The Flourishing Life Year is a new measure of the American good life — a framework, grounded in federal data, that describes whether a life is flourishing across five domains: purchasing power, wealth, family, health, and useful education. Take the assessment to see where your life stands against the national age trajectory, or scroll to see how the country looks county-by-county on the live map below.
Hover any county for its scores. Higher is better; gray means insufficient federal data (≥ 4 of 5 domains required). Two summary metrics are available — pick which question you want answered, then color the map by it. Or switch to a single-domain view to see one indicator at a time.
A Flourishing Life Year is the amount of human flourishing produced in one year, benchmarked against a national reference. It's the unit we use to describe places and people with the same math — the same federal data, the same five domains, the same arithmetic.
The definition is normative — a positional claim about what makes a year of American life worth counting as flourishing — and the index states it explicitly rather than smuggling it in through the mechanics:
A Flourishing Life Year is a year in which a citizen is:
The methodology page documents the empirical grounds for each condition — and, for the family condition specifically, the editorial position the index is taking.
Applied to a county, the index produces a map of how well places sustain flourishing lives by this definition. Applied to a person, it produces a Flourishing Read — a categorical assessment per domain against the average for Americans your age. Same five conditions. Same federal data. Same honesty.
How much flourishing is the country actually producing? Lived is the headline — total per-person flourishing aggregated nationally, with below-2019 domains weighted 2× per loss aversion. Anchored at 2019 (last full pre-COVID year). The nominal total — the additive face-value figure with no loss-weighting — stays alongside in lighter type so the gap between the two ledgers (what we produced on the books vs what we produced in lives) is visible at every step. § Why losses count double explains the rule.
2020 gap. Census did not release a 2020 ACS 1-year PUMS due to COVID-era survey-response quality concerns. The trend interpolates visually across the gap; no synthetic 2020 estimate is published.
Methodology. Every adult (18+) in PUMS is scored Strong / Adequate / Below across the five domains using absolute 2019-frozen thresholds. Adult Nominal FLY = 0.20 × #Strong + 0.10 × #Adequate (range 0.00–1.00). Children inherit the weighted-mean adult FLY of their household — a 5-year-old in a Strong-flourishing home is flourishing; her score is the average of the adults raising her. Group-quarters children (foster care, juvenile detention, institutional placement) score 0, since institutional placement is, by the framework's own logic, a family-domain failure. National total FLY is the PWGTP-weighted sum across the full US population. The thresholds are the same across years — so an increase in the index means more Americans clearing the 2019 bar, not just outranking each other.
Most-recent-year cutoff: the trend extends to 2024, the most recent year with ACS 1-year PUMS published. The 2025 number arrives in fall 2026.
The loss-weighted view in § The national balance isn't a math trick — it's an accounting principle this site uses everywhere it reports change. Levels stay raw; changes get loss-weighted. This page is the one place that defines the rule, so every other figure on the site can link back to it.
The principle. Lived flourishing isn't symmetric. A year of decline registers in households more heavily than an equal-sized year of progress: the lost income, the closed clinic, the broken family, the foreclosed home — these are felt acutely and remembered. Equal-sized gains arrive quietly. Behavioral economics has measured this asymmetry empirically; the standard finding is that people weight losses about twice as heavily as gains (Kahneman & Tversky 1992, "Advances in Prospect Theory," with median λ ≈ 2.25 across replications).
Why the index has to reflect this. If we average the five domains equally and report only the headline number, a country that is rapidly enriching and educating itself can paper over collapsing health and family formation indefinitely. The headline keeps moving up; the lived experience moves the other way. The tension between national accounts and lived reality is the failure mode of every aggregate index, and the loss-weighted view is how this index refuses it.
The rule, in one line. When summarizing change across domains: above-baseline gains count once, below-baseline losses count twice.
Where the rule applies, where it doesn't. Loss aversion is a property of change, not state. So the rule applies to every figure on this site that reports a delta — year-over-year, vs-2019, vs-cohort, projection. It does not apply to per-person scores or per-county composites: those are levels, and loss-weighting them would silently change what an individual's score means. Loss-weighting is layered on top when the question is "how is the country moving?", not "what is it now?"
The lived FLY total — derived, not redefined. The site's headline national total has been switched from nominal FLY (the additive face-value figure) to lived FLY. Lived FLY is a derived figure, defined precisely:
The 2019 baseline is unchanged: lived 2019 = nominal 2019 (loss-weighting is zero at the baseline year by definition). For every later year, the country's nominal production of flourishing is multiplied by the loss-weighted change to get the lived equivalent. The gap between nominal and lived is the cost of unbalanced flourishing — what the country produced on the additive ledger but lost on the lived ledger because some domains backslid. That gap is visible directly on the primary balance chart: the dashed line (nominal) sits above the solid line (lived) whenever any domain is below 2019.
What this preserves, and what it doesn't. Per-person FLY scores stay unchanged — the §02 equation (0.20 × #Strong + 0.10 × #Adequate) still describes a person, and the additive sum across people still gives the nominal national total. County map values stay nominal-equivalent — geography is comparative. What changes: the headline national figures, the cost-per-FLY denominator, and the year-over-year deltas now read as lived figures, with the nominal numbers shown beside them in lighter type for audit.
What you'll see on this site. Lived FLY is the headline metric in §The national picture, §02 (rollup), §The national balance (chart and summary), and §05 (cost-per-FLY). Nominal figures still appear in lighter type beside the lived ones, so the loss-weighting drag is never hidden — only made the primary reading. Domains below their 2019 bar carry a small badge on their card. The point is to make the asymmetry structural, never decorative.
Reference. Kahneman, D. & Tversky, A. (1992). "Advances in Prospect Theory: Cumulative Representation of Uncertainty." Journal of Risk and Uncertainty 5(4): 297–323. The Liebig's-Law-of-the-Minimum analogy from agronomy (a crop's yield is set by its scarcest nutrient, not its average) and the UNDP's 2010 switch to a geometric-mean Human Development Index are the same idea applied to different aggregation problems.
The same five domains measure a place and measure you. For each, we build a national age-trajectory curve from federal microdata, then see where your inputs — and your county's indicators — land on it. The assessment returns a per-domain read against the cohort average for Americans your age. These are the benchmarks the read is measured against.
Real purchasing power: median household income deflated by the local cost of living. BEA publishes Regional Price Parities at state and metro level — not at county level — so each county gets its metro RPP if it sits inside a metro area, or its state's overall RPP otherwise. The imputation rule is documented, including the known limitation that the state-overall figure is metro-weighted. A dollar in Mississippi and a dollar in San Francisco buy different lives; this domain measures what the dollar actually does.
A four-proxy stack for household balance-sheet position: homeownership rate, median home value, the share of county adjusted gross income that comes from dividends and capital gains, and (per Q8.5) the share of returns claiming an IRA or self-employed retirement-plan deduction. There is no federal wealth survey at county level — these proxies together form the most defensible substitute. Three are stocks (what you've accumulated); the fourth is the flow (whether you're still adding). Q8.5 requires both.
The indicator is the share of children under 18 living with married parents. Per Q8.6 (Definition of Flourishing), the family domain anchors specifically on stable married-couple household structure — cohabiting two-parent households are not counted toward this indicator. The methodology section below states the editorial position openly. Family structure is the most direct, county-resolvable measure of the conditions in which children grow up.
Life expectancy at birth from the NCHS Small-area Life Expectancy Estimates Project (USALEEP), with age-adjusted mortality from CDC WONDER as the suppression-friendly fallback for small counties. The most legible single measure of how long lives last here.
Education here means human capital, not academic credentialism. Three components, geometric-mean'd: the share of adults with a post-secondary credential (associate's or bachelor's-and-above), the share employed in skilled trades — construction, installation, maintenance, repair — and (per Q8.3) the share of working-age adults who BOTH hold a bachelor's-or-higher credential AND are currently employed. The first two measure capability and trade depth; the third measures Q8.1.2's "meaningfully employed" — credential AND deployment, not merely credentialed. The master electrician and the BA count equally; the BA driving rideshare doesn't.
The county FLY map shows a snapshot — one release at a time. To answer "is the nation moving forward or back?" we need absolute levels over time. These are the national-level headline indicators per domain, drawn from the same federal sources, year by year. Change figures show relative growth from the first year to the last.
Sources: U.S. Census Bureau ACS 1-yr (5-yr 2016–2020 used for 2020 since the 1-yr release was suspended); BLS CPI-U all-items for income deflation; CDC NCHS National Vital Statistics for life expectancy. The line charts show one indicator per domain — the one that's directly comparable across years; multi-component domains use only the longest-available comparable component. We deliberately do not roll all five into a single "national FLY" line: combining absolute levels in different units across decades requires methodology calls the data can't justify.
Every indicator is sourced from a public federal dataset, transformed in the open, and scored transparently. If a county-level number doesn't exist honestly, we don't fabricate one. Some gaps stay gaps — and we say so.
Pull from the American Community Survey (Census), the Bureau of Economic Analysis (BEA), IRS Statistics of Income (SOI), and CDC's National Center for Health Statistics. No proprietary data, no private licenses, no black-box APIs.
Crosswalk tract and county geographies to a consistent unit (50 states + DC). Where data is too unreliable to publish, suppress and label. Never impute, model, or fill from neighbors.
Each domain is scored against absolute frozen-2019 thresholds (Strong / Adequate / Below). Per-person Nominal FLY = 0.20 × #Strong + 0.10 × #Adequate. Lived FLY layers loss-weighting on top: Below contributions actively subtract at 2× weight, so a deficit in any one domain visibly pulls the score down.
Release data, code, and methodology together. Versioned annually (vYYYY.MM, anchored to the ACS December release). Each prior version preserved and recomputable from the same inputs.
A county's summary score requires data on at least 4 of 5 domains. With one domain suppressed (structurally — see clarification below), the score is computed from the four available and flagged on hover. With two or more suppressed, the county shows as gray on the map. High-noise survey estimates are flagged for reliability, not suppressed — they still contribute to the score, but the map fades them and the tooltip shows the MOE so a reader can decide. No imputation, ever — no state averages, no nearest-neighbor fill, no modeled estimates. The number is either reproducible from federal data, or it isn't.
The family-domain editorial position (Q8.6). The index treats stable two-parent households with children as the family-domain definition of flourishing past age 30, on the empirical grounds that this household configuration produces the strongest measured outcomes on income, wealth accumulation, child educational attainment, and adult mortality. This is a normative position, not a neutral observation. People who flourish outside this configuration — never-married, divorced, child-free, single-parent, same-sex — are real and not failing; the index simply does not register their household structure as flourishing on this domain. The position is empirical, not moral; the moral question of whether American policy should support this configuration is downstream. The county-map indicator (married-couple share of children under 18) reverses the prior cohabiting-couple inclusion to anchor the indicator on the position the index actually takes.
Three clarifications. (1) ACS MOE: flagged, not suppressed (revised v2026.05). Earlier drafts dropped any ACS estimate where the margin of error exceeded 30% of the estimate. That hid roughly 650 counties from the education domain alone — most of them rural, where the underlying populations are real but the survey samples are small. The current rule keeps the estimate, ranks it like everyone else's, and surfaces the MOE on the map (the county fades to ~65% opacity and the tooltip shows e.g. ±42% MOE). Suppression is reserved for structural data gaps — denominator zero, source unavailable. (2) Suppression vs. reliability for CDC WONDER. WONDER itself suppresses death counts of 9 or fewer; the additional n < 20 rule is an additional rate-reliability threshold applied on top — different conceptually from the ACS MOE flag because mortality is rate-based, not survey-based. (3) The county-RPP imputation in the purchasing-power domain is our own deterministic rule (metro RPP for metro-attached counties; the state's overall RPP otherwise) — it is not a BEA product. BEA does not publish a per-state non-metropolitan RPP, only a single national non-metro figure, so the state-overall figure is the most local available signal for non-metro counties even though it is metro-weighted. Documented because it is the one place this methodology relaxes the no-imputation rule, and only because BEA does not publish below the metro level.
This product uses the Census Bureau Data API but is not endorsed or certified by the Census Bureau.
A rank-order that's right is worth more than a decimal score that's wrong. Where the data can't support county-level precision, we aggregate up rather than invent.
The summary matters less than its components. Every page shows the underlying indicators driving the score — drill down until the claim is falsifiable.
Methodology, source list, transformation code, and raw outputs are all public. If you can't reproduce our number, it's a bug.
Government doesn't buy flourishing. Citizens flourish — or don't — through the choices they make and the circumstances they live in. But choices happen inside an environment: schools that work, hospitals that exist, roads that connect, laws that hold. That environment is built and maintained with public dollars. Stack a year of public spending against a year of flourishing produced, and you get the public investment per Flourishing Life Year. Not a price tag — an accounting.
The state doesn't buy flourishing — but it does build the floor. A Flourishing Life Year is produced by a person, not purchased by a government. Citizens hold jobs, raise families, get educated, age in place. The state doesn't make any of those things happen. But it does build and maintain the environment those things happen in — public schools, courts, public health, infrastructure, security, social insurance. The number above isn't what flourishing costs. It's what the country spent maintaining the conditions inside which flourishing happens. Without those conditions, fewer Americans would clear the 2019 bars on the five domains; that's the conceptual link the dollar figure stands in for.
Of which $1.83T was borrowed. The federal deficit in FY2024 — the gap between what Washington collected and what it spent — is what the Treasury added to the national debt that year. State and local governments mostly balance their general funds by law, so at the consolidated level the borrowed share is essentially the federal deficit. We didn't pay for $11K of every flourishing life year produced; we promised future taxpayers would.
Why we subtract federal grants. Of the $6.75T in federal outlays, ~$1.1T flows to states and localities (Medicaid is by far the largest piece) and is then spent by them on direct services. Adding federal outlays to state and local direct spending without netting this out double-counts those dollars. Gross unconsolidated spending is ~$10.95T; the consolidated figure here is what taxpayers actually paid for.
What's outside the frame. This is public spending only — federal, state, and local government outlays. Private spending on the things that produce flourishing (employer-provided healthcare, household education spending, private retirement contributions, mortgage payments that build wealth) is not in the numerator. The cost-per-FLY here is therefore a lower bound on the total social cost — it's the public share, not the all-in cost.
What this number is and isn't. It's an average, not a marginal cost. It doesn't say what the next FLY costs to produce, or which dollars produced flourishing and which didn't. A per-domain attribution — federal health spending ÷ health-domain FLY, education spending ÷ education-domain FLY — is a plausible next cut, and a more demanding one. Keep this number in the spirit of directional honesty: the order of magnitude is the message.
Vintages. Federal: Treasury / CBO Monthly Budget Review FY2019–FY2024 finals (FY2024: $6.75T outlays, $1.83T deficit). Federal grants to state & local: Office of Management and Budget / CBO (FY2024 ~$1.1T, of which $618B is Medicaid). State and local: Census Annual Survey of State and Local Government Finances through FY2022 (most recent comprehensive), grown forward to FY2024 at trend. CPI-U (BLS, all-urban consumers, calendar-year average) is used as the deflator, indexed to 2019 = 1.000. FLY divisor: ACS 1-year PUMS, most-recent year currently 2024. Census did not publish a 2020 ACS 1-year PUMS due to COVID-era survey-quality issues, so the trend chart shows a gap there.
David Verchere built the Flourishing Life Year Index to better understand how the United States is doing as a nation — county by county, year over year, against a measure that isn't GDP. The economy is what we produce. The Flourishing Life Year is how we're living.
The framework is reproducible from public federal data; the code, sources, and methodology are open. If a number on this page is wrong, it can be checked. If the methodology has a flaw, it can be argued with on the merits.
Questions, corrections, collaborations — dverchere@gmail.com · GitHub.
The economy measures what you make. The Flourishing Life Year measures how you're doing.
Eleven questions. Three minutes. Nothing saved. Every number returned is traceable to a public federal dataset — the same data the county map is built on. Each of the five domains comes back as a categorical read: strong, adequate, or below your cohort. Family is shown descriptively — the index does not put a value judgment on household structure.
Take the assessment