[HC-06] IRMAA: The Two-Year Lookback, the SSA-44 Appeal, and Bracket Management
A one-year toll charged on a two-year delay — priced per person, at a cliff
Section titled “A one-year toll charged on a two-year delay — priced per person, at a cliff”Pillar: Healthcare Navigation · Applies to: Anyone on Medicare, and anyone doing Roth conversions from age 63 onward Last verified: August 2026 · Refresh cadence: Annual (November, when CMS releases next-year premiums and brackets) Related: HC-02 Medicare 101 · HC-01 ACA Bridge · FN-02 Key Numbers · TX-02 Roth Conversion Strategy · ER-02 Roth Conversion Ladder · SS-03 Taxation of Benefits · SS-05 Coordinating SS + Medicare
Not advice. Educational reference. IRMAA determinations are made by SSA on your individual record, and appeal outcomes depend on documentation. Verify current brackets at Medicare.gov/SSA.gov and confirm your own determination notice before acting.
- IRMAA (Income-Related Monthly Adjustment Amount) is a surcharge on Medicare Part B and Part D, triggered by your MAGI from two years ago. Your 2026 premiums are set by your 2024 tax return. ✅
- It is a cliff, not a phase-in. One dollar over a threshold moves you a full tier. Crossing the first line costs a couple $2,296.80 for the year — an effective marginal rate of ~230,000% on that dollar. ✅
- It is charged per person. Both spouses pay. A couple in the top tier pays $13,872/year in surcharges alone. ✅
- The misconception this article exists to kill: IRMAA is not a tax bracket you’re stuck in. It’s a toll charged for exactly one year, because only one tax year feeds each premium year. A one-time spike self-cures in 24 months. People routinely wreck good Roth-conversion plans to dodge a cost that’s smaller than the tax it saves.
- The thing people under-fear: the widow(er) cliff. Single thresholds are half of joint, so a survivor with less income can jump two tiers. And the top bracket ($500k/$750k) is frozen until 2028, so inflation drags people into it. ✅
- Form SSA-44 knocks IRMAA down after eight qualifying life-changing events — retirement (“work stoppage”) is the big one. A Roth conversion, capital gain, or home sale is not an appealable event. ✅
1. The Mechanic
Section titled “1. The Mechanic”Medicare’s standard Part B premium covers about 25% of the program’s per-capita cost; the government subsidizes the rest. IRMAA claws that subsidy back from higher-income beneficiaries by raising their share to 35%, 50%, 65%, 80%, or 85% of cost. Part D gets a parallel surcharge. ✅
Premium year 2026 ← MAGI from tax year 2024 ← return filed in 2025SSA pulls the most recent return IRS has on file — normally two years back. If that return isn’t available, it reaches to three years back and corrects later.
IRMAA MAGI is its own definition — and it is not the ACA MAGI from HC-01:
IRMAA MAGI = AGI + tax-exempt interest (muni bonds)That’s it. No Social Security add-back beyond what’s already in AGI, no foreign-income add-back for most filers. Everything that lands in AGI counts: Roth conversions, IRA/401(k) withdrawals, RMDs, capital gains, dividends, interest, pension income, the taxable portion of Social Security, and net rental income. Muni interest — tax-free for income-tax purposes — still counts here. Tax-free Roth distributions do not. ✅
Three structural features that make it bite:
- Cliff, not slope. Cross a threshold by $1 and you pay the entire tier.
- Per person. Each spouse pays their own Part B and Part D surcharge off the same household MAGI. A married couple’s cost is always double the table below.
- No hold-harmless protection. The provision that caps most beneficiaries’ Part B increase at their Social Security COLA does not apply to IRMAA payers. ✅
2. The 2026 Table (2024 MAGI)
Section titled “2. The 2026 Table (2024 MAGI)”Per person, per month. Thresholds ✅; premiums ✅.
| 2024 MAGI — Single | 2024 MAGI — MFJ | Part B total | Part B surcharge | Part D surcharge | Combined surcharge | Annual, per person |
|---|---|---|---|---|---|---|
| ≤ $109,000 | ≤ $218,000 | $202.90 | — | — | — | $0 |
| – $137,000 | – $274,000 | $284.10 | $81.20 | $14.50 | $95.70 | $1,148.40 |
| – $171,000 | – $342,000 | $405.80 | $202.90 | $37.50 | $240.40 | $2,884.80 |
| – $205,000 | – $410,000 | $527.50 | $324.60 | $60.40 | $385.00 | $4,620.00 |
| – < $500,000 | – < $750,000 | $649.20 | $446.30 | $83.30 | $529.60 | $6,355.20 |
| ≥ $500,000 | ≥ $750,000 | $689.90 | $487.00 | $91.00 | $578.00 | $6,936.00 |
Double every annual figure for a married couple. Top tier: $13,872/year. → FN-02
Two structural quirks worth planning around:
- Single thresholds are exactly half the joint thresholds — except at the top, where single is $500,000 but joint is $750,000, not $1,000,000. High-income couples hit the top tier at 1.5× the single line, not 2×. A genuine marriage penalty. ✅
- The top bracket is frozen. The Bipartisan Budget Act of 2018 created the $500,000/$750,000 tier and barred inflation indexing until 2028. The lower four brackets index annually to CPI-U; the top one does not, so bracket creep pushes more people in every year. ✅
Part D note: the Part D surcharge is billed by Medicare (usually deducted from Social Security), not by your drug plan — and you owe it whether your Part D coverage is a standalone PDP or bundled inside a Medicare Advantage plan. Dropping Part D entirely doesn’t avoid it if you’re enrolled in a plan; going without creates late penalties instead (HC-02 §2). ✅
3. The Cliff Math, With Real Numbers
Section titled “3. The Cliff Math, With Real Numbers”A couple, both 66, 2024 MAGI of $218,000 — exactly at the line. Surcharge: $0.
Now the same couple at $218,001:
| Per person/mo | Per person/yr | Couple/yr | |
|---|---|---|---|
| Part B surcharge | $81.20 | $974.40 | $1,948.80 |
| Part D surcharge | $14.50 | $174.00 | $348.00 |
| Total | $95.70 | $1,148.40 | $2,296.80 |
One dollar of MAGI cost $2,296.80. That is the entire argument for buffer.
The tier-to-tier steps are smaller but still lumpy. Moving from tier 1 to tier 2 (crossing $274,000 MFJ) adds $144.70/month each — $3,472.80/year for the couple. Every threshold is a step of this shape.
Now the correction most people need. That couple’s $2,296.80 is a one-year cost. Their 2024 MAGI sets their 2026 premiums and nothing else. Their 2025 MAGI sets 2027, independently. There is no compounding and no permanence.
So compare honestly. Suppose converting an extra $40,000 to Roth at 63 pushes a couple one tier up:
| Amount | |
|---|---|
| Federal tax at 24% | $9,600 |
| IRMAA toll (tier 1, couple, one year) | $2,296.80 |
| Total cost of the conversion | $11,896.80 |
| Effective rate on the $40,000 | 29.7% |
If that $40,000 would otherwise come out as an RMD at 73+ — at 24%+ then, inside a bracket possibly compressed by a spouse’s death — the conversion still wins. IRMAA added 5.7 percentage points, not a veto. ✅
The reverse is also true: crossing a line to convert $1,200 is indefensible. Size matters, and the toll is fixed regardless of how far over the line you go. Once you’ve decided to cross a threshold, cross it decisively — the marginal cost of the next $20,000 within a tier is zero IRMAA.
4. The Widow(er) Cliff — The Under-Feared One
Section titled “4. The Widow(er) Cliff — The Under-Feared One”Single brackets are half of joint. When one spouse dies, the survivor keeps a large share of the household’s portfolio income but files under thresholds that halved.
Worked example. Couple, joint 2024 MAGI $260,000 → tier 1, $95.70/mo each.
The husband dies in 2025. By 2027 the survivor files single. Her income drops to $200,000 — she lost his Social Security benefit and a pension survivor reduction:
| Household (MFJ, $260,000) | Survivor (Single, $200,000) | |
|---|---|---|
| Tier | 1 | 3 |
| Surcharge/mo | $95.70 each | $385.00 |
| Annual surcharge | $2,296.80 (couple) | $4,620.00 (one person) |
Her income fell 23% and her personal surcharge quadrupled. This is the same compression that drives the widow(er)’s tax trap in TX-02 — IRMAA is simply the healthcare-premium expression of it. It is the strongest argument for converting aggressively while both spouses are alive and filing jointly.
One relief valve: death of spouse is a qualifying life-changing event on Form SSA-44 (§5). A survivor whose income actually dropped should file rather than wait two years for the lookback to catch up.
5. Form SSA-44: The Appeal That Actually Works
Section titled “5. Form SSA-44: The Appeal That Actually Works”IRMAA is set on two-year-old data, which is wrong for anyone whose income just changed — which describes every new retiree. Form SSA-44 lets you ask SSA to use your estimated current-year MAGI instead, if a qualifying life-changing event caused the drop.
The eight qualifying events (20 CFR 418.1205): ✅
| # | Event | Typical retiree relevance |
|---|---|---|
| 1 | Work stoppage | The big one — you retired |
| 2 | Work reduction | Went part-time / phased retirement |
| 3 | Marriage | Changes filing status and thresholds |
| 4 | Divorce or annulment | Same |
| 5 | Death of spouse | The widow(er) cliff (§4) |
| 6 | Loss of income-producing property | Casualty, disaster, fraud — not a market decline or a voluntary sale |
| 7 | Loss or reduction of pension income | Plan termination, PBGC takeover |
| 8 | Employer settlement payment | From employer bankruptcy/reorganization |
What does NOT qualify — the single most common misunderstanding: ✅
- a Roth conversion
- a large capital gain or business sale
- a home sale above the §121 exclusion
- an RMD or a big IRA withdrawal
- an inheritance or IRD
- investment losses or a market drop
These are income spikes, not life changes. SSA will deny the appeal. The consolation: a spike falls off by itself two years later, because the next premium year reads a different return.
Mechanics:
- File SSA-44 with evidence — a retirement letter, a final pay stub, a death certificate, a pension-termination notice. Estimate the current year’s MAGI on the form.
- You can file proactively. Don’t wait for the bill — submit at Medicare enrollment if you’re retiring, or as soon as you get the predetermination/initial-determination notice. You generally have 60 days from the determination notice.
- If denied: request reconsideration on Form SSA-561-U2 within 60 days, then an ALJ hearing if needed.
- File every year the event still governs. SSA-44 fixes the year you file for. A 2026 retiree may need one for 2026 and 2027 before the lookback naturally catches up.
- Overpayments are refunded once corrected — but the cash-flow gap is real, so file early.
6. Bracket Management: The Age-63 Wall
Section titled “6. Bracket Management: The Age-63 Wall”If you turn 65 in year X, your first Medicare premium year is X — priced on your MAGI from year X−2, i.e., the year you turned 63. The planning window closes earlier than almost anyone expects.
| Your age in the income year | Premium year it prices | What it constrains |
|---|---|---|
| 63 | 65 | First Medicare year |
| 64 | 66 | — |
| 65 | 67 | — |
| 71 | 73 | First RMD year — RMDs stack on top |
The strategic shape of a retirement:
- Ages 60–62 — the free window. Pre-lookback. Nothing you convert here touches IRMAA. If you’re also off the ACA exchange (or accepting the subsidy cost, HC-01 §5), this is the cheapest conversion airspace you will ever have. Convert hard here.
- Ages 63–64 — the wall. Conversions now carry an IRMAA toll two years out. Size deliberately: either stay under a threshold with buffer, or cross decisively and fill the tier.
- Ages 65–72 — managed. Each year’s MAGI prices a premium year. Keep a running target; check December.
- Age 73/75+ — RMDs arrive and are non-optional MAGI (TX-04). Whatever you failed to convert now sets a permanent floor under your IRMAA tier. This is the cost of under-converting earlier.
Buffer discipline. Leave $3,000–$5,000 of headroom below your target threshold. Late-arriving 1099s, mutual-fund capital-gain distributions in December, and reclassified dividends routinely add a few thousand dollars after you thought you were done. The cliff punishes precision failures, not bad strategy.
Sequencing note. IRMAA thresholds index annually to CPI-U (except the top tier), so next year’s lines are slightly higher than this year’s — but they are not published until roughly November of the prior year. When planning a conversion two years ahead, use the current thresholds minus your buffer and treat indexing as upside, not budget. ◻️
7. Three Scenarios
Section titled “7. Three Scenarios”A. The new retiree over-charged by the lookback (65, single, retired at 64). 2024 MAGI was $240,000 — peak earnings. Filing single, that sits in the fourth tier ($205,000–<$500,000), so her 2026 premiums land at $529.60/mo of surcharge — $6,355.20 for the year on top of the standard $202.90. But she now lives on $70,000. She files SSA-44 citing work stoppage, estimating 2026 MAGI at $70,000, attaching her retirement letter. SSA re-rates her to the standard $202.90 and refunds the overcharge. She files again for 2027, since her 2025 return still shows a partial-year high income. Cost of not knowing about SSA-44: $6,355.20 — for a form that takes 20 minutes. Note how much harder the single thresholds bite: the same $240,000 as a couple’s joint MAGI would have been tier 1, at $1,148.40 per person.
B. The over-optimizer (couple, 63, $1.9M traditional IRA). Their advisor’s spreadsheet says “avoid IRMAA,” so they cap conversions at $217,000 of MAGI every year. Result: they convert ~$60,000/year less than bracket-filling would allow, and at 75 their RMDs are large enough to put them in tier 2 permanently — plus the survivor faces §4’s cliff. Running the math the other way: accepting tier 1 ($2,296.80/year) to convert an extra $60,000 annually at 24% costs 3.8 percentage points of extra effective rate, and drains the IRA years faster. They were optimizing the small number against the big one. → TX-02
C. The one-time spike (68, sells a rental for a $310,000 gain). 2026 MAGI: $395,000. In 2028 he’ll pay tier 4 — $529.60/mo, $6,355.20 for that year. He calls SSA to appeal and is told, correctly, that a property sale is not a life-changing event. It isn’t appealable. What he can do: recognize that the toll is one year only, verify he isn’t also tripping the NIIT and capital-gains-rate thresholds in the same year (TX-01), and — had he planned ahead — consider an installment sale to split the gain across two tax years, or pair the sale with a low-MAGI year. The lesson is timing, not appeals.
💡 Pro-Tips
Section titled “💡 Pro-Tips”- File SSA-44 in your first retirement year, always. Work stoppage is a qualifying event and the two-year lookback is guaranteed to over-charge you. It is the highest hourly-rate paperwork in retirement.
- Keep a $3,000–$5,000 buffer below every threshold. December fund distributions are the classic overshoot.
- Once you decide to cross a tier, fill it. The toll is identical whether you’re $1 over or $27,000 over — so don’t cross for a trivial conversion, and don’t stop halfway.
- Convert hardest at 60–62, before the lookback window opens. That airspace never comes back.
- Model the survivor. Run every conversion plan against single brackets at half the thresholds — the widow(er) cliff is the real long-run cost of under-converting.
- Remember muni interest counts. Tax-exempt bonds reduce your tax bill and not your IRMAA MAGI.
- Check your determination notice against your own return. SSA sometimes prices off a three-year-old return when the two-year one is late — appealable and correctable.
- Don’t let IRMAA veto a good conversion. Price it: one tier for a couple is $2,296.80–$13,872 for one year, against decades of RMDs and a survivor’s compressed brackets.
⚠️ Common Pitfalls
Section titled “⚠️ Common Pitfalls”- Assuming IRMAA phases in. It doesn’t — one dollar buys the whole tier.
- Forgetting it’s per person. Every published table is per-beneficiary; couples pay double.
- Trying to appeal a Roth conversion, capital gain, or home sale. Not life-changing events; the appeal will be denied.
- Never filing SSA-44 after retiring and silently overpaying for two years.
- Missing the 60-day window on the determination notice, or failing to re-file for the second affected year.
- Planning conversions at 63–64 as if IRMAA doesn’t exist — that’s the first lookback year for your age-65 premiums.
- Ignoring the widow(er) cliff — single thresholds are half, so a survivor can jump tiers on falling income.
- Assuming the top bracket indexes. It’s frozen until 2028; inflation is pushing people into it.
- Expecting hold-harmless to cap your increase. It does not apply to IRMAA payers.
- Cutting conversions to the bone to dodge a one-year toll, then meeting far larger RMDs at 73–75.
- Overlooking muni interest and the taxable portion of Social Security when totalling MAGI.
✅ Actionable Checklist
Section titled “✅ Actionable Checklist”Ages 60–62 (the free window)
- Convert aggressively — pre-lookback MAGI carries no IRMAA cost
- Coordinate against ACA subsidies if still on the exchange (HC-01 §5)
Age 63 — the wall goes up
- Recognize this year prices your age-65 premiums
- Choose a target threshold and a $3,000–$5,000 buffer
- Model the couple and survivor outcomes (TX-02)
At Medicare enrollment
- Read the IRMAA predetermination/initial-determination notice and verify the tax year used
- If retired or income dropped: file SSA-44 with evidence — don’t wait for a bill
- Calendar the 60-day appeal deadline; note Form SSA-561-U2 if denied
Every year on Medicare
- October: estimate fund distributions before sizing December conversions
- December: confirm MAGI against your target line with buffer intact
- Re-file SSA-44 if the life-changing event still governs the next premium year
- Recheck brackets when CMS publishes November figures (FN-02)
Ages 71+
- Project RMDs against IRMAA tiers (TX-04) — unconverted balances set a permanent floor
- Consider QCDs at 70½+ to reduce MAGI directly
State notes (→ ST-01, ST-04): IRMAA itself is federal and uniform — the same thresholds and premiums apply in all 50 states, and no state can waive or alter them. What varies is the cost of the income that triggers it. A Roth conversion sized to fill an IRMAA tier is free of state income tax in Washington, Florida, Texas, Nevada, Tennessee, and the other no-income-tax states, but stacks a state marginal rate on top of the federal toll in high-tax states — the same $40,000 conversion in §3 can cost several thousand dollars more in CA or NY, changing where the break-even sits. ✅ Washington adds a further wrinkle: no income tax on conversions, but a 7% capital-gains excise tax above ~$278k that can raise the cost of harvesting gains as an IRMAA-managed alternative ◻️ (FN-02). Separately, state Medicare Savings Programs and State Pharmaceutical Assistance Programs (SPAPs) help with Part B/D costs at the low-income end — they don’t reach IRMAA payers, but they matter for the same households in later, lower-income years. Model your state’s rate on the conversion, not just the federal toll.
Sources & further reading (verified Aug 2026)
Section titled “Sources & further reading (verified Aug 2026)”- CMS 2026 Medicare Parts A & B premiums and deductibles release (Part B $202.90; IRMAA tiers to $689.90)
- CMS 2026 Part D national base beneficiary premium ($38.99) and Part D IRMAA amounts ($14.50–$91.00)
- SSA — Form SSA-44, Medicare Income-Related Monthly Adjustment Amount — Life-Changing Event, and its instructions; Form SSA-561-U2 (reconsideration)
- 20 CFR 418.1205 (the eight qualifying life-changing events); 20 CFR 418.1201 (IRMAA MAGI definition)
- Bipartisan Budget Act of 2018 §53114 (top bracket at $500,000/$750,000, indexing barred until 2028)
- Medicare.gov — Monthly premium for drug plans (Part D IRMAA billing) and Part B costs
- FN-02 Key Numbers 2026 — the canonical bracket table for this wiki
Not advice. Educational reference only. Decisions with real money should be confirmed against primary sources — IRS publications, SSA.gov, Healthcare.gov, CMS — or a fee-only CFP/CPA.
Dollar figures, thresholds, and brackets are stated for the plan year named in each article’s header, and tax and healthcare rules change annually. Check theLast verified date at the top of the page before relying on a number.