Statistical flags indicate unusual patterns — not proof of fraud or wrongdoing. Read our methodology

Geographic AnalysisFebruary 17, 2026·5 min read

Geographic Risk Hotspots

Where do fraud signals concentrate? We normalized flagged provider counts by state population to find where suspicious billing is most dense. The results are surprising: it's not always the biggest states that have the most problems.

1.08

Highest per 100K (VT)

114

Most total flags (NY)

$2,411

Highest $/capita (MA)

982

Total flags nationwide

When people think of Medicaid fraud hotspots, they often think of large states like New York, California, or Texas. And in absolute numbers, they're right: New York leads with 114 flagged providers, followed by California (104) and Arizona (71). But raw counts obscure a more revealing picture.

Small states, high rates

When we normalize by population, Vermont leads with 1.08 flagged providers per 100,000 residents — nearly the rate of New York. The District of Columbia (1.03) and Maine (1.00) round out the top three.

Arizona stands out as a large-population state with an unusually high per-capita rate of 0.96 per 100K — driven by the influx of new behavioral health clinics billing massive amounts. With 71 flagged providers across 7.4M residents, Arizona's rate is nearly 4× that of similarly-sized states like Ohio (0.31) and Pennsylvania (0.28).

Spending doesn't always correlate with flags

Massachusetts spends the most per capita at $2,411/person, but ranks only 9th in flags per capita (0.67). Meanwhile, Nevada has a high flag rate (0.78) despite relatively low Medicaid spending ($302/person). This suggests that flag rates reflect billing behavior patterns, not just spending volume.

Another striking pattern: Nevada gets 21 of its 25 flags from the ML model rather than statistical tests — the highest ML-to-statistical ratio of any state. Maryland is similar, with 11 ML flags vs 11 statistical flags. This may indicate different types of anomalies in these states — patterns that traditional rule-based tests miss but machine learning picks up.

Top 15 States by Flags Per 100K Population

#StateFlags/100KFlaggedSpending/CapitaPopulation
1Vermont1.087$1,439647,464
2District of Columbia1.037$1,657678,972
3Maine1.0014$9011,395,722
4Arizona0.9671$9017,431,344
5Alaska0.957$1,230733,406
6Delaware0.879$8461,031,890
7New Hampshire0.8612$6871,402,054
8Nevada0.7825$3023,194,176
9Massachusetts0.6747$2,4117,001,399
10West Virginia0.6211$8321,770,071
11New York0.58114$1,82419,571,216
12Rhode Island0.556$1,0301,095,962
13Kansas0.5416$2402,940,546
14New Mexico0.5211$9732,117,522
15New Jersey0.5046$1,0599,290,841

Top 15 States by Total Flag Count

#StateFlaggedStatisticalMLTotal Spending
1New York1141068$35.70B
2California104995$30.89B
3Arizona71683$6.69B
4Texas61574$9.24B
5Massachusetts47434$16.88B
6New Jersey46442$9.84B
7Ohio37325$7.00B
8Pennsylvania36360$3.66B
9Florida33285$2.78B
10Nevada25421$965.4M
11Michigan24222$8.18B
12Tennessee23185$7.34B
13Kentucky22184$3.54B
14Maryland221111$1.17B
15Missouri22220$4.03B

Important caveat: Small-state bias

Per-capita rates can be misleading for small states. Vermont's 7 flagged providers in a population of 647K is enough to top the rankings, while Georgia's 4 flags across 11M residents push it to the bottom. A single fraudulent provider in Wyoming would give it a rate of 0.17 per 100K. Higher detection may also reflect better state-level reporting and audit systems rather than more actual fraud. These rankings should be read as starting points for investigation, not conclusions.

Key Takeaways

  • Vermont, DC, and Maine have the highest fraud flag rates per capita, despite being among the smallest states in the dataset.
  • New York leads in total flags (114), driven heavily by its massive home care industry — 106 from statistical tests alone.
  • Arizona (0.96 per 100K) is notable as a large state with a high per-capita rate, fueled by rapid new-entrant clinics in Phoenix.
  • Spending per capita doesn't predict flags. Massachusetts spends $2,411/person but ranks below smaller states in flag density.