Where Consumer Debt Is Managed Most Successfully: State-by-State Ranking (All 50)

A single metric cannot define responsible debt management: a state can post low delinquency because credit is hard to obtain, or high balances because incomes are high and borrowing is affordable. Responsible management is a set of conditions that hold together:

• (a) Obligations are met on schedule. Delinquency is the primary observable. For personal loans this means installment payments — fixed obligation, fixed term — which behave differently from revolving credit.
• (b) Problems are resolved before they become terminal. Low incidence of third-party collections and bankruptcy indicates borrowers either avoid unmanageable obligations or resolve difficulty earlier through refinancing, consolidation or reduced borrowing.
• (c) Borrowing is proportionate to capacity. A $20,000 loan differs at a $100,000 median income versus $60,000. The relevant measure is burden relative to income, not absolute debt.
• (d) Borrowers retain access to affordable credit. A strong credit file lets a borrower refinance, consolidate or absorb a shock without a 200% APR product; a damaged file frequently cannot. This self-reinforcing mechanism is why credit standing is a pillar, not a summary.

Deliberately excluded: low borrowing volume is not treated as responsible in itself, nor high credit-score attainment among a narrow, already-advantaged population. Within-state variation on every indicator exceeds between-state variation (Source: Debt in America: An Interactive Map).

3.1 Geographic scope

Geographic scope. All 50 states ranked; DC reported separately (§5.3) and excluded from means, standard deviations and boundaries — its 9.6% installment delinquency (highest of any jurisdiction) alongside the lowest debt-burden ratio is a combination no state exhibits. Puerto Rico and territories are excluded, as the source files do not report them.

3.2 Data vintages, publication dates, and coverage

Indicators are drawn from different periods and are not fully time-aligned — from August 2025 (Urban Institute extract) to June 2026 (bankruptcy filings), with income from calendar 2024.

4.1 Why each metric was selected

4.2 Normalisation

Each indicator becomes a z-score across the 50 states only (DC excluded), sign inverted where a higher value is unfavourable, combined within pillars then across pillars, and rescaled linearly so the strongest state = 100.0 and the weakest = 0.0. Consequently the score is ordinal-relative: 50 means mid-range among the 50 states, not "half as responsible," and scores are not comparable across editions. Z-scores were not winsorised — Mississippi's 10.2% installment delinquency and Alabama's 4.12/1,000 bankruptcy rate are genuine outliers that pull the scale but do not alter the top-25 ordering.

4.3–4.4 Missing data and double-counting

No state has missing data on any scored indicator; all 50 appear in all five sources. (Arkansas reports jointly across its Eastern/Western districts — the combined 7,338 filings are used against full state population; multi-district states are summed before per-capita rates.)

Overlap was addressed, not ignored: collections share and average FICO correlate at ρ = −0.961 (a collections account is itself a major FICO negative), so their combined weight is capped at 40% and a FICO-removed sensitivity run was performed (ρ = 0.996 vs. baseline) — the credit-score pillar adds almost nothing independent and should be read as confirmatory. Card and installment delinquency are drawn from different bureaus and combined within one pillar. New York Fed auto 90+ dpd was deliberately kept out to avoid duplicating the Urban Institute installment measure (it is used for confidence instead). Bankruptcy and collections correlate at only ρ = 0.474, confirming the bankruptcy pillar carries independent information.

4.5 Sensitivity analysis

The ranking is not an artefact of the chosen weights. Any reasonable alternative weighting produces substantially the same ordering.

4.6 Confidence levels

Each state gets High/Medium/Low by rule: (1) rank gap between New York Fed auto delinquency and Urban Institute installment delinquency ≤8 → High, ≤15 → Medium, >15 → Low; (2) population below 1.5 million → downgrade one level (both panels are samples). Result: 35 High, 9 Medium, 6 Low. The six Low-confidence states are Hawaii, Montana, North Dakota, Rhode Island, Michigan and Indiana — the last two are large states where the panels genuinely disagree; the other four are small-population.

The index measures association, not causation — it describes co-occurrence of favourable indicators, controls for no confounders, and supports no causal claim.

Column sources (each links directly to the primary dataset): Delinquency = auto/retail installment loans 60+ days delinquent, Aug 2025 · Collections = share of adults with any debt in collections / median balance among them, Aug 2025 · Bankruptcy = nonbusiness filings per 1,000 residents, 12 months ending 30 Jun 2026, using Census population · Credit score = average FICO Score 8, Sep 2025 · Debt burden = non-housing consumer debt per capita (Q4 2025) ÷ median household income (2024).

All rows share the same data period: credit-bureau indicators Aug–Dec 2025; bankruptcy 12 months to Jun 2026; income CY2024.

National reference points: 22.7% of adults have debt in collections (median balance $2,528); installment delinquency 5.7%, card delinquency 5.6% (Source: Debt in America file). Average U.S. FICO was 713 in 2025 (Source: Experian). Nonbusiness bankruptcy filings totalled 581,570 in the 12 months to 30 June 2026 (Source: Table F-2), ~1.70 per 1,000.

How to read the score. The distribution is not uniform: 19 states score above 78, 12 cluster 45–70, 10 fall below 35. The gap from rank 1 to 20 (100.0→76.5) is smaller than rank 40 to 50 (35.9→0.0) — stress is more differentiated at the bottom than excellence at the top.

5.1 District of Columbia (supplemental, not ranked)

DC's profile is internally contradictory in a way no state's is — the highest installment delinquency of any jurisdiction (9.6%, higher than Mississippi) alongside the lowest debt-burden ratio (0.106) and below-average bankruptcy. This is characteristic of extreme internal inequality: a high-income, low-leverage population coexisting with severe credit distress, averaged into one figure. The composite above should not be used to rank DC against states.

A recurring pattern: strong on delinquency and collections, far more variation on bankruptcy and debt burden. None leads on every pillar; several have a clear weak point. Full figures are in the §5 table; the notes below flag each state's standout strength and its one qualification.

1. Vermont — RBI 100.0 · Medium. Leads or near-leads on four pillars (installment delinquency 3.0%, lowest median collections balance $2,079, bankruptcy 0.44, FICO 737). Only qualification: mid-range debt burden (0.159), and Medium confidence on a small population (~645,000).

2. Minnesota — RBI 97.9 · High. Leads outright on the lowest collections share (12.8%) and highest FICO (741). Weak point is bankruptcy (1.74/1,000, 27th). The strongest large-population case in the ranking.

3. Massachusetts — RBI 94.5 · Medium. Second-lowest burden ratio (0.125), driven by a $104,569 median income. Relative weakness is revolving credit (10.33% card 90+ dpd, 18th); high income does considerable work in the score.

4. Wisconsin — RBI 92.1 · High. Lowest card delinquency in the country (8.03%) on a below-median income — yet one of three states whose statutes require only that a loan not be "unconscionable" rather than a numeric APR cap (Source: NCLC 2025). The clearest counterexample to the argument that rate caps drive outcomes.

5. New Hampshire — RBI 91.0 · Medium. Lowest poverty rate in the dataset (7.3%) and low bankruptcy (0.69). Weak spot: median collections balance $2,737 (above national) — fewer residents in collections, but larger amounts.

6. Maine — RBI 87.3 · Medium. Lowest bankruptcy in the country (0.40), but the weakest top-ten burden ratio (0.179) and a collections share (19.2%) only modestly better than national. Rests heavily on the bankruptcy pillar — falls several places in the variant that removes it.

7. Washington — RBI 87.1 · High. Third-lowest collections share (14.3%) on a $99,309 median income. Distinctive weakness: median collections balance $3,197, fourth-highest of any state — the fewer-but-larger pattern again.

8. Hawaii — RBI 86.0 · Low. Installment 3.7%, low bankruptcy (0.79). Low confidence: the two bureau panels diverge by >15 places and population is below 1.5m — both downgrade rules trigger. Placement is provisional.

9. Connecticut — RBI 85.9 · High. Strength concentrated in burden (ratio 0.133, third-best), but collections (18.3%), card delinquency (10.53%) and FICO (724, lowest in the top ten) are only moderately better than national. Ranks ninth largely on capacity, not payment performance.

10. Iowa — RBI 85.8 · High. Strongest low-income entrant: on a below-median income ($75,623) it posts the lowest top-ten non-housing debt per capita ($11,530) and third-lowest median collections balance ($2,022). The most direct evidence that low borrowing and consistent repayment can produce top-decile outcomes independently of high income.

What the top ten share — and don't. Shared: all have installment delinquency ≤4.0% (national 5.7%), collections shares ≤19.2% (national 22.7%), FICO ≥720. Not shared: bankruptcy ranges 0.40 (Maine) to 1.74 (Minnesota) — a fourfold spread; burden 0.125 (Massachusetts) to 0.179 (Maine); incomes $75,623 (Iowa) to $104,569 (Massachusetts). There is no single profile of a high-performing state — the strongest argument against reading the ranking as resident character.

7.1 The bottom ten

Three distinct stress patterns, calling for different responses:

• Pattern A — Broad-based distress (Mississippi, Louisiana, Alabama, Arkansas): elevated on every pillar; median incomes $59,301–$66,704, poverty 15.2–18.6%. Mississippi's 10.2% installment delinquency is nearly double the national rate. The constraint appears to be income capacity as much as borrowing decisions.
• Pattern B — High collections, contained insolvency (Texas, South Carolina, North Carolina, West Virginia): Texas has the second-highest collections share (32.9%) but bankruptcy of 1.18 — below national. Likely procedural, not behavioural: exemption laws, garnishment rules and filing cost differ. A significant limitation on the bankruptcy pillar (§9).
• Pattern C — Insolvency-heavy (Alabama, Tennessee, Nevada, Georgia): Alabama's 4.12 and Tennessee's 3.06 filings/1,000 are the two highest; both reflect long-established Chapter 13 filing practice — the route distress takes as much as its volume.

None of these states can be characterised as having less responsible residents. Mississippi's median income ($59,301) is the lowest in the country and its poverty rate (17.8%) among the highest; Louisiana's poverty is 18.6%. The data cannot separate constrained capacity from borrowing decisions.

8.1 Regional averages

Using Census regional definitions, with DC excluded:

The regional gradient is the single largest structural pattern. The South's mean RBI (34.3) is less than half the Northeast's (83.4), and every one of the ten lowest-ranked states is in the South or the West's Nevada — mirroring the Urban Institute's long-standing finding that debt in collections concentrates in southern states. The Midwest and West have near-identical means but different compositions: the Midwest is tightly clustered (Wisconsin 4th, Minnesota 2nd, but Indiana 34th, Ohio 32nd); the West is bimodal (Washington, Hawaii, Oregon, Montana top-12; Nevada, Arizona, New Mexico bottom-15).

8.2 Socioeconomic associations

Poverty is a stronger correlate than income (−0.770 vs. +0.594): the share of households at the low end matters more than where the middle sits. New Hampshire (poverty 7.3%, rank 5) versus Maryland (higher poverty, rank 29) despite similar medians. Income is far from decisive — Iowa (median $75,623) ranks 10th while Maryland (29th) and Virginia (25th) sit on much higher incomes; Vermont ranks 1st on $82,836.

8.3 The legal-protection overlay

As of December 2025, for a $2,000 two-year installment loan: 31 states and DC cap APR at 17–36%; 11 permit 37–60%; five (Alabama, Idaho, South Carolina, Utah, Wisconsin) impose no numeric cap and require only "unconscionability"; two (Delaware, Missouri) impose no cap at all (Source: NCLC 2025).

Recent movement is towards higher permitted costs: Mississippi extended its Credit Availability Act (permitting APRs above 300%) from 2026 to July 2030 and raised the max loan from $2,500 to $3,250 with inflation indexing; Tennessee raised a maximum rate from 30% to 36% and an origination fee from 10% to 12.5%, lifting the max APR on a $500 six-month loan from 106% to 114%; Oklahoma's inflation adjustments lifted the max APR on a $2,000 two-year loan from 54% to 56%.

The overlay produces an informative split:

• Weak protection, high stress: Mississippi (50), Alabama (48), South Carolina (41), Oklahoma (42), Tennessee (40), Texas (44), Missouri (31), Delaware (33).
• Weak protection, low stress: Wisconsin (4), Utah (17), Idaho (18) — all lack numeric caps yet rank top-18.

This refutes a simple causal story both ways. What the overlay identifies is where a consumer meeting a high-cost offer has the least statutory recourse — a question about the information a borrower needs, not about character.

8.4 Composition effects that the index cannot remove

Not controlled for: population age structure (FICO rises with age — 678 Gen Z vs. 760 Silent Generation nationally, so younger/high-in-migration states show lower averages); medical debt reporting changes (bureaus changed reporting from 2022; seven states restrict medical debt on reports, so collections are not comparable on identical rules); credit access (low delinquency can reflect restrictive lending as easily as strong repayment); bankruptcy procedure (exemption statutes, garnishment, local filing practice vary by district — the Texas–Alabama contrast in §7).

This ranking establishes association. It does not establish causation, and cannot.

What can legitimately be said: Vermont's residents in aggregate showed lower delinquency, collections and bankruptcies than Mississippi's over the periods measured — a well-supported description of outcomes. These indicators co-occur (pillar correlations mostly 0.5–0.9). Poverty is more strongly associated with the composite than median income.

What cannot be said:

• That top-ranked residents are more financially literate — no literacy measure met the source criteria; low delinquency is compatible with literacy, stable incomes, restrictive credit supply, employer benefits or a favourable age distribution.
• That low delinquency proves better behaviour — identical behaviour produces different outcomes under different income volatility, healthcare exposure and employment stability. The Urban Institute's data show collections rates differing sharply between predominantly white communities and communities of colour within the same states — a structural finding, not behavioural.
• That a state's rank predicts an individual's situation — within-state variation exceeds between-state on every indicator.
• That improving one indicator would improve the composite — the index has no causal model; making bankruptcy filing harder would raise a state's score while worsening residents' actual position, demonstrating the index measures observable outcomes, not welfare.

Plausible mechanisms, as hypotheses only (none tested here): income stability rather than level (consistent with the stronger poverty correlation); healthcare cost exposure (collections include medical obligations); credit supply composition (affordable mainstream credit gives shock-hit borrowers better options); demographic composition (age, migration, household structure affect aggregate credit-file measures). Offered to indicate the range of explanations compatible with the same data — which is why a causal reading is unwarranted.

For consumers — your state's rank tells you about your environment, not about you. A Mississippi resident with a stable income and strong file faces the same personal-loan mathematics as a Vermont resident with the same profile. What differs is the market: legally available APRs, density of high-cost lenders, and statutory recourse. Three practical consequences:

1. Where there is no numeric APR cap, comparison falls entirely on the borrower — Delaware and Missouri (no ceiling); Alabama, Idaho, South Carolina, Utah, Wisconsin (only "unconscionability"). A licensed lender's offer cannot be assumed reasonable.
2. APR, not monthly payment, is the comparable figure — it incorporates interest, fees and term. A longer term lowers the monthly payment while increasing total cost.
3. Fees can dominate the stated rate — NCLC documents a Pennsylvania line of credit at 5.98% stated interest producing a 431% effective APR once a recurring monthly fee is included.

For lenders and underwriters. State-level averages are weak underwriting inputs — within-state dispersion exceeds between-state on every indicator. Installment and revolving delinquency diverge meaningfully by state (Wisconsin 1st on card, 12th on installment; Massachusetts 2nd on installment, 18th on card), so portfolio expectations built on one product's state pattern won't transfer. The bankruptcy pillar is procedurally contaminated (the Texas result shows filing propensity is a poor proxy for distress across state lines).

For the informational/comparison sector. The clearest gap: no public dataset reports personal-loan performance at state level. Every state-level personal-loan claim in circulation is derived from proxies or unverifiable. A platform that states this openly and labels its proxies occupies a defensible position most competing content does not.

• Federal Reserve Bank of New York, Center for Microeconomic Data — State Level Household Debt Statistics 2003–2025 (XLSX), Feb 2026 (sheets: creditcard_delinq, auto_delinq, total, mortgage, studentloan; column Q4_2025; NY Fed Consumer Credit Panel / Equifax; excludes territories, subject to sampling variation). Quarterly Report 2026 Q2 (PDF) — total household debt $18.8tn in Q2 2026, 4.7% of balances delinquent. Panel: 5% random sample of individuals 18+ with an Equifax file (1% for student loans).
• Urban Institute — Debt in America — State/national data file (CSV), 23 Oct 2025 (fields: totcoll, totcollpos_p50, autoretdelrate, carddelrate, household_income_avg). Interactive Map — ~4% panel, August 2025, >10 million records. Technical Appendix 2025 (PDF) — variable definitions; auto/retail delinquency covers people 60+ days delinquent on an auto loan, lease or retail installment loan.
• Administrative Office of the U.S. Courts — Table F-2, 12 mo. ending 30 Jun 2026 (XLSX) · PDF · landing page (nonbusiness filings aggregated district→state; total 581,570). Bankruptcy Filings Rise 11 Percent — trend context (+11.2% to 549,577 in 12 mo. to 31 Dec 2025).
• U.S. Census Bureau — State Population Totals, Vintage 2025 (CSV) (POPESTIMATE2025). SAIPE 2024 (XLS) — median household income and poverty rate (county FIPS = 0), model-based.
• Experian — What Is the Average Credit Score in the U.S.? (30 Mar 2026, data through Sep 2025; national average 713; personal-loan delinquency 3.76%). Personal Loan Debt Study (average balance $19,333; combined $597.6bn, $207.1bn unsecured; 38.0% of consumers hold a personal loan). 2025 Consumer Credit Review.
• National Consumer Law Center — Predatory Installment Lending in the States (2025) (19 Dec 2025) · Full report (PDF) · Appendix A · Appendix C. APRs computed under the Truth in Lending Act for closed-end installment loans by licensed non-bank lenders.
• Consumer Financial Protection Bureau — Consumer Use of Buy Now, Pay Later (PDF, Jan 2025) (BNPL not furnished to bureaus — why it is excluded). Consumer Credit Trends.
• Reproducibility. Extract the Q4_2025 column from the named NY Fed sheets; read the five named fields from the Urban Institute CSV; sum the nonbusiness column of Table F-2 across districts by state (Arkansas reports as one combined district) ÷ POPESTIMATE2025 × 1,000; read the 2025 column of the Experian state table; read median income and poverty from SAIPE rows. Z-scores on the 50 states only, weighted per §4, rescaled linearly to 0–100.

Sources deliberately not used: state credit-score tables without dated sources or disclosed methodology; pages citing figures that differ from the reports they attribute; content untraceable to any original dataset. Where secondary coverage was encountered, the primary source was retrieved and used directly.

This ranking reflects the data available and the methodology chosen; different indicators, weights, periods or normalisation would produce a different ordering. Scores are meaningful only relative to the other states in the same edition, from the same sources.

Not personalised financial advice. Nothing here assesses any individual's circumstances, creditworthiness, or product suitability. BankGuider.com is an informational and comparison platform — not a lender, credit provider, adviser, broker, debt-relief provider or government agency, and it does not originate, price, approve or service loans. Readers needing individual assessment should consult a licensed professional or a nonprofit credit counselling agency.

The ranking does not demonstrate that any state's residents are inherently more or less financially responsible. Credit outcomes are shaped by income level and volatility, healthcare costs, insurance coverage, employment structure, credit availability and pricing, state lending and collection law, and demographics — none of which are attributes of character. The composite correlates with median income at only 0.594: income does not explain the pattern, and it does not follow that behaviour does. Several indicators are proxies — no public dataset reports personal-loan delinquency, balances or losses for all 50 states; the closest proxy (auto/retail installment delinquency) shares the repayment structure but is not the same product. Bureau indicators derive from ~5% and ~4% samples with proportionally larger variation for small states; medical debt reporting rules changed from 2022 and differ across states; bankruptcy rates reflect exemption statutes, garnishment law and local practice as much as distress; consumers without credit files are absent from every bureau-based measure, so the populations under greatest pressure may be under-represented. Data periods span August 2025 to June 2026 with income from calendar 2024 — the composite is a blended picture, not a single-point measurement.

Prepared as independent research. This document contains no individualised financial, legal or tax advice.