No — across states, more federal money per capita has not tracked longer life over the last decade; the exception is targeted health transfers like Medicaid expansion

Summary

No. Looking across the 50 states over roughly the last decade, states that receive more total federal money per person have not seen bigger gains in life expectancy — if anything, the simple state-level correlation runs the other way (r = -0.41, p = 0.003). But that finding is heavily confounded: nearly every state's life expectancy is still below its pre-pandemic level, so this decade's "change" is dominated by the COVID-19 mortality shock, and the states receiving the most federal dollars per capita in this data are disproportionately defense-contracting and federal-workforce states (Virginia, Alaska, Connecticut, Maryland), not necessarily places getting more health or welfare aid. When the question is narrowed to a specific, well-identified federal-money channel — Medicaid expansion, which is federally funded at 90%+ — the peer-reviewed literature finds a real, causal, life-saving effect: roughly a 2.5% reduction in mortality among low-income adults in expansion states, at a documented cost of about $5.4 million per life saved. So the honest answer is channel-dependent: undifferentiated federal spending shows no positive association with state life expectancy in this data; targeted federal health-insurance transfers do improve it.

Step 1 — What the literature already says

Three lines of published research bear directly on this question, and they point in different directions depending on what "federal money" means:

• State (not federal) government social spending and mortality. MacKenzie et al. (PLOS ONE 2019) linked 2005 state tax burden and expenditures to county-level mortality from 2006-2015: a 1-point rise in state tax burden was associated with a 5.8% mortality reduction (1.1% after controlling for income/education), and state K-12 education spending specifically was associated with a 4.7% mortality reduction per 10% spending increase. A companion study (J Urban Health 2021) found a 10% rise in STATE government expenditure associated with a 1.4% mortality reduction in American cities (p=0.008), while city-level spending effects differed sharply by race. These are about states taxing and spending their OWN money, not about federal transfers into the state — an important distinction from the question as posed.
• Federal money via Medicaid expansion — the cleanest "federal money into a state" natural experiment. A quasi-experimental study exploiting staggered state adoption timing found expansions raised Medicaid enrollment ~12 points and reduced mortality among low-income adults by 2.5%, costing about $5.4M per life saved / $179,000 per life-year (cited via Evidence-Based Policy review). An earlier NEJM study of three expansion states (Sommers et al., NEJM 2012) found expansion associated with reduced mortality relative to non-expansion neighbors. A national cohort study (Lancet Public Health 2021) found the mortality reduction was highly variable by state — from -63.8 deaths/100k in Delaware to +30.4/100k in New Mexico — meaning even a genuine federal-money effect is not uniform across states.
• Cross-national context. The Institute for New Economic Thinking and Peterson-KFF Health System Tracker both note the U.S. is an outlier: it spends far more on healthcare (public and private combined, federal and state) than peer countries with no corresponding life-expectancy advantage, which is the backdrop against which any single spending-channel result should be read.

Step 2 — Data and method used for the state-level cut

Life expectancy: AskAmerica's corpus has no life-expectancy table, so this analysis pulled CDC/NCHS data directly: the USALEEP state-level life expectancy at birth, 2010-2015 (used as the "~10 years ago" baseline, all 50 states + DC) and a 2023 both-sexes cross-section compiled by World Population Review from CDC/NCHS National Vital Statistics Reports (used as the current point). These are two different vintages/estimation methods bridged as a before/after comparison — the best available given that CDC's more recent state life-table PDFs (NVSR 74-12, 2022) returned HTTP 403 from this environment and no machine-readable state-level series for 2018-2022 or 2023 alone could be fetched directly.

Federal money: fiscal.usaspending_by_state declares a 2017-2025 window, but a live scan found only 2022-2025 actually loaded (2017-2021 are not populated despite the declared range) — so a true 10-year federal-spending panel matching the life-expectancy window does not exist in this corpus. The analysis instead uses the average of FY2022 and FY2023 obligated_amount_excl_loans_excl_cms_admin per capita (loans excluded because they geocode to the lender, not the borrower; CMS-administered awards excluded because they geocode to the Medicare Administrative Contractor's location, e.g. inflating North Dakota/Minnesota/Indiana) — divided by 2023 ACS population (census.acs_population). This is total federal obligations across all award types and agencies (contracts, grants, direct payments) — it is NOT a health- or welfare-specific figure, which matters for interpreting the result below.

Step 3 — What the state-level correlation shows

Across the 50 states (DC and territories excluded), federal obligations per capita (2022-23 avg) correlate negatively with the change in life expectancy from the 2010-2015 baseline to 2023: r = -0.41, p = 0.0035, n = 50. In a regression controlling for each state's 2010-15 baseline life expectancy (to net out simple mean reversion), the federal-$-per-capita coefficient remains negative and significant (coef = -9.6e-05 per dollar, p = 0.002); a leave-one-out check found no single state's removal flips the sign or crosses p=0.05 (max standardized influence 0.94, on Virginia).

Restricting the same bivariate correlation to the 45 states outside the top-5 federal-$-per-capita states (Alaska, Virginia, New Mexico, Connecticut, Maryland — all defense-contracting or federal-workforce-heavy) weakens it to r = -0.24, p = 0.11 — no longer significant. That sensitivity matters: it shows part of the negative association is carried by a handful of states whose federal dollars are mostly defense procurement and federal payroll, not health or welfare transfers, and whose already-high pre-pandemic life expectancy gave them more room to fall during COVID.

46 of 50 states have LOWER life expectancy in 2023 than their 2010-2015 average (mean change: -0.74 years). This is consistent with published national trends: U.S. life expectancy peaked around 78.9 in 2019, collapsed to 76.4-77.1 during 2020-2021, and by 2023 had recovered only to about 78.5 — still below the pre-pandemic peak (CDC/NCHS, Dec. 2024; NCHS Data Brief No. 521). The "decade of decline" this analysis measures for most states is substantially a COVID-mortality artifact, not evidence that federal spending failed to help — or hurt.

Step 4 — Why this is not a causal test, and what would be needed

Three reasons the correlation above cannot be read as "federal money reduces longevity":

• Reverse causality / targeting. Federal transfers (Medicare, Social Security, Medicaid, SNAP) flow disproportionately toward older, poorer, and sicker populations by design — exactly the states and demographics with lower baseline life expectancy. A raw cross-sectional correlation cannot separate "money causes worse outcomes" from "money is targeted at places with worse outcomes to begin with."
• Composition of the spending measure. Total USAspending obligations mix defense contracts, federal R&D, infrastructure grants, and safety-net transfers into one number. A dollar of F-35 subcontracting in Virginia and a dollar of Medicaid expansion funding in Kentucky are both counted as "federal money into a state," but only one has any plausible mechanism to affect resident mortality.
• Window mismatch. The federal-spending figure available (2022-2023, the only years actually loaded in this corpus's usaspending_by_state table despite its declared 2017-2025 window) does not span the same ten years as the life-expectancy comparison (2010-15 to 2023). A state's current federal-dollar level is being compared to a decade-long health outcome change, which is a proxy, not a matched panel.

The Medicaid-expansion literature in Step 1 solves exactly these three problems (it isolates one federal-money channel, exploits staggered adoption timing as a quasi-experiment, and directly targets the intended beneficiary population) — which is why it is the stronger basis for a causal claim than the correlational cut above.

Caveats and data-quality notes

• fiscal.usaspending_by_state's declared 2017-2025 coverage window is not what is actually loaded; only FY2022-2025 held rows on a live scan. Report treats 2022-2023 as the spending window, not 2017-2023.
• No life-expectancy table exists in this corpus; figures were fetched directly from CDC's Socrata endpoint (2010-2015 USALEEP) and from a secondary aggregator (World Population Review) for 2023, because CDC's own 2022 state life-table PDF (NVSR 74-12) and STACKS 2018-2022 map page both returned HTTP 403 from this environment. The 2023 cross-section is one step removed from a primary CDC document; treat it as CDC-sourced-via-secondary-compilation, not as independently verified against CDC's own table.
• DC and the five populated territories were excluded from the state-level regression (n=50) per the corpus's own mixed-unit-kind warning.