Washington's budget more than doubled in a decade, outpacing population and inflation

1. Concern (minor) Context is complete

   Report whether all-funds state operating spending also doubled from 2013-15 to 2023-25, since the claim says 'state spending' and the dataset already holds all-funds totals.

   The article supplies the context a reader most needs: the post-recession low base (tested from every start year back to 2005-07), McCleary and the levy swap (K-12 vs non-K-12 growth, OSPI per-student data), government-specific inflation, the income-share counter-argument with its 2005 vs 2013 dependence, caseloads and Medicaid, revenue and new taxes, appropriations vs expenditures, and the NGF basket change. One gap: the claim is about "state spending," but the article measures only the Near General Fund. The cited dataset's ngf-by-biennium table is described as including "all-funds totals by biennium," yet the article never reports whether the all-funds operating budget (which includes federal and dedicated funds) grew similarly. The article does not address this; it is unlikely to reverse the impression, since the NGF is the budget both sides argue over and Gildon's own Gregoire comparison uses NGF-scale figures, so Minor. The dataset's limitation that 2021-23 included about $7.6B of one-time items is not in the article, but it affects only the 2021-23 to 2023-25 step, not the decade comparison. Confidence: Medium.

   About conclusion, evidence:budget-doubled, source:facthub-dataset The finding

2. Not checked Central claims are accurate

   One-sentence summary: the article rates Sen. Gildon's claim that state spending doubled in the past decade Mostly true, finding Near General Fund appropriations up 114.6% (2013-15 to 2023-25), well ahead of population times inflation (52%), but below the 2005-07 share of personal income. Claim-by-claim: 1. Provenance (Gildon said it, acknowledging inflation and population): TRUE. The release reads "Inflation and population growth are factors, but the governor's administration is leaving out the fact that state spending has doubled in the past decade alone," and cites Gregoire's "$33 billion when she left office in 2013 to over $80 billion now." High confidence. 2. NGF $33.90B to $72.74B, +114.6%: UNVERIFIABLE here; both Legislative Budget Notes exceed the 5 MB snapshot limit. Consistent with WPC's independently stated NGFO figures ($31.2B in 2011-13, $80.2B in 2025-27) and with the dataset's saved analysis. 3. NASBO +121.5% FY2014-FY2025: UNVERIFIABLE (PDFs unreadable). 4. Population x CPI +52%, benchmark $51.53B: MOSTLY TRUE as far as checkable; WPC independently computes a $51.6B population-plus-inflation benchmark from 2013-15. BLS snapshot gives 2024 CPI-U halves 312.145/315.233 (annual about 313.7), consistent with the stated 32.5% rise from 2014, but 2014 values are not in the snapshot. 5. Share of income 5.95%/4.78%/5.39%: UNVERIFIABLE (BEA zip too large). 6. Area breakdown, K-12 OSPI per-student, ERFC revenue and tax-law estimates: UNVERIFIABLE (sources unreadable). 7. 2025 package about $9B over four years: TRUE; WRC: "estimated to increase revenues ... (NGFO) by $9.0 billion over four years." 8. WPC benchmark $51.6B and 55%: TRUE, quoted accurately. 9. NPI position: NPI page unreadable (zstd), but WPC's account matches the article (per-$1,000 income measure, $189.98 in 2005 to $170.25 in 2023; "in relative terms" quote). Peripheral discrepancy: the article says Gregoire's last budget (2011-13) "ended at $31.5 billion"; WPC gives $31.2B for 2011-13, likely a different account definition; cannot resolve without the LBN. No readable source contradicts any central claim, and every figure matches the cited dataset's saved analyses, but the dataset is the author's own compilation, and the primary records behind the central figures could not be read. Result NotChecked. Confidence: Low for the central figures; High for provenance and the WPC/WRC items.

3. Pass Claims are supported and labeled

   Every major claim is tied to an evidence item with named sources and a dataset saved analysis, and the figures in the summary, keyPoints, answer and conclusion match the dataset's results exactly (e.g., growth-2013-to-2023: "Final NGF rose 114.6% ($33.90B to $72.74B) ... population x US CPI-U rose 52.0%"; existing-base-counterfactual: "$58.99B ... about $7.46B to spare ... about $13.75B short"). Inferences are labeled as such: "Which framing a reader prefers depends on whether the low-spending years after the recession were a trough or a reset; the data cannot settle that"; the revenue counterfactual is called "an estimate" and "static"; tax-law revenue is given as a range ($7.05-9.25 billion). The conclusion separates what is shown from what is not ("it shows that the tax increases paid for growth beyond population and inflation, not that the growth was unjustified"). No opinion is presented as fact. Confidence: High.

4. Pass Sourcing is strong and transparent

   Central claims rest on primary official records: the Legislature's Legislative Budget Notes (appropriations, with page references such as p. O-16, O-3, O-9), ERFC Table 3.17 (revenue), OFM population, BLS CPI-U series CUUR0000SA0, BEA SAINC1 and the S&L price index B829RG3A086NBEA, OSPI F-196 district reports, and NASBO's Fiscal Survey. Partisan or advocacy sources (Senate Republican Caucus, NPI, WPC, WRC) are used only as claim provenance, the opposing claim, a critic's benchmark being checked, and the size of the 2025 package, which WRC states accurately and which the revenue evidence also ties to LBN 2025 p. O-9. Each evidence item lists its sourceIds, so it is clear which source supports which claim, and the cited dataset records input figures with sources and a build script. Source quality is strong even though several primary files could not be read by this auditor. Confidence: High.

5. Pass Framing is fair and neutral

   The article represents Gildon's claim accurately (quote matches the release, including "Inflation and population growth are factors"); the summary's paraphrase "beyond what inflation and population explain" slightly extends his wording but is a fair reading of "factors, but ... leaving out the fact," and the article tests that harder version. Both sides are held to the same standard: it concedes NPI is "right that the budget is smaller relative to the economy than in 2005" while showing critics "are right that it has grown faster than the economy for the past decade," and it checks numbers from Gregoire/Gildon (different budgets compared) and WPC (method not shown; benchmark nearer 49%) alike. WPC's source confirms the article states its figures accurately. Emphasis follows the claim's decade scope. No loaded language or advocacy. Confidence: High.

6. Pass Uncertainty is stated

   Uncertainty is stated where it exists: the trough-vs-reset question is called unresolvable by the data; tax-law revenue is a range with a confidence column the reader is told to consult; the counterfactual is "static" and ignores behavioral effects; appropriations are not audited expenditures; benchmarks are "a common yardstick, not a measure of what services should cost"; the K-12 figure includes about $0.43B of COVID relief with an ex-COVID figure given; and the conclusion lists "What this check does not establish" (outcomes, legally compelled vs chosen growth, fiscal-note detail). Where evidence conflicts (income share by start year; CPI vs government price index), both are presented and the article explains which applies to the claimed decade. High confidence for the doubling itself is reasonable given two records and a margin well above 100%. Confidence: High.

Coverage: Read the full article payload. Requested all 15 cited sources with get_source_snapshot. Readable: gildon-release (archive, full text), wpc-frost (archive, full text), wrc-tax-2025 (archive, full text), bls-cpi (archive; shows monthly CPI-U 2016-2026 only, no 2014 values). Unreadable: lbn-2015, lbn-2025, nasbo-2015, bea-sainc1 (larger than 5120 KB); nasbo-2026, erfc-sep-2026, ofm-pop (PDF not supported); fred-sl-price (CSV not supported); ospi-f196 (XLSX not supported); npi-post (zstd encoding not decodable); facthub-dataset source entry has no URL. Also read the cited dataset node dataset:wa-general-fund-growth:2 (digest matched) and compared its saved-analysis results, limitations and statement records with the article; the dataset's CSV artifacts themselves were not read. Because the primary appropriation, NASBO, population, income, revenue and OSPI records could not be read, the central figures could not be verified against primary sources, so claims-accurate is NotChecked. Other checks were examined from the article, the readable sources and the dataset.

1. Concern (minor) Rates the real claim

   State whether 'state spending' in the claim means the Near General Fund only, and note whether all-funds operating spending also doubled over the decade.

   Claim: "State spending has doubled in the past decade alone." Method 1 defines it as "state general-fund (Near General Fund) spending roughly doubled over about ten years, by more than population growth and inflation explain." Adding the population/inflation clause follows the editorial note's account that Gildon acknowledged those factors, and it makes the claim harder, not easier, so that is fair. Time scope (2013-15 to 2023-25, last completed biennium) fits "past decade" and the conclusion stays within it. The remaining gap: a reader repeating "state spending" may mean the whole operating budget (all funds, including federal Medicaid money) or total state spending, but the article measures only the Near General Fund and never says why that is the right reading or whether the all-funds figure tells the same story. The context says only that the budget is paid "mostly" from the NGF. This is a small scope mismatch; it would not likely change the reader's belief, so Minor. Confidence: Medium.

   About conclusion, method:1, evidence:budget-doubled The finding

2. Concern (minor) Rating follows from the evidence

   Explain why a 2005-07 income comparison, outside the claim's decade, lowers the label to Mostly true, or rate it True since every past-decade measure supports the claim.

   Before reading the finding I would rate this True or Mostly true. Evidence as stated: appropriations up 114.6% (107.6% enacted-to-enacted), NASBO 121.5%; population times CPI 52%; real per-resident growth 30-41% from 2013-15 and positive from every start year back to 2005-07; and over the decade the budget "has grown faster than personal income (114.6% vs 90.3% from 2013-15)." Every measure within the claimed decade supports the claim. The conclusion demotes to Mostly true because "one reasonable yardstick points the other way": the budget is 5.39% of income vs 5.95% in 2005-07. But that comparison runs from 2005, outside the "past decade" the claim names; within the decade the same yardstick agrees with the claim. A defensible reason for Mostly true exists (the low post-recession base colours the implication), and the article says so, so either adjacent label is defensible; but the stated rationale leans on an out-of-scope comparison. Borderline adjacent-label choice: Minor. The answer and finding do not contradict each other. Confidence: Medium.

   About conclusion, evidence:share-of-income The finding

3. Concern (minor) Evidence is independent

   Stop calling NASBO an 'independent' record in the evidence title and conclusion: its figures are state-reported; and note evidence 3-5 reuse the same LEAP appropriations series.

   The doubling rests on two items: LEAP appropriations (budget-doubled) and NASBO (nasbo-cross-check). NASBO's own limitation says "the underlying numbers still originate with the state," yet its title calls it "An independent national record" and the conclusion says "an independent national record (NASBO)." It does add some independent measurement (General Fund-State expenditures by fiscal year, not NGF appropriations, and unaffected by the NGF basket change), so it is partial rather than no confirmation. Evidence items beats-population-inflation, share-of-income and where-it-went all reuse the same LEAP appropriations series as numerator; they are analyses of one series, not independent lines, though the editorial note counts "seven evidence lines" from "five independent record-keepers." Independent support does remain (OSPI district books for K-12, ERFC revenue), and for a budget-total question the Legislature's own records are authoritative, so the overlap does not undermine the finding. Minor. Confidence: High.

   About evidence:nasbo-cross-check, evidence:budget-doubled, evidence:beats-population-inflation, evidence:share-of-income, evidence:where-it-went, conclusion The finding

4. Pass Confidence matches the evidence

   High confidence rests on the Legislature's own final appropriations (114.6%), confirmed in direction and size by NASBO's fiscal-year general fund figures (121.5%), with a margin well above 100% under both enacted and final comparisons. Key limitations are stated: appropriations vs audited expenditures, the NGF basket change (largest addition about 1.5% of budget), benchmark limits (aging, enrollment), the income-vs-tax-base caveat, and that the revenue counterfactual is static and estimated. The title ("more than doubled in a decade, outpacing population and inflation") and summary match the evidence, and the summary also states the income caveat. The more uncertain parts (tax-law revenue estimates, $7-9 billion range) are presented as estimates with ranges and are not central to the rating. I would also give High confidence for the doubling. Confidence: High.

5. Pass Strongest objections answered

   Strongest objections I would raise: (1) the 2013-15 post-recession trough inflates growth; (2) McCleary court-ordered K-12 funding and the levy swap explain it; (3) government costs outpace CPI; (4) relative to the economy the budget shrank (NPI); (5) caseload/Medicaid growth; (6) the 2023-25 endpoint is a temporary, reserve-funded peak. The article answers 1-5 explicitly with evidence: every start year back to 2005-07, K-12 vs non-K-12 growth (122.5% vs 104.9%) and OSPI per-student data, the S&L price index (34.0% vs 32.5%), the share-of-income series, and HCA/Corrections growing below the benchmark. On (6), the context and evidence note the 2025-27 budget stands at $80.2 billion, higher still, so the endpoint is not a one-off peak. Responses are evidence-based, not dismissive, and concede where the objection is right (NPI against 2005; government prices over 2006-2024). The all-funds scope question is noted under right-claim. Confidence: High.

Coverage: Read the full article payload (claim, finding, answer, confidence, title, summary, keyPoints, context, conclusion, all seven evidence items with methods, institutions, years and limitations, seven objections, eight method steps, editorial note, sources list). Per the Skeptic prompt, no source snapshots or datasets were read; evidence was taken as stated.

1. Concern (minor) Title states the finding

   Make seoTitle state the finding, e.g. "Washington's state budget more than doubled in a decade, fact check", instead of asking a question.

   Title: "Washington's budget more than doubled in a decade, outpacing population and inflation" leads with what is true, names the subject, resolves the claim and is neutral: meets the TITLE-FORM rule. seoTitle: "Did Washington's state budget double in a decade? Fact check" asks the question rather than stating the finding. The rule says seoTitle "may use the wording people search with as long as it states the finding"; a seoTitle that does not state the finding is Minor.

   About conclusion The finding

2. Concern (minor) Verdict fields agree and stand alone

   Name the state in the answer so it stands alone when quoted, e.g. "Washington's general-fund appropriations rose 115% in a decade...".

   Read as: FactHub rates the claim "State spending has doubled in the past decade alone." Mostly true (High confidence): "Appropriations rose 115% in a decade, far beyond population and inflation, though spending is below its 2005-07 share of income." The claim is worded as it circulates and is not its correction. Title, claim, label, answer, summary and conclusion agree (more than doubled; beyond population and inflation; below the 2005-07 income share), with no stale wording. However, neither the claim nor the answer names Washington or says whose appropriations, so the verdict quoted alone does not tell a reader which state or budget it concerns. Rule: "The answer stands alone when quoted and says what is actually true." An accurate answer that does not stand alone is Minor.

   About conclusion The finding

3. Concern (minor) Each section does its job

   Cut the hand-written Independence paragraph from editorialNote (the page generates it) and make the summary one sentence naming the strongest independent evidence.

   editorialNote contains "Independence: Seven evidence lines draw on five independent record-keepers ... using four methods ... All are U.S. sources covering 2005 to 2025". Rule: the independence note is generated by the page and "the author must not write one"; a hand-written independence note is Minor. Also, the summary rule is "one sentence naming the strongest independent lines of evidence"; the summary is two sentences, the first restating the claim ("Sen. Chris Gildon said state spending has doubled..."). The conclusion does its job well: it explains how evidence fits together, weighs the income yardstick, connects to the label, and states what is not established. editorialNote states scope, caveats and external-review status. Context is one paragraph of framing. keyPoints state what the evidence establishes. Worst item: Minor.

   About conclusion The finding

4. Concern (minor) Consistent house style

   Round each figure the same way everywhere (114.6% vs 115%, $33.9 vs $33.90 billion, $72.7 vs $72.74 billion) and spell out LEAP and FTE and explain the McCleary levy swap.

   Rule: "the same rounding for the same figure everywhere." The decade growth appears as 115% (answer, description) and 114.6% (conclusion, keyPoints, evidence); appropriations as $33.9/$72.7 billion (conclusion, keyPoints) and $33.90/$72.74 billion (evidence budget-doubled); the benchmark as $51.5 and $51.53 billion. Values agree, so this is a format inconsistency, not a different figure: Minor. Rule: an abbreviation is spelled out at first use and technical terms are explained: LEAP is never spelled out; "FTE" in evidence k12-per-student is unexplained; "the McCleary levy swap" is used in evidence without explanation. Fiscal years appear as both "FY2025" and "fiscal 2025." No typos or leftover drafting text found.

   About conclusion, evidence:budget-doubled, evidence:k12-per-student, evidence:beats-population-inflation The finding

5. Pass Limitations are specific

   Each limitation is specific to its item: budget-doubled (appropriations not audited expenditures; NGF basket growth, WEIA about 1.5%); nasbo-cross-check (state definitions, GF-S only, fiscal year, nominal); beats-population-inflation (yardstick ignores aging, enrollment, new obligations; mid-year pegging); share-of-income (personal income is not the tax base); where-it-went (agency reorganizations, inseparable components, one-time items); k12-per-student (operating only, $0.43 billion COVID relief, year alignment); revenue-and-taxes (estimated items, static counterfactual). None is boilerplate.

6. Pass Tone is measured and respectful

   Language is measured; "far beyond" and "far exceeded" describe 114.6% vs 52% and are not absolutes like "proves" or "definitively." Objections are answered with figures and concede where the objector is right ("Against 2005, yes"; "The Northwest Progressive Institute is right that..."). No ridicule, sarcasm or contempt for either side; critics' numbers are checked even-handedly (Gregoire, WPC).

7. Pass Readers can verify it

   All 15 sources carry specific role labels ("Primary record of appropriations", "Independent cross-state record", "Claim provenance", "Opposing claim"), full titles, stable identifiers (series ids CUUR0000SA0, B829RG3A086NBEA, SAINC1; table and page numbers such as p. O-16, Table 3.17), archived URLs, and descriptions of why each was chosen. Key figures are anchored to primary records and the cited dataset. Method steps cover claim definition (step 1), evidence not resting on one source (step 2 NASBO cross-check, step 5 OSPI), sources reviewed (steps 2-6), and recorded limits (step 8).

Coverage: Used only the article payload: title, seoTitle, description, claim, finding, answer, confidence, summary, keyPoints, context, conclusion, seven evidence items and limitations, seven objections, eight method steps, editorialNote, chart captions, and 15 sources with role, title and description. No sources or web were consulted, per the Standards prompt.