A retiree in Naples executes a Roth conversion in the spring. The math is sound, the amount is deliberate, the bracket space is used carefully. For eighteen months, nothing seems to happen. Then, without warning, a letter arrives from the Social Security Administration. The Medicare premium is going up — not a little, but by several thousand dollars over the course of the year. The surcharge is technically not a tax. But it has the effect of one. And it is being charged, the letter explains, because of income reported two years prior.
This is IRMAA — the Income-Related Monthly Adjustment Amount — and it is one of the most avoidable and most frequently paid costs of retirement. Unlike most taxes, it does not appear on a return. It arrives as a change in Medicare premiums, often with no explanation that clearly ties the current bill to the two-year-old decision that caused it. The retiree who receives the letter is often startled, occasionally angry, and almost always unaware that the conversion, the sale, or the withdrawal that generated the income had this consequence attached to it.
There is a version of retirement planning in which IRMAA is treated as a surprise. There is a better version — the one this issue is written for — in which it is treated as a constraint. The distinction matters. A surprise is something that happens to you. A constraint is something you plan around. And IRMAA is one of those constraints that rewards attention generously and punishes inattention with a persistence most retirees do not see coming.
A Surcharge Disguised as a Premium
IRMAA is a structural feature of Medicare Part B and Part D, introduced to means-test premiums for higher-income beneficiaries. The mechanics are straightforward in principle: the Social Security Administration looks at the modified adjusted gross income reported on a retiree’s tax return, compares it to a schedule of thresholds, and applies a surcharge on top of the standard Medicare premium for anyone whose income exceeded the lowest threshold. The higher the income, the higher the surcharge — up to premiums that, for the highest bracket under the current schedule, can amount to several thousand dollars per person per year above the baseline.
What makes IRMAA unusual is the lag. The premium charged this year is based on the modified adjusted gross income from the tax return filed two years ago. A retiree reviewing their 2024 return in April of 2025 is not looking at a historical document. They are looking at the precise income figure that will determine their Medicare premiums in 2026. This is why IRMAA deserves its own category of attention in retirement planning. The decision that triggers the surcharge is made long before the surcharge arrives, and by the time the bill shows up, the income that caused it is no longer reversible.
The structure rewards a kind of discipline that is otherwise rare in tax planning: a two-year forward calendar. A retiree who is disciplined about bracket management within the current tax year, but who does not think about how the current year’s income will affect Medicare premiums two years from now, is optimizing for one constraint while violating another. The result is often a tax plan that looks excellent on this year’s return and produces an unpleasant surcharge letter eighteen months later.
The Cliff Problem
The second feature that makes IRMAA dangerous is its cliff structure. Unlike federal income tax brackets, which apply a higher rate only to the dollars earned above each threshold, IRMAA applies the full surcharge on the entire premium the moment income crosses a threshold. One dollar over the line produces the full increase. The thresholds are fixed — they do not phase in.
The practical consequence is that IRMAA creates sharp, invisible edges throughout the income landscape. A retiree whose income lands one hundred dollars under a threshold pays one premium level. A retiree whose income lands one hundred dollars over the same threshold pays thousands of dollars more per year, per spouse, for every Medicare-eligible household member. The difference is produced by a difference in income so small that it would be invisible on any other tax or planning document. On the IRMAA schedule, it is the difference between two premium brackets.
For a retiree executing a Roth conversion, this cliff logic changes the optimization calculus entirely. The question is not merely “how much should I convert this year?” The question is “how much can I convert before I cross the next IRMAA threshold?” Because the federal bracket itself is a continuous gradient for income tax purposes but a cliff for Medicare purposes, the optimal conversion amount is often bounded not by the federal tax bracket but by the IRMAA line sitting somewhere inside it. A retiree who converts an additional fifty thousand dollars at what appears to be a favorable marginal rate may find that the conversion cost, once the Medicare surcharge is included, is materially higher than the spreadsheet suggested.
This is why Roth conversion planning resists spreadsheets alone. The spreadsheets that matter are not only the ones that show federal tax brackets. They are the ones that overlay the IRMAA schedule on top of the conversion plan and identify the points at which one more dollar of income crosses a line that produces a disproportionate cost.
How IRMAA Compounds With Everything Else
In isolation, IRMAA is a manageable constraint. In combination with the other forced and optional income events of retirement, it becomes a coordination problem of real significance.
The most common source of IRMAA surprises is the interaction between required minimum distributions and other income. Consider, for example, a retiree taking a one-hundred-and-twenty-thousand-dollar RMD in a year when they also sell appreciated stock to fund a home renovation — or trigger a capital gain in a rebalancing move, or execute a partial Roth conversion, or claim a lump-sum Social Security payment. Any of these can push modified adjusted gross income past one or more IRMAA thresholds without the retiree realizing the combined effect. Each decision, viewed individually, appears reasonable. The combination produces a surcharge that neither decision alone would have triggered.
The coordination problem is compounded for married couples, because IRMAA thresholds apply to the joint return and the surcharges apply to both spouses’ premiums. A single income event that crosses a threshold for a married couple produces two surcharges, not one. This doubling is rarely visible in the planning conversation until the letter arrives.
Withdrawal sequencing is the most important lever for managing this. A retiree who understands that qualified Roth withdrawals are invisible to IRMAA, that qualified charitable distributions from an IRA can (within current annual limits and eligibility rules) offset the RMD without increasing modified adjusted gross income, and that long-term capital gains on a stepped-up basis can fund meaningful spending without crossing a threshold, has tools that a retiree drawing mechanically from a single account simply does not have. The difference between an IRMAA-aware withdrawal strategy and a default one can be thousands of dollars per year — every year, for as long as both spouses are enrolled in Medicare.
And then there is the widow’s penalty, which arrives here as well. When one spouse dies, IRMAA thresholds for a single filer are not half of the joint thresholds. They are considerably lower in proportion. The same income that kept a married couple below a threshold can push the surviving spouse two brackets higher. This is another one of those silent structural shifts that the well-constructed retirement plan anticipates and the default plan walks into.
When Life Changes Faster Than the Calendar
For retirees who have already been surcharged, one of the least-known and most useful features of the IRMAA system is the life-change exception. The Social Security Administration recognizes that the two-year lookback can produce absurd results when income has fallen dramatically due to a major life event. A retiree whose 2024 return reflected pre-retirement earnings, but who retired in early 2025 and now has materially lower income, should not be paying 2026 premiums based on income they no longer earn.
The remedy is Form SSA-44, which allows a beneficiary to request a re-determination of their IRMAA based on expected current income following a qualifying life-changing event. The qualifying events are specific: marriage, divorce, death of a spouse, work stoppage or reduction, loss of income-producing property, loss of pension income, and a handful of others. Not every income drop qualifies. A retiree whose Roth conversion created the surcharge cannot appeal on the grounds that the following year’s income will be lower — the event must be on the SSA’s list.
When a qualifying event has occurred, the form is often worth filing. The re-determination can adjust or eliminate the surcharge in many cases, sometimes with meaningful savings. It is also worth knowing about in advance of the event, because a retiree who does not know the form exists may spend an entire year paying a surcharge that could have been removed with a few pages of paperwork filed at the right time.
This is one of those small administrative details that separates retirees with structured planning from retirees without it. The form itself is not complicated. The awareness that it exists, and the discipline to file it at the moment it becomes relevant, is the kind of thing that does not show up in a financial plan but shows up in the results.
Planning With Memory
The deeper lesson of IRMAA is not about Medicare premiums. It is about the kind of planning that retirement actually requires.
In the accumulation years, most tax and financial decisions produced consequences within a twelve-month window. A contribution deduction, a capital gains event, a tax-loss harvesting transaction — each reconciled on the return for the year it occurred, and then receded into history. Retirement does not work this way. Roth conversions cast shadows into the years ahead. Withdrawal decisions shape future RMDs. Income events ripple into IRMAA calculations that will not produce bills for another twenty-four months. The planning horizon that worked during accumulation is too short for the distribution phase.
IRMAA is the most concrete example of this. It is a tax in everything but name, levied with a two-year memory, on decisions the retiree has likely forgotten by the time the bill arrives. The retirees who manage it well are not the ones who react faster. They are the ones who think further ahead — who consider not only what an income decision does this year, but what it will do in two years, and in five, and in the arc of the household’s entire distribution phase.
That discipline is not unique to Medicare planning. It is the same discipline the Roth conversion window requires, the same discipline the Social Security claim requires, the same discipline that separates a retirement plan from a retirement account. The question is not whether the decision looks good this year. It is whether the decision still looks good in the years that follow — when its consequences, invisible at the time of the decision, begin quietly showing up in the mail.
