Comparing State Fiscal Outlooks

Scope and What Is Deliberately Left Out

This page uses three headline measures: the audited pension and retiree health care liabilities scaled to personal income, the three agencies’ ratings, and the NASBO rainy-day fund balances. It follows the metro page’s structure, with two changes. The composite rating (average notches across agencies) was dropped: the state ratings are so compressed (46 of 50 within three notches of AAA) that an average adds little, and the weakest single rating is used instead. Personal income was added as a second denominator because it is what the agencies use and because it is available for states but not cities. Plan-level pension funded ratios were considered as a way around the teachers’-plan inconsistency described below and omitted: states run between one and a dozen plans with different measurement dates, discount rates and cost-sharing arrangements, and no single funded ratio per state can be built from primary sources on a comparable basis without a great deal of judgment. General Fund reserves (committed, assigned and unassigned balance as a share of expenditures) are carried in the workbook only; at the state level the General Fund is so differently defined (North Dakota’s balance is six times its spending because most state activity runs through other funds; Louisiana’s taxes bypass it entirely) that the measure is not a ranking. Revenue mix, capital-gains dependence, Medicaid exposure and muni bond spreads are real drivers of state fiscal health and are listed under What to Watch For; each would need its own primary-source pull.

What the ACFR Pension Figure Is and Is Not

Under GASB 68 a state reports its own proportionate share of each plan’s net pension liability as an employer, plus any share it carries as a “nonemployer contributing entity” for other governments’ employees, which in practice means the teachers’ plan. Whether the state carries the teachers’ share depends on state law, not on how well funded the plan is. Connecticut, Kentucky, Maryland, Texas, Illinois, Rhode Island, California, Alaska, Maine and Tennessee are among the states that carry all or part of it; Missouri, Michigan, Georgia, Ohio and Virginia carry none because school districts are the employers, and Kansas carries only the state’s own share of KPERS. The figure on this page is therefore the state government’s bill, which is the right number for a state’s own budget and credit, but it understates the pension problem in the states where the districts carry the teachers. The workbook notes each state’s treatment.

Balance-Sheet Measures

Liabilities: net pension liability plus net OPEB liability of the primary government (governmental plus business-type activities), as reported on the government-wide Statement of Net Position or, for the 19 states that aggregate them into long-term liabilities (Alabama, Connecticut, Florida, Idaho, Illinois, Indiana, Kansas, Kentucky, Maryland, Michigan, Missouri, North Carolina, Ohio, South Carolina, Tennessee, Utah, Virginia, Washington, Wisconsin), in the long-term obligations note. Where a statement splits the liability into current and noncurrent lines (Texas, Colorado, New Hampshire and others) the two are summed. Pay-as-you-go retiree health plans report a “total OPEB liability”, used as reported. Pension and OPEB assets are not netted; four states (Alaska, South Dakota, Utah, West Virginia) report a net OPEB asset and no OPEB liability and are shown at zero for that component, and South Dakota also reports a net pension asset. North Carolina’s pension figure includes $341 million of GASB 73 separation-allowance liability. Michigan’s note is printed in millions and is precise to $50,000. Unrestricted net position is the governmental-activities figure. All figures were converted to dollars from the printed unit (thousands, millions or dollars; Nevada prints the government-wide statement in thousands and the fund statements in dollars).

Extraction and Verification

The 50 ACFR PDFs were converted to text and the three statements located by heading; Arkansas’s report has no text layer and its statement pages were read by OCR at 300 dpi. Every General Fund figure was checked by footing (fund-balance components sum to the printed total; all 50 tie to zero) and the pension and OPEB figures against the MD&A condensed tables and the pension notes where available. The PDF page of each statement is recorded in the scorecard table and the workbook, and the verbatim lines are on the workbook’s ACFR Source Lines tab.

Ratings

Each agency’s long-term scale is mapped to notches below AAA (Aaa/AAA = 0, Aa1/AA+ = 1, Aa2/AA = 2, Aa3/AA− = 3, A1/A+ = 4, A2/A = 5, A3/A− = 6). The ranking uses the weakest rating among the agencies that rate the state. Outlooks are recorded and shown on the chart but not scored. The agencies’ own sites do not permit automated retrieval, so every rating was taken from a state document, a governor or treasurer release, or The Bond Buyer, with the source URL in the table; cells that could not be confirmed against a dated 2025–26 source are flagged rather than guessed.

Vintages and Known Gaps

Ratings are as of September 2, 2026. Balance-sheet figures are fiscal 2025 (fiscal years ending March 31 to September 30, 2025), except Arizona, Mississippi, Nevada and Oklahoma at fiscal 2024. Personal income is BEA’s April 2026 SAINC1 release and will be revised September 30, 2026. Rainy-day balances are NASBO’s Spring 2026 survey; Pew’s days-of-cover figures, computed from an earlier survey vintage, differ slightly and are carried in the workbook for reference. Open items are listed on the workbook’s Open Items tab.