What a state’s score means
Each score answers one question: of the other places, how many are under less strain? Louisiana ranks 1st with a score of 100, so it is more distressed than every other state and D.C. North Dakota ranks 51st and scores 0.
We round scores down to a whole number, so “more distressed than” is always true as written. Only the most distressed place can score 100, and only the least distressed scores 0.
Ranks run from 1st, the most distressed, to 51st, the least. Two places share a rank only when their results are exactly equal, and the page then says “tied for” that rank. Right now Connecticut and Illinois are tied for 13th. Tied states get the same score: the share of places below both of them. (Why D.C. is included is in the questions below.)
State pages print a label next to each score. Each label covers ten points of the score:
| Score | Label on state pages | Places now |
|---|---|---|
| 0–9 | exceptionally low state distress | 5 |
| 10–19 | very low state distress | 5 |
| 20–29 | low state distress | 5 |
| 30–39 | low-moderate state distress | 5 |
| 40–49 | moderate-low state distress | 5 |
| 50–59 | moderate state distress | 5 |
| 60–69 | moderate-high state distress | 5 |
| 70–79 | high state distress | 5 |
| 80–89 | very high state distress | 5 |
| 90–100 | extreme state distress | 6 |
What goes into the score: four parts, equal weight
Each part is 25% of the score. Inside a part, each measure gets an equal share of that part. For every measure, a higher figure means more strain.
The county score has a fifth part, the safety net, and the national score’s fifth part is saving. The state score has four parts. It once counted SNAP enrollment as distress, but enrollment also depends on program rules and sign-ups, so renters’ housing costs took its place (more under What changed).
Late payments
Three measures: among people with credit card debt, the share with a card account 90 days or more behind, and the same for auto loans and for mortgages. That’s about three missed payments or more.
Collections and bankruptcy
The share of people with any debt in collections, and personal and business bankruptcy filings for every 100,000 residents over a full year. Both are late-stage signs of debt trouble.
Jobs
The state’s unemployment rate, adjusted for normal seasonal swings and averaged over the latest three months, so one unusual month doesn’t move the score on its own.
Renters’ housing costs
The share of renter households that pay half or more of their income on rent and utilities. Homeowners’ costs aren’t in it.
| Part | Measure | Share of score, % | Source | Latest data | Source updates |
|---|---|---|---|---|---|
| Late payments | Credit card borrowers 90+ days behind | 8.3 | Federal Reserve Bank of Philadelphia | Q2 2026 | Quarterly figures; no fixed release date |
| Auto loan borrowers 90+ days behind | 8.3 | ||||
| Mortgage borrowers 90+ days behind | 8.3 | ||||
| Collections and bankruptcy | People with debt in collections | 12.5 | Federal Reserve Bank of Philadelphia | Q2 2026 | Quarterly figures; no fixed release date |
| Bankruptcy filings per 100,000 residents | 12.5 | Administrative Office of the U.S. Courts | 2025 | Yearly; we use the latest full calendar year | |
| Jobs | Unemployment rate, three-month average | 25.0 | U.S. Bureau of Labor Statistics | July to September 2026 | Monthly |
| Renters’ housing costs | Renter households paying half or more of income on rent | 25.0 | U.S. Census Bureau | 2025 | Yearly |
Late payments
Credit card borrowers 90+ days behind · 8.3%
Auto loan borrowers 90+ days behind · 8.3%
Mortgage borrowers 90+ days behind · 8.3%
Collections and bankruptcy
People with debt in collections · 12.5%
Bankruptcy filings per 100,000 residents · 12.5%
Jobs
Unemployment rate, three-month average · 25.0%
Renters’ housing costs
Renter households paying half or more of income on rent · 25.0%
How the score is calculated
- Line up each measure across all 51 places, so each state’s figure sits beside every other state’s.
- Turn each figure into a percentile. A percentile places a figure on a 0 to 100 scale by its rank: the place with the most strain on a measure gets the highest number. Equal figures share an average rank.
- Average the measures inside each part. That gives the state four part scores.
- Average the four part scores. That gives each state one overall result, and the ranking sorts states by it.
- Turn the ranking into the score. The score is the share of the other 50 places whose overall result is lower, rounded down to a whole number.
Why the four part scores don’t average to the score
Indiana’s part scores are 60.8 for late payments, 75.5 for collections and bankruptcy, 24.5 for jobs, and 53.9 for renters’ housing costs. They average 53.7. Its score is 50, because that’s the share of the other places with a lower average.
The last step is there because an average of percentiles bunches toward the middle. No state is at the top of every measure, so the averages rarely reach the ends of the scale. Turning the averages into a share of places makes the score say exactly what it sounds like.
The formula and the missing-data rules
Each measure’s percentile is: percentile = (average rank − 0.5) ÷ n × 100. The average rank is the place’s rank on that measure, with equal figures sharing a rank, and n is the number of places with data.
A measure goes into an update only when at least 48 of the 51 places have a figure for it. Otherwise it is left out for every state, and the other measures in its part share the weight. If that leaves a part with no measures, the update stops and the last published scores stay up. If a state is missing one measure in a part, its part score uses the measures it has. If it is missing a whole part, its result averages the parts it has. A state missing two or more parts gets no rank. We never fill a gap with zero.
Why the data come from different dates
The current score uses the newest figure each source has published (for bankruptcy filings, the latest full calendar year), and the sources publish on different schedules. Right now the credit figures are from the second quarter of 2026, bankruptcy filings cover 2025, the renter figures are from 2025 and unemployment covers July to September 2026.
We check the sources every weekday. State scores change only in a named release, about twice a year.
The credit figures come from the Federal Reserve Bank of Philadelphia’s Consumer Credit Explorer, built on credit records from the New York Fed and Equifax. They run 3 months behind the jobs figures.
The renter figures come from the U.S. Census Bureau’s yearly American Community Survey. Each year’s estimate normally comes out the following September.
If a source goes too long without new data, we stop updating the scores and leave the last ones up: credit after eight quarters, unemployment after six months, and bankruptcy or renter figures 24 months after the end of their year.
Why a state’s history chart shows a different rank
A state’s history chart can show a different rank and score from the one at the top of its page. The history ends at the second quarter of 2026, the last quarter all four parts have data. Compared with that quarter, the current score uses the same credit figures, newer jobs and renter figures, and bankruptcy filings for the latest full calendar year.
We score every quarter back to the first quarter of 2005 with the same method. In the history, each quarter uses figures from that quarter: credit for that quarter, unemployment averaged over its three months, bankruptcy filings over the 12 months ending then, and the latest renter estimate published by then. To see how a state has moved, compare history quarters with each other, not with the current score.
State pages are dated editions. Their charts stop at the first quarter of 2025, the latest quarter we had when those pages were written. A state’s score today comes from the latest release, which reads each input’s newest figures (for bankruptcy, the latest full calendar year). Our quarterly history reads each input as it stood at the end of that quarter, so its latest quarter is not the same as today’s score.
The renter figures start later than the others, so quarters before the third quarter of 2006 are scored on the other parts. No standard one-year survey estimates were published for 2020, so the earlier estimate stayed in use until the next one came out. The fourth quarter of 2025 is scored without the jobs part: its three-month unemployment average needs October 2025, and the Bureau of Labor Statistics has no figure for that month.
What changed, and when
On September 29, 2026, the score became the share of the other places with a lower overall result. Before that, the published score was the overall result itself, the average that bunches toward the middle. Scores published before that date are on the older scale. This is version 3 of the method.
In July 2026, renters’ housing costs replaced SNAP enrollment, the share of people getting food assistance. Enrollment shows how much help reaches people, which depends on program rules and sign-ups as well as need. When the score counted enrollment as distress, the end of the pandemic’s extra SNAP benefits made households look better off. SNAP figures are still on the SNAP statistics page; they just aren’t part of the score.
What to keep in mind
- Bankruptcy filings depend on state law as well as hardship. Where creditors can take part of a paycheck, people file more often than in states with similar strain. We show the figure as it is and don’t adjust it.
- The renter measure covers about a third of households, and a different share in each state. It measures what rent costs, not evictions or homelessness. A state can score low here because rent is cheap or because renters earn more.
- A rank hides how big the gaps are. Two states next to each other in the ranking may be nearly the same or far apart.
The State Distress Index is not a quality-of-life ranking, and it says nothing about state government budgets or debt. It is not a forecast. It describes where households stand, as of the latest data, compared with households in other states. It is also not the same as the American Distress Index, which compares the whole country with its own past.