[ST-02] "Retirement-Friendly States," Honestly

The rankings price the wrong household at the wrong age; here is the cost sheet that prices yours

Pillar: State Dimension · Applies to: Anyone comparing states — especially pre-Medicare early retirees and households in the $2–10M estate band Last verified: August 2026 · Refresh cadence: Annual for figures (via ST-04 tables); the framework is evergreen Related: ST-01 Six State Dimensions · ST-03 Changing Domicile · ST-04 Quick-Reference Tables · HC-01 ACA Bridge · HC-02 Medicare 101 · EP-05 Estate & Inheritance Taxes · HC-08 Long-Term Care

Not advice. This page teaches a method and demonstrates it on invented households. Every input is state- and county-specific and most of them change annually. Price your own numbers against ST-04 and the state’s own agencies before moving on the strength of any figure here.

---

TL;DR

• Published “retirement-friendly state” rankings compute state and local taxes as a share of income for the average resident — a working-age wage earner. That household shares almost no income lines with a retiree, so the ranking answers a question you did not ask.
• The single largest correction is age. Every published ACA benchmark premium table — KFF’s included — is quoted for a 40-year-old. On the federal default age curve a 62-year-old’s factor is 2.873 vs. 1.278, so a 62-year-old pays 2.25× the number in the table. ✅
• Age-rating the table reverses the ranking. Vermont is the most expensive benchmark state at 40 ($1,299/mo) and only 6th of 11 for a 62-year-old couple — $27,633/yr cheaper than Wyoming — because Vermont and New York are the two states that prohibit age rating entirely. ✅ New York goes from mid-pack to cheapest of the eleven.
• Below the 400% FPL cliff, the premium line is identical in every state. Your required contribution is a percentage of your MAGI, so the credit absorbs 100% of the state difference: $8,366/yr in Florida, Washington, Pennsylvania, and Vermont alike. The premium column in every ranking is noise until you cross the cliff.
• The § below prices one couple across four states. Florida wins and Pennsylvania — which taxes no retirement income at all — finishes last, because a 4.5% inheritance tax and a 1.26% property rate outweigh an income tax it never charges.

---

1. What the rankings actually measure

Nearly every “most tax-friendly states for retirees” list is built from one of two ingredients, and both are the wrong instrument.

The tax-burden ratio. Aggregate state and local taxes collected, divided by aggregate state personal income. It is a real statistic and a useful one — for legislators. It describes the average resident, who is working, earning wages, and paying income tax on nearly every dollar. It tells you almost nothing about a household living on Social Security, qualified dividends, and IRA withdrawals.

The composite score. Income tax rate, plus sales tax, plus property tax, plus a “does it tax Social Security” flag, weighted by editorial judgement. Better, but it still fails on four axes at once.

Mismatch	What the ranking assumes	What a retiree actually faces
Household	Wage income, taxed at the headline marginal rate	SS exempt in 42 states + DC ✅; IRA withdrawals exempt in PA/IL/MS ✅; LTCG possibly taxed at 0% federally (TX-01)
Age	Premium tables quoted for a 40-year-old	A 62-year-old is rated at 2.25× that figure ✅
Geography	One number per state	ACA benchmark, property tax, and homeowners insurance are set at the county level
Horizon	This year’s tax bill	A 30-year plan with three phases and a one-time bill at death that no ranking scores

The fourth is the one that costs the most. No mainstream ranking prices the state estate or inheritance tax, Medicaid long-term-care rules, or Medigap switching rights — the three lines most likely to produce a six-figure number. ST-01 catalogues those dimensions; this page converts them to dollars.

The counterintuitive thing, stated early: for a subsidized pre-Medicare retiree, a state with expensive health insurance is not worse. It is frequently better. §3 does the arithmetic.

2. The age-curve correction

Insurers in most states rate individual-market premiums off the federal default standard age curve, capped at a 3:1 ratio between the oldest and youngest adults. The published factors: ✅

Age	21	40	50	55	60	62	64+
Factor	1.000	1.278	1.786	2.230	2.714	2.873	3.000

KFF’s state benchmark table — the source nearly every ranking reuses — is quoted for a 40-year-old. So the conversion for a 62-year-old is 2.873 ÷ 1.278 = 2.248×, and for a couple, twice that again.

Three states break the curve, and it changes their ranking completely. Vermont and New York prohibit age rating outright — a 1:1 ratio, the same premium at 25 and 64 ✅. Massachusetts compresses its curve (2.365 at age 60 against the default 2.714) ◻️, and DC runs its own (2.099 at 60) ◻️.

Applying this to the 2026 benchmark averages ✅ (second-lowest silver, county-weighted, national average $625/mo at 40):

State	KFF table (age 40)	Age-62 couple, per year
Wyoming	$1,090	$58,809
West Virginia	$1,073	$57,892
Alaska	$1,032	$55,680
Florida	$683	$36,850
Washington	$612	$33,019
Vermont	$1,299	$31,176	no age rating
Pennsylvania	$572	$30,861
Virginia	$455	$24,549
Maryland	$414	$22,337
New Hampshire	$401	$21,635
New York	$817	$19,608	no age rating

Vermont is the most expensive state in the published table and sixth of eleven once you age-rate it — $27,633/yr cheaper than Wyoming. New York is fifth in the table and cheapest on this list for a 62-year-old couple. Both inversions are created entirely by a rating rule that the ranking never mentions, and both run in the direction that penalises anyone who trusted the ranking.

3. Below the cliff, the premium line is free

Now the correction that matters more than the age curve.

Your premium tax credit is set so that you pay a fixed percentage of your own MAGI toward the benchmark plan; the credit is the remainder (HC-01 §3). The benchmark premium enters the formula only on the credit side. So for a household under the 400% FPL cliff:

Dana & Marcus, both 62, MFJ, no dependents. Bridge-year income: $60,000 IRA withdrawal + $18,000 realized LTCG + $6,000 dividends = $84,000 MAGI — 397% FPL against the 2026 household-of-2 cliff of $84,600 ✅ (FN-02), with $600 of headroom. Applicable percentage in the 300–400% band: 9.96% flat ✅ (Rev. Proc. 2025-25).

Required contribution = $84,000 × 9.96% = $8,366/yr — in every state.

	Florida	Washington	Pennsylvania	Vermont
Benchmark, age-62 couple	$36,850	$33,019	$30,861	$31,176
Their required contribution	$8,366	$8,366	$8,366	$8,366
Premium tax credit	$28,484	$24,653	$22,495	$22,810

The state’s premium level does not change what they pay. It changes what the government pays. A $6,000 swing in benchmark premium moves the credit, not the bill.

And the credit is portable across metal tiers: it can be applied to any plan except catastrophic, and if a Bronze plan costs less than the credit, the surplus is real money — a bigger benchmark buys a bigger credit to spend on a cheaper plan (HC-01 §2). Which is why the ranking’s premium column, for a subsidized retiree, is not merely noise but inverted.

Above the cliff the sign flips violently. At $86,000 of MAGI — $1,400 over — the credit is $0 and they pay full freight:

	Florida	Washington	Pennsylvania	Vermont
Cost of crossing the cliff	$28,484	$24,653	$22,495	$22,810

$1,400 of extra MAGI costs $22,495–$28,484 depending on the state. That is the real state healthcare variable: not the premium, but how deep your state’s cliff is. Florida’s is 27% deeper than Pennsylvania’s.

4. The five-line cost sheet

Replace the ranking with five lines, priced on your household, split by phase — because the lines that dominate change as you age.

#	Line	Bridge (pre-65)	Medicare (65+)	At death
1	State income tax on your actual plan	●●●	●●●	—
2	Healthcare	●○○ subsidized / ●●● over cliff	●●○ Medigap	—
3	Property tax + insurance	●●●	●●●	—
4	Sales & excise	●○○	●○○	—
5	Estate / inheritance tax	—	—	●●●●

How big can each line actually get? This is the part worth memorising, because it tells you what to compute carefully and what to skip:

Line	Realistic annual swing, best state → worst
ACA premium, subsidized	$0 — identical everywhere ✅
ACA premium, over the cliff (62-yo couple)	NH $21,635 → WY $58,809 = $37,174/yr ✅
Property tax, $600k home	HI 0.29% $1,740 → NJ/IL 1.88% $11,280 = $9,540/yr ◻️
State income tax	$0 → roughly $8,000/yr at this income ◻️
Medigap Plan G, couple	ND ~$2,928 → NY ~$8,496 = $5,568/yr ◻️
Sales tax, ~$45k taxable consumption, 3-point spread	~$1,350/yr ◻️
Estate / inheritance	$0 → $162,000, once ✅ (see §5)

Two readings. First, the sales-tax column that rankings weight is the smallest line on the board — roughly a tenth of the property-tax swing. Second, the healthcare line is either the largest line or literally zero, depending on one binary you control: which side of the cliff you land on. Manage the cliff before you shop for states.

A note on Medigap. New York is the most expensive Medigap market in the country ◻️ and one of the few with year-round guaranteed issue (ST-04 §5). Rankings score the protection as a pro and the premium as a con without noticing they are the same fact: guaranteed issue with no underwriting is a benefit that gets priced. Decide whether you want the protection; do not expect it free.

5. Worked example: one couple, four states

Dana & Marcus from §3. Both 62. They will buy a $600,000 home, spend ~$95,000/yr, hold MAGI at $84,000 for three bridge years, go on Medicare at 65, and leave a projected $3.6M estate to two adult children. Horizon to the second death: 28 years.

The four candidates are deliberately chosen: two states every ranking calls friendly (Florida, Washington — no income tax), one that taxes wages but exempts all retirement income (Pennsylvania), and one every ranking calls unfriendly (Vermont — graduated to 8.75%, and one of the eight states still taxing Social Security ✅).

The death line first, because it is the biggest number and the one no ranking shows:

	Amount	Why
Florida	$0	No estate or inheritance tax ✅
Vermont	$0	$5,000,000 exemption; the estate is below it ✅
Washington	$60,000	$600,000 over the $3,000,000 exclusion at the 10% first bracket ✅ — and only because there is no state portability; credit-shelter planning takes it to $0 (EP-05)
Pennsylvania	$162,000	4.5% inheritance tax on the whole $3.6M to adult children, with no exemption threshold ✅

The state with no income tax on retirement income has the largest death bill on the board, by 2.7×.

The full sheet, annualising the death line over 28 years:

Line	Florida	Washington	Pennsylvania	Vermont
1 State income tax	$0 ✅	$0 ✅	$737 ✅	~$2,317 ◻️
2 ACA net premium	$8,366	$8,366	$8,366	$8,366
3 Property tax, $600k home	$4,680 ◻️	$5,520 ◻️	$7,560 ◻️	$9,060 ◻️
5 Death line ÷ 28	$0	$2,143	$5,786	$0
Total, bridge year	$13,046	$16,029	$22,449	$19,743

Line 1 detail: Pennsylvania exempts IRA distributions after 59½ entirely ✅ but taxes interest, dividends, and capital gains at 3.07% ✅ — so their bill is 3.07% × $24,000 of investment income, not 3.07% × $84,000. Vermont’s figure applies its 3.35% first band to MAGI less an approximate $14,850 standard deduction ◻️ — verify the current schedule with the Vermont Department of Taxes.

Cost order: Florida ($13,046) < Washington ($16,029) < Vermont ($19,743) < Pennsylvania ($22,449).

Pennsylvania is the result worth sitting with. It charges this couple $737 of income tax — less than 4% of its total — and still finishes last, because a 4.5% inheritance tax and a 1.26% property rate are lines the “no tax on retirement income” headline does not cover. Vermont, the state every ranking places near the bottom, beats it by $2,706/yr.

The same four states, $1,400 over the cliff

	Washington	Florida	Vermont	Pennsylvania
Total, bridge year	$40,682	$41,530	$42,553	$44,944

The order changes — and more importantly, the spread collapses from 1.72× to 1.10×. Once you are over the cliff, every state is roughly equally bad, because a $22,000–$28,000 healthcare bill swamps every other line. The cliff decision is worth more than the state decision, and it is entirely within their control (HC-01, TX-03).

When the answer flips

Change one input and the ranking moves:

• Estate of $2.5M instead of $3.6M → Washington’s death line goes to $0 and it takes second place outright; Pennsylvania’s inheritance tax scales with the estate and still charges $112,500.
• Renting instead of buying → line 3 disappears, and Vermont and Pennsylvania close most of the gap.
• Past 65, drawing Social Security → Vermont begins taxing benefits ✅ while Florida, Washington, and Pennsylvania do not, so Vermont’s line 1 grows exactly when the ACA line vanishes. The state that wins your bridge is often not the state that wins your Medicare years.

Three Scenarios

A. The cliff-managed early retiree, 58, single, MAGI $52,000. Line 2 is fixed by her MAGI, so the entire premium column is irrelevant to her. Her decision collapses to lines 1 and 3 — and since she rents, effectively to line 1 alone. She should ignore healthcare rankings entirely and compare income-tax schedules on her actual withdrawal plan. The one thing she must check is not a cost at all: whether the state runs a Basic Health Program that would raise her subsidy floor to 200% FPL (HC-03).

B. The over-the-cliff consultant, 61, MAGI $140,000. Line 2 is his largest expense by a wide margin, and he is paying full benchmark. For him the age-rated premium table in §2 is the ranking — a move from Wyoming to New Hampshire is worth $37,174/yr, more than any income-tax difference in the country. He should also price whether cutting MAGI below the cliff via a solo 401(k) beats moving at all ([IN-02]).

C. The $6M couple choosing between Washington and Idaho, both 64. Their bridge is nearly over, so line 2 barely matters. Line 5 decides it: at $6M, Washington’s frozen $3M exclusion with no portability exposes half the estate, while Idaho has no estate tax ✅. Washington’s zero income tax cannot catch up — but credit-shelter planning can, which makes this a document question, not a moving question (EP-05, ST-01 §3).

---

💡 Pro-Tips

1. Age-rate every premium you read. Multiply the published 40-year-old figure by 2.25 for a 62-year-old, 2.12 at 60, 2.35 at 64+. If the source doesn’t say what age it quotes, it quotes 40.
2. Settle which side of the cliff you’re on before comparing states. Below it, delete the premium column from your spreadsheet; above it, it’s the only column that matters.
3. Price the death line first, not last. It is the largest single number in most comparisons and the only one that is effectively irreversible once you have died in the wrong state.
4. Check for an inheritance tax separately from an estate tax. They are different taxes with different payers, and a state can have one, both, or neither — Pennsylvania has no estate tax and the worst death line in §5.
5. Get the county, not the state, for benchmark premiums, property tax, and homeowners insurance. State averages hide spreads that routinely exceed the interstate difference.
6. Re-run the sheet for each phase. A state that wins ages 62–65 on healthcare can lose ages 70–90 on Social Security taxation — ST-04 §1.2 tells you which eight states still tax benefits.
7. Treat guaranteed-issue Medigap as a purchase, not a freebie. If your state protects switching, expect to see it in the premium (ST-04 §5).
8. Don’t buy a state to avoid a tax you can plan around. A credit-shelter trust neutralises the no-portability trap in every estate-tax state; moving to avoid it is the expensive way to solve a drafting problem.

⚠️ Common Pitfalls

1. Reading a benchmark premium table as though it applied to you — it is quoted for a 40-year-old and understates a 62-year-old by more than half.
2. Assuming a cheap-premium state saves a subsidized household money. It doesn’t; it shrinks the credit and can leave less surplus to spend down on a Bronze plan.
3. Treating “no income tax” as “no tax.” Washington’s death line in §5 is $60,000; Florida funds itself through property tax and some of the highest homeowners insurance in the country ◻️.
4. Treating “exempts all retirement income” as the same claim. Pennsylvania exempts IRA distributions and still taxes dividends and gains at 3.07% ✅ — and takes 4.5% of the estate.
5. Comparing this year’s tax bill across a 30-year decision and never pricing the one-time bill at death.
6. Scoring sales tax as a major line. At realistic retiree consumption it is roughly a tenth of the property-tax swing.
7. Using a state average for a county-level cost — benchmark premium, property tax, and insurance are all set below the state level.
8. Forgetting that Vermont and New York don’t age-rate, and therefore screening them out on a 40-year-old’s premium that a 62-year-old will never pay.
9. Moving for the bridge and staying for the Medicare years without re-running the sheet for the phase that lasts three times as long.
10. Deciding on the ranking’s weights rather than your own. The weighting is the answer; a composite score is someone else’s household imposed on yours.

✅ Actionable Checklist

Before you compare anything

• Write down your projected MAGI by phase and mark which side of the 400% FPL cliff each phase lands on (FN-02, HC-01)
• Write down your projected estate at the second death, and who inherits — children, siblings, or non-relatives (ST-04 §2.2)

Build the sheet (one column per candidate state)

• Line 1: run your actual withdrawal/conversion plan through each state’s schedule — not the headline rate (ST-04 §1)
• Line 2: pull your county’s benchmark from the exchange, then age-rate it; if subsidized, enter your required contribution instead
• Line 3: effective property tax on the home you’d actually buy, plus a real homeowners insurance quote
• Line 4: sales tax on realistic taxable consumption — then confirm it’s as small as §4 says before spending more time on it
• Line 5: estate and inheritance tax at the second death, assuming no portability (ST-04 §2, EP-05)

Sanity-check the result

• Re-run the sheet for the Medicare phase, adding Medigap and Social Security taxation
• Flip your two biggest assumptions (estate size, rent vs. buy) and see whether the winner changes
• Confirm nothing in the sheet is doing work that a document could do more cheaply (EP-05)
• Before acting, read ST-03 and sequence income events around the move date

Annually

• Refresh every figure against ST-04; re-verify anything marked ◻️ against the state’s own agency

Sources & further reading (verified August 2026)

• CMS, federal default standard age curve and state-specific variations (State Specific Age Curve Variations, read directly for the 1.278 / 2.714 / 2.873 / 3.000 factors) and Market Rating Reforms for the 3:1 cap and New York’s and Vermont’s 1:1 ratios
• KFF State Health Facts, Marketplace Average Monthly Benchmark Premiums, 2026 — second-lowest silver, 40-year-old, county-weighted by plan selections; US average $625
• IRS Rev. Proc. 2025-25 (2026 applicable percentages); FN-02 for the FPL and cliff figures
• Washington DOR, estate tax exclusion and rate tables ($3,000,000 exclusion, frozen; 10%–20% schedule restored for deaths on or after July 1, 2026 — see ST-04 §2.1)
• Pennsylvania DOR, PA Personal Income Tax Guide — the eight classes of taxable income, and the exemption for IRA distributions after 59½; PA inheritance tax rates by relationship class
• ST-04 for every state list referenced here, with its own refresh date and per-section primary sources; ST-01 for the dimensions this page prices; HC-01 for the subsidy mechanics behind §3
• Property-tax effective rates are aggregator figures (◻️) derived from Census ACS data; confirm against the county assessor for any state you are seriously considering