How to choose a retirement age

What actually changes at 55, 59½, 62, 63, 65, 67 and 75 — the ACA bridge cost of going early, and what an extra working year is really worth.

"What age should I retire?" is not really a question about age. It is a question about money, health coverage, tax, and what you want your days to look like — and the age is the number those things produce. That is the argument in how to plan for retirement, and this spoke of the Decision tools hub turns it into something you can actually choose with.

Because you do still have to pick a year. And the American retirement system has hard edges at specific ages, which means some years are genuinely better than the ones on either side of them.

What changes at each age

Two rows in that table are routinely misread, so they get their own sections below: 55 and 63.

The rule of 55 has a trap in it

The rule of 55 lets you take money out of a workplace plan without the 10% early-distribution penalty if you separate from service during or after the calendar year you turn 55. It is the main reason a 55-year-old can retire at all without an elaborate workaround.

But it is narrower than people think:

• It applies only to the plan of the employer you separated from. Not your old employer's plan from three jobs ago, and not your spouse's.
• It does not apply to IRAs at all. IRS Topic 558 is specifically about "retirement plans other than IRAs" for exactly this reason.
• Which means: rolling your 401(k) into an IRA when you leave destroys it. This is the trap. The rollover is the standard advice, the paperwork is easy, and doing it at 55 can lock the money up for four and a half years.

One bonus most people never hear: distributions from a governmental 457(b) plan are generally not subject to the 10% additional tax at any age after you separate. Public-sector employees with a 457(b) have an early-access route that private-sector savers do not.

Age 63 is a decision you make without noticing

Medicare's income-related monthly adjustment amount — IRMAA — is set from the tax return you filed 2 years earlier. So the income you report in the year you turn 63 determines your Part B and Part D premiums at 65.

Cross the first tier — $109,000 single, $218,000 joint — and you pay roughly $1,148 extra for the year, per person on Medicare, and it is a cliff rather than a ramp. Sell a rental property at 63, take a big severance, or do an unplanned Roth conversion, and the bill lands two years later when you have forgotten why.

The real cost of going before 65: the health-insurance bridge

This is the number that moves retirement ages, and most people discover it late.

Medicare starts at 65. Retire at 60 and you are buying your own health insurance for five years. As a national reference point, an unsubsidised benchmark silver plan for a 60-year-old costs about $15,914 a year — and premiums rise steeply with age, so the last years before Medicare are the most expensive ones.

The subsidy that used to soften this got much harsher. The enhanced premium tax credits expired on 31 December 2025, restoring the cliff: household MAGI one dollar over 400% of the federal poverty line — $62,600 for one person, $84,600 for a couple — forfeits the entire premium tax credit. Not tapered. Zero.

The consequence for choosing an age is specific and unintuitive: before 65, your retirement income is not just an income decision, it is a health-insurance decision. Every dollar you withdraw from a traditional account, every capital gain you realise, every conversion, raises the MAGI that the cliff is measured against. A retiree who wants to fill their 12% tax bracket and a retiree who wants to stay under the cliff are being pulled in opposite directions.

Below the cliff there is still help — the income-based cap on the benchmark premium survived, it simply became narrower and steeper — and below 100% of the poverty line the subsidy vanishes again, so it is possible to squeeze your income too far. Price your own bridge with the ACA bridge tool, and read the ACA subsidy cliff is back for 2026 and health care before Medicare.

Why one more year is worth so much — and why that argument is dangerous

An extra working year does four things at once, and they all push the same way:

1. You add a year of contributions — up to $8,000 of catch-up on top of the regular limit if you are 50+, or $11,250 at 60–63. See the catch-up planner and catch-up contributions in 2026.
2. You do not withdraw for a year — so the portfolio compounds on a bigger base instead of a shrinking one.
3. Your retirement is one year shorter, so the same pot has to cover fewer years.
4. You may raise your Social Security benefit, both by replacing a weak year in your top 35 and by delaying the claim.

That is why "just work one more year" closes gaps that saving more cannot. It is also why the advice is dangerous when it becomes the only plan. The lever works by compounding several small effects — but it is spending the scarcest thing you have, and it assumes a decision that may not be yours to make. People leave work early because of their own health, a parent's health, or a company reorganisation, far more often than they expect to.

So use the lever, but do not lean the whole plan on it. Test your numbers at an age one or two years earlier than your target with the retirement age tradeoff tool. If the plan only works at 67 and fails at 65, it is not a plan — it is a hope with a spreadsheet.

Run three ages, not one

Instead of guessing a single number, model three — early, target, and late — and compare the funding gap each one leaves (how much is enough covers the gap arithmetic):

• Early: the biggest gap, the longest ACA bridge, the most dependence on flexibility and part-time income.
• Target: your base case.
• Late: the smallest gap, more compounding, a higher Social Security floor, and fewer years to enjoy it.

Seeing the three side by side turns an anxious guess into a comparison. Run them through Am I on track? or the Retirement Checkup.

The half of this that is not arithmetic

The financial half of the decision is the easy half. The rest of it is not, and pretending otherwise is how people end up with a well-funded retirement they do not want.

Health is asymmetric. The years right after you stop working are the ones in which you are most likely to be well enough to use them. An extra year of work is a year of money, taken from the specific part of your life when money is most usable. That trade may still be worth it. It is not free, and no calculator will show you the cost.

Caring responsibilities usually decide it, not you. A parent needing care, or a spouse's illness, ends careers on a timetable no one chose. If that is foreseeable in your family, plan the age around it rather than around the market.

Some people should not retire, and that is fine. Work supplies structure, status and a large part of most people's social contact. If you have no idea what you would do on the first Monday, the honest answer may be to keep working — or to shift to part-time work, which happens to be excellent for the portfolio too.

The point of getting the numbers right is not to obey them. It is to know what the trade actually is, so the non-financial half is a choice rather than an accident.

Key takeaways

• The system has hard edges: 55 (rule of 55), 59½ (everything), 62 (earliest claim), 63 (Medicare's lookback begins), 65 (Medicare), 67 (full retirement age), 70 (maximum benefit), 73/75 (RMDs).
• Rolling a 401(k) to an IRA at 55 destroys the rule of 55. It does not apply to IRAs.
• Before 65, your withdrawal decisions and your health-premium decisions are the same decision. The ACA cliff is back for 2026.
• One more working year moves four levers at once — but a plan that only works if you can work one more year is fragile.
• Model three ages, and stress-test the early one. Then decide the non-financial half deliberately.

Educational only — not financial or tax advice.