Illinois, Kentucky, New Jersey, Mississippi and Connecticut carry the biggest pension funding gaps — and it does show up in bond pricing, though rating trajectory matters as much as the raw shortfall

Summary

Illinois, Kentucky, New Jersey, Mississippi and Connecticut carry the deepest pension funding gaps in the country, all below 60% funded as of the end of FY2024 (Reason Foundation's Annual Pension Solvency Report), against a national median funded ratio of 79%. Illinois is the clear outlier on every measure: worst funded ratio in the nation (52%), and by far the largest unfunded liability per resident at $15,804 — more than 50% higher than the next-worst state (Connecticut, $10,151). By raw dollar amount, California ($265B) and Illinois ($201B) top the list, but California's ratio is far healthier once its much larger economy and pension-fund base are accounted for. Three states — Tennessee (104%), Washington (102%) and South Dakota (100%) — reported no aggregate unfunded pension liability at all at the end of FY2024.

Yes, this gap shows up in bond pricing — but the relationship runs through credit-rating trajectory and fiscal management more than through the static size of the shortfall. Chicago's own general-obligation bonds have recently traded 155-175 basis points over the AAA municipal benchmark, widening rapidly through September 2026 (up from 124 bps just over a week earlier) as market participants say pricing is running ahead of what the rating agencies have formally acted on. Yet at the state level, Illinois — still explicitly called "the lowest-rated state government" by S&P even after its 2026 upgrades — has now had 11-12 rating upgrades since 2021 specifically because it strung together eight consecutive balanced budgets and enacted (if only partial) pension-liability-reduction measures, not because its funded ratio actually improved much. New Jersey's KBRA upgrade to AA- in September 2026 tells the identical story: the rating agency explicitly still calls the state's unfunded pension and retiree health-benefit obligations "very high," but rewarded it anyway for six consecutive years of paying its full actuarially required pension contribution. Bond markets price the direction the state is moving as much as the depth of the hole itself.

1. Which states have the biggest gap between promised and funded pensions

Reason Foundation's Annual Pension Solvency Report, drawn directly from states' own audited financial statements for the fiscal year ending 2024 (the most recent year with complete data across all states), gives three distinct rankings depending on what "biggest gap" means:

• Relative gap (funded ratio: cents on the dollar of promises actually funded) — Illinois is worst in the nation at 52% funded. The other states under 60% are Kentucky (54%), New Jersey (55%), Mississippi (56%), and Connecticut (59.5%). The national median across all state and local plans was 79% funded at the end of FY2024 — meaning governments nationally have about 79 cents saved for every dollar of pension benefit already promised.
• Absolute dollar gap — California ($265B) and Illinois ($201B) carry by far the largest unfunded liabilities in raw dollars, followed by Texas ($92.2B) and New Jersey ($92B), then Pennsylvania ($67B) and Ohio ($61B). This ranking mechanically favors large, populous states; California's funded ratio is materially better than Illinois's despite the larger dollar total.
• Per-capita gap — the cleanest "promise vs. delivery" burden measure, since it normalizes for state size. Illinois again leads at $15,804 per resident, with Connecticut second ($10,151), followed by Alaska ($9,990), Hawaii ($9,784), New Jersey ($9,688), Mississippi ($9,033), New Mexico ($8,641) and Kentucky ($8,626).

Illinois is the consistent outlier across all three cuts — worst funded ratio, second-largest dollar total, and by a wide margin the worst per-capita burden. Tennessee, Washington and South Dakota are the only three states reporting zero aggregate unfunded pension liability at all.

2. Does it show up in bond pricing

Yes — both in spreads and in ratings — but rating trajectory and fiscal management dominate the raw funding-ratio number.

Spreads: Chicago's general-obligation debt (layered on top of the state's own pension pressure) has recently traded at a widening spread over the AAA municipal benchmark: a GO bond maturing January 2036 carried an evaluated yield of 5.583% as of September 25, 2026 — a spread of roughly 155 basis points to AAA, according to Municipal Market Analytics data reported by Bond Buyer, up from 148 bps on September 17 and 124 bps on September 16. A separately cited 10-year GO spread reached 175 bps, described by one market participant as higher than the average sub-investment-grade bond in the Bloomberg high-yield municipal index, even though Chicago's own ratings (BBB+/BBB+/BBB/Baa3 from KBRA/Fitch/S&P/Moody's) remain investment grade. Analysts quoted directly attributed this to the market pricing in a possible downgrade "before any of the rating agencies have taken that step," citing Chicago's underfunded pensions, structural budget imbalances and reliance on refinancing gimmicks as the underlying credit concerns — alongside a general widening in BBB-rated municipal spreads across the market that is not specific to Chicago.

Ratings: At the state level, Illinois remains — in S&P's own words as of September 2026 — "the lowest-rated state government," citing "elevated pension liabilities and chronically underfunded pensions" by name as the reason. Yet Illinois has been upgraded 11 times since 2021 (Moody's to A1 from A2 in August 2026; S&P to A from A-minus in September 2026), specifically because of eight consecutive on-time balanced budgets, a rebuilt $2.53B rainy-day fund, and the governor's incremental pension-liability-reduction measures — improvements in fiscal management and trajectory, not in the actual funded ratio, which both rating agencies still flag as the state's central structural weakness. New Jersey's KBRA upgrade to AA- on September 2, 2026 followed the same pattern almost exactly: KBRA cited six consecutive years of the state paying its full actuarially required pension contribution as a key driver, while explicitly still warning that New Jersey's unfunded pension and retiree health-benefit obligations remain "very high relative to residents' income and the state economy."

The upshot: the size of the pension gap is priced into both credit ratings (it is cited by name as the binding constraint keeping Illinois in last place among states, and New Jersey's obligations are still flagged as a risk despite the upgrade) and into market spreads (Chicago's widening GO spread, tied explicitly by analysts to legacy pension underfunding among other structural issues). But the rate of change — whether a state is making its full required contribution and running balanced budgets — moves ratings and, with a lag, spreads, even while the absolute funding gap stays enormous. A state can be simultaneously the worst-funded in the country and one of the year's biggest rating-upgrade stories, which is exactly what happened to both Illinois and New Jersey in 2026.

Data and methodology note

The askamerica/govdata connector was searched first (search_catalog for "pension," "liabilities," "state pension") and returned zero matching tables or columns across its roughly 26 topic schemas — there is no pension-liability table and no municipal-bond-pricing/yield-spread table in this corpus. Every figure in this report is instead drawn from primary sources read directly: Reason Foundation's own published Annual Pension Solvency Report page (fetched and read in full, not a search snippet) for every funded-ratio and unfunded-liability figure, and Bond Buyer's own news articles (fetched and read in full) for every spread and rating-action figure, cross-referenced against the rating agencies' own stated rationale as quoted in that reporting. This is disclosed as an external-source-based answer rather than presented as connector-derived; no askamerica query, statistical tool, or table underlies any number above, so no claims-validation scoring was run — this is a research synthesis, not a fact-check of a third-party claim.