The EvvyTools National Index (ENI) — the headline number behind the State of America tracker — is a 0-to-100 composite that synthesizes every live tracker on EvvyTools into a single grade for the U.S. economy. This page is the canonical methodology document: how each tracker is normalized, how the six economic fronts are aggregated, how the persona weights are applied, and how the final score is mapped to a letter grade and an emotional read.
This document is intentionally exhaustive. The intended audience is journalists, researchers, and curious readers who want to know exactly how the number is built before citing it. If you want the high-level overview, read the State of America tracker page instead.
1. The Six Fronts
Every tracker on EvvyTools is assigned to one of six economic fronts. The fronts are intentionally broad — each one bundles trackers that move together and matter together to the typical American household:
- Cost of Living — CPI, core CPI, PCE, food-at-home inflation, medical-care inflation, gas prices, retail prices for eggs and coffee.
- Jobs & Wages — unemployment rate, initial jobless claims, average hourly earnings, JOLTS job openings, labor-force participation rate.
- Housing — 30-year and 15-year mortgage rates, median home price, Case-Shiller index, existing-home and new-home sales, housing starts, building permits, real-estate loans outstanding.
- Borrowing & Saving — federal funds rate, prime rate, Treasury yields (3-month, 2-year, 10-year, 30-year), HYSA yields, CD rates, auto-loan rates, credit-card APR.
- Markets & Wealth — S&P 500, Bitcoin, gold spot, oil (WTI), VIX, USD exchange rates against the major reserve currencies.
- Macro & Fiscal — GDP, retail sales, M2 money supply, Federal Reserve balance sheet (WALCL), national debt, federal deficit.
The tracker-to-front mapping is taken from each tracker’s group_id in the database. The mapping is fixed by design — we do not move trackers between fronts after release, because that would break the comparability of historical index values. (The Currency group rolls into the Markets & Wealth front; the Special group rolls into the Cost of Living front for weighting purposes.)
2. Normalization
Different trackers live on wildly different scales — mortgage rates in percentage points, Bitcoin in tens of thousands of dollars, jobless claims in hundreds of thousands. To make them comparable, every tracker is normalized against its own historical distribution before contributing to a front score.
For each tracker, the live value is converted to a 20-year rolling percentile:
percentile = (count of historical values ≤ current value) / (total count of historical values) × 100
So a 30-year mortgage rate of 6.78% in May 2026 is compared against every weekly mortgage rate observation from May 2006 forward. If 6.78% sits at the 95th percentile of those 20 years, the tracker contributes a percentile score of 95 to its front (before direction-flipping; see below).
Trackers with less than 20 years of history use whatever history they have, with a minimum of 5 years required for inclusion. Trackers below the 5-year floor are tracked on the site but are excluded from the ENI until they accumulate enough history to be statistically meaningful.
3. Direction-Flipping
Higher is not always better. Some trackers move opposite to economic health — high unemployment is bad, high inflation is bad, a high mortgage rate is bad for housing affordability. For those trackers, the percentile is flipped so that the score still rises with economic health:
adjusted = 100 − raw_percentile (for lower-is-better trackers)
Direction is set per tracker in the database via a single boolean flag. The "lower is better" set as of this writing includes: all inflation series, the unemployment rate, jobless claims, all mortgage rates, all consumer-credit rates (auto-loan, credit-card APR, HELOC), the VIX, the national debt, and the federal deficit. Markets & Wealth series, retail sales, GDP, employment, and earnings keep their natural direction.
4. Outlier Clipping
Composite indices are sensitive to single-point outliers — one panic month can swing the headline so far that the whole signal is lost in noise. To prevent that, every tracker’s contribution is clipped to the range [5, 95] before going into the front average:
clipped = min(95, max(5, adjusted))
This means even a tracker at its all-time worst contributes a 5 (not a 0), and even one at its all-time best contributes a 95 (not a 100). The trade-off is intentional: a 0-or-100 reading on a single component shouldn’t dominate a six-front average. In practice, very few trackers hit the floor or ceiling in any given month.
5. Front Aggregation
Within each front, the clipped tracker scores are averaged with equal weight to produce the front score. So if the Cost of Living front contains seven trackers (CPI, core CPI, PCE, food, medical, gas, eggs), the Cost of Living front score is the simple mean of those seven normalized values. Trackers within a front are treated equally because the front itself is the unit of weighting in the next step — over-engineering the within-front weights would introduce subjective judgment that the percentile normalization is specifically designed to avoid.
If a tracker is temporarily missing data (e.g., a release is delayed, the source API is down), it is excluded from that month’s front average. Fronts always require at least three contributing trackers to produce a score; below that, the front carries forward last month’s value.
6. Persona Weights
The six front scores are then blended with persona-specific weights to produce the headline ENI. Each of the six personas has a different weight vector reflecting what matters most to that reader’s lived experience of the economy. The weights below are the live values from StateOfAmerica::PERSONAS as of this page’s last review:
| Persona | Cost of Living | Jobs & Wages | Housing | Borrowing & Saving | Markets & Wealth | Macro & Fiscal |
|---|---|---|---|---|---|---|
| Average American | 20% | 22% | 17% | 13% | 13% | 15% |
| College Student | 22% | 25% | 13% | 25% | 5% | 10% |
| Home Buyer | 10% | 15% | 35% | 20% | 8% | 12% |
| Blue Collar Worker | 28% | 28% | 14% | 15% | 5% | 10% |
| Investor / Wealthy | 8% | 10% | 10% | 12% | 35% | 25% |
| Retiree | 25% | 5% | 10% | 25% | 25% | 10% |
The weights are calibrated by hand against the lived experience of each persona. A retiree on Social Security feels medical inflation and CD rates much more than they feel the entry-level job market — so Cost of Living and Borrowing & Saving get heavier weights, and Jobs & Wages gets a deliberately light one. An investor with most of their wealth in equities feels the S&P 500 daily — so Markets & Wealth gets the heaviest single weight in that profile.
Weights are reviewed annually. Material changes are version-bumped (see the version history at the bottom of this page) so historical ENI values remain reconstructible.
7. Final Score
The final ENI score for a given persona is the weighted average of the six front scores:
ENI = Σ (front_score × persona_weight)
Result is rounded to the nearest integer for display. The headline score on the State of America page is always the Average American persona unless the reader has selected a different one with the “Who’s asking” selector.
8. Letter Grade Mapping
The 0-to-100 score is mapped to a standard A-through-F letter grade with the same +/− bands a school transcript would use. The thresholds are deliberately stricter than typical academic grading to keep the upper grades meaningful — an A+ requires the economy to be near its best 20-year reading on the persona’s most-weighted fronts.
| Score | Grade | Reading |
|---|---|---|
| 95 – 100 | A+ | Once-in-a-generation strong. |
| 87 – 94 | A | The economy is at or near a 20-year best for this persona. |
| 80 – 86 | A− | Better than 8 of the last 10 years. |
| 77 – 79 | B+ | Strong, expanding economy. |
| 73 – 76 | B | Solidly good. |
| 70 – 72 | B− | Above average; comfortable. |
| 67 – 69 | C+ | Average expansion; not booming, not slowing. |
| 63 – 66 | C | An average decade. Not bad, not great. |
| 60 – 62 | C− | Below average; some fronts stressed. |
| 57 – 59 | D+ | Visibly weakening across multiple fronts. |
| 53 – 56 | D | Significant economic pain across this persona’s priorities. |
| 50 – 52 | D− | Approaching recession-level readings. |
| 0 – 49 | F | Crisis-level readings — consistent with recession or financial shock. |
For reference: the index hit 38 in April 2020 (COVID lockdown), 52 in October 2022 (peak inflation), 82 in January 2020 (pre-COVID peak), and is currently in the C / C+ range. The all-time high of 82 and the all-time low of 38 are anchors that future readings will be compared against.
9. Mood Mapping
Alongside the letter grade, every score maps to one of five emotional reads — the “portrait tag” shown on the State of America gauge:
- Booming — score 80+. “Thriving.”
- Confident — score 70–79. “Feeling solid.”
- Steady — score 60–69. “Holding steady.”
- Strained — score 50–59. “Stretched thin.”
- Crushed — score below 50. “Hurting.”
The mood mapping is intentionally cruder than the letter grade — five buckets instead of thirteen — because it’s designed for a single emotional read that matches how Americans actually talk about the economy.
10. Update Cadence
The ENI recomputes whenever any underlying tracker updates. Because the tracker stack updates at mixed cadences — daily for markets, weekly for mortgages, monthly for CPI and jobs, quarterly for GDP — the index in effect re-prints whenever the slowest input that month posts. The “Last reviewed” timestamp on the State of America page reflects the most recent constituent update.
Month-over-month deltas are computed against the prior calendar-month close. Year-over-year deltas are computed against the same month one year prior. Daily intra-month fluctuations are not exposed in the headline metric — the index is designed for medium-term context, not day-trading.
11. Data Sources
Every input to the ENI comes from a free, public, official source. We do not use scraped, estimated, or vendor-proprietary data:
- FRED (Federal Reserve Economic Data, St. Louis Fed) — the umbrella source for most rate, inflation, GDP, and labor series.
- BLS (Bureau of Labor Statistics) — CPI, PCE components, unemployment, jobless claims, JOLTS, average hourly earnings, retail prices for eggs and coffee.
- BEA (Bureau of Economic Analysis) — GDP, personal consumption, savings rate.
- Freddie Mac PMMS — the 30-year and 15-year mortgage rate weekly averages.
- U.S. Treasury / TreasuryDirect — Treasury yields, I Bonds composite rate, federal debt outstanding.
- NAR (National Association of Realtors) — existing-home sales, median sale price, pending sales.
- U.S. Census Bureau — new-home sales, housing starts, building permits.
- S&P / Case-Shiller — the Case-Shiller Home Price Index.
- Stooq — S&P 500 close, oil (WTI), gold spot.
- CoinGecko — Bitcoin price, BTC dominance.
- European Central Bank (via Frankfurter) — daily reference exchange rates for major currency pairs.
- EIA (U.S. Energy Information Administration) — retail gasoline prices.
Each individual tracker page lists its specific upstream source and series ID. We never cite a number on the State of America page that does not have a corresponding tracker page with the raw source linked.
12. Known Limitations
The ENI is a methodology, not a forecast. Some known limitations are worth being upfront about:
- It is not a recession indicator. The index has historically dropped sharply during recessions (April 2020 hit 38), but a falling score does not predict a recession — it confirms one once enough constituent series have weakened.
- It is not regional. The U.S. economy is genuinely several economies stacked — coastal vs. interior, urban vs. rural, large metro vs. small metro — and the ENI is a single national grade. A reader in a fast-growing Sunbelt metro will experience the economy meaningfully differently from a reader in a Rust Belt town with the same headline ENI.
- The persona set is incomplete. Six personas cannot cover every situation. The Average American persona is the catch-all; readers whose lives don’t fit any single persona cleanly should watch Average American alongside whichever of the five specific personas is closest.
- Direction-flipping is a judgment. Marking a tracker as “lower is better” is an editorial choice. Mortgage rates are bad for buyers but good for HYSA savers — we flip them because the persona weighting handles the asymmetry, but reasonable analysts could choose differently.
- The 20-year window is a choice. A 10-year window would weight more recent history more heavily; a 50-year window would catch the high-inflation 1970s and 1980s. We picked 20 years because it spans two full business cycles plus the COVID anomaly — a reasonable middle ground.
13. Citation
If you cite the State of America Index in an article, paper, or external dashboard, please use the following format:
EvvyTools. (2026). State of America — EvvyTools National Index (ENI). Retrieved from https://evvytools.com/trackers/state-of-america/. Methodology: https://evvytools.com/methodology/state-of-america/.
The index is published under a CC BY 4.0 license — free to reuse with attribution, including for commercial purposes.
14. Version History
- v1.0 — May 19, 2026 — Initial release. Six fronts; six personas; 20-year percentile normalization; clipping at [5, 95]; equal within-front weighting.
Any future version that changes the weighting tables, the percentile window, the clipping bands, or the front mappings will be tagged with a new version number here. Historical ENI values published before a version change are not retroactively recomputed — they remain as published under the prior methodology, with the version number recorded so the calculation is reproducible.
Questions or corrections? Get in touch. The State of America tracker itself lives at /trackers/state-of-america/, and the full list of constituent trackers is at /trackers/.