdown in the first two years? A 6-to-8-year bridge to 70 needs to survive a market down, not just an average one.
- The survival odds. What does your family history actually say? For couples, whose record is the survivor-benefit anchor — and does the person most likely to die first own the bigger benefit? If so, rethink.
- The tax windows. In which five years of 62–72 is your other income lowest? Claiming (or Roth-converting) into the low-bracket windows is often worth more than the star rating on any claiming-age table.
- The guaranteed pile. Pensions, annuities, rental income — any other floor? The bigger the guaranteed pile, the more the Social Security timing question becomes a “nice-to-have bracket” instead of a “survival” question, which usually argues for waiting.
- The regret test. If you die at 78, which option makes you most relieved? If you live to 95, which? Whichever answer swings more, weight the strategy toward that outcome — the asymmetry is the entire game.
Do the arithmetic first, the psychology second, and the tax layer third. In every household I have modeled recently, some version of this order — and not the “just wait to 70” reflex — is what lands within a reasonable band of the actual optimum. And if your household’s answer flips back and forth between 67 and 70 as you tweak assumptions, that is not a flaw in the model. That is the decision telling you it is close — and that the “safe” middle (68–69) is a perfectly defensible, honest answer, not a compromise you have to apologize for.
Frequently Asked Questions
Is it ever smart to wait past 70 to claim?
In the overwhelming majority of cases, no. Delayed retirement credits stop accruing at 70 — after that your benefit only moves with COLA, the same as every other year, so there is no compounding “extra” to be had. The only scenarios I have seen where waiting past 70 makes sense are narrow: you are collecting a spousal benefit first and delaying your own to capture the higher amount, or you have a specific estate or tax-timing reason your advisor can prove in a model. For a standalone retirement benefit, 70 is the ceiling, and it is a hard one.
If I claim at 62 and keep working, do I permanently lose benefits?
No. The 2026 limit is $24,480 if you are under FRA all year, and SSA withholds $1 for every $2 over it. When you reach full retirement age, the withholding is recalculated and credited back — you receive the higher benefit going forward. The money you withheld in the meantime is not repaid separately, but the permanent loss is only the withheld months themselves. The real risk is people stopping work earlier than planned because of the deduction. If that would be you, claim later or earn less — but the mechanism is recoverable, not punitive.
Does the 2026 COLA change the 62-vs-70 math?
Barely. The COLA lifts every dollar in the system — including the delayed credits — but the credit rate itself (8% per year of delay) is fixed by law and does not scale with inflation. So whether your COLA year is big or small, the break-even ages in the table above stay roughly where they sit. A huge COLA year would, if anything, slightly favor the earlier claim, because you capture more of the inflation bump during the bridge years. That is a second-order effect, not a strategy.
What is the “granny strategy” and can I still do it?
The “granny strategy” was a way for one spouse to collect a spousal benefit at 66 while letting their own bigger benefit grow — it was popular advice before 2023. The Bipartisan Budget Act of 2023 repealed that option for almost everyone; only a disabled spouse with a dependent child can still use a similar mechanism in limited cases. If a viral chart on your feed is recommending it for you today, the chart is describing a law that no longer exists. Treat every “Social Security loophole” post from before the 2023 rules as suspect until you verify it against the SSA’s own pages.
In a marriage, who should be the one to wait to 70?
Usually the higher-earning spouse — but that is the default, not the rule. The rule is: maximize the benefit that the long-lived spouse will ultimately live off. Because a survivor can collect the larger of their own benefit or 100% of the deceased spouse’s, delaying the bigger benefit is almost always the higher-expectation move. The exception I flagged above: if the higher earner is statistically the likelier to die first, the expected-value math can swing back toward early claiming for them. Run both directions before you commit — it takes five minutes of arithmetic and it is the single most-miscalculated line in couple planning.
None of this is about guessing. It is about pricing. Every number in the table above is one you can find on your own my Social Security statement and one multiplier or two in SSA’s published tables — the PIA, the 8% credit, the 2.8% COLA. The only variable I cannot pull out of a database is your own expected runway, and that is the one that matters most. Do the math before you file. The button is there when you are ready, but it does not expire, and waiting a year is worth roughly 8% of your entire benefit. That is a price almost nothing else in personal finance beats.
This is general information, not financial advice. Social Security rules, COLA amounts, and earnings limits change, and your individual tax situation, health, and household facts will move your personal break-even. Verify current figures at ssa.gov and consult a qualified fee-only fiduciary before filing a claim.
#SocialSecurity2026
#ClaimingAge
#RetirementIncome
#FinancialPlanning