Customer protections · dataset 2026.08
Urgent customer policies are often hard to find publicly.
A directory-wide measurement of whether public material described shut-off procedures, leak adjustments, household assistance, and emergency reporting routes.
34%
At least one shut-off detail identified
1,192 of 3,490
35%
Leak-adjustment detail identified
1,233 of 3,490
31%
Utility assistance detail identified
1,079 of 3,490
34%
Emergency reporting route identified
1,172 of 3,490
Interpretation
This is a visibility finding, not a policy grade.
When a rule is difficult to find, a customer may still receive it on a bill, inside an account portal, by telephone, or through local law. The analysis therefore measures only what the directory identified in public provider-specific material.
The distinction is especially important for assistance. A missing utility-owned program field does not mean a household has no local charity, municipal, state, tribal, or temporary funding option. Likewise, a missing shut-off timeline does not establish that service can be interrupted without notice.
The practical conclusion is narrower: public documentation is uneven enough that a directory should show its evidence and uncertainty, rather than publish a confident-looking answer for every system.
State detail
Coverage by indexed records
Percentages show records with a provider-specific research note in the named field, or at least one such note for shut-off coverage.
| State | Systems | Shut-off detail | Leak research | Assistance research | Emergency route |
|---|---|---|---|---|---|
| District of Columbia | 3 | 67% | 67% | 67% | 67% |
| California | 418 | 56% | 54% | 55% | 54% |
| Texas | 400 | 53% | 58% | 54% | 45% |
| Alaska | 9 | 44% | 56% | 33% | 56% |
| Florida | 261 | 42% | 46% | 37% | 42% |
| Virginia | 66 | 42% | 50% | 33% | 44% |
| Maryland | 27 | 41% | 41% | 26% | 30% |
| Maine | 10 | 40% | 40% | 40% | 40% |
| Colorado | 136 | 38% | 38% | 35% | 52% |
| South Dakota | 8 | 38% | 50% | 25% | 25% |
| New Mexico | 19 | 37% | 37% | 26% | 37% |
| Nevada | 11 | 36% | 55% | 45% | 55% |
| North Dakota | 11 | 36% | 27% | 27% | 36% |
| Missouri | 40 | 35% | 33% | 33% | 28% |
| Washington | 107 | 34% | 35% | 34% | 32% |
| Montana | 9 | 33% | 22% | 22% | 22% |
| Indiana | 50 | 32% | 32% | 28% | 26% |
| Tennessee | 101 | 31% | 32% | 22% | 27% |
| Louisiana | 42 | 31% | 33% | 19% | 40% |
| Nebraska | 13 | 31% | 38% | 31% | 31% |
| West Virginia | 13 | 31% | 31% | 23% | 46% |
| New Jersey | 99 | 30% | 21% | 21% | 31% |
| Wisconsin | 46 | 30% | 28% | 15% | 30% |
| Connecticut | 33 | 30% | 33% | 36% | 39% |
| Arizona | 58 | 29% | 31% | 26% | 22% |
| North Carolina | 97 | 27% | 28% | 23% | 28% |
| Ohio | 115 | 26% | 23% | 23% | 31% |
| Georgia | 86 | 26% | 31% | 27% | 33% |
| Pennsylvania | 107 | 25% | 22% | 21% | 27% |
| Utah | 61 | 25% | 21% | 18% | 23% |
| Oklahoma | 36 | 25% | 22% | 17% | 25% |
| Idaho | 16 | 25% | 31% | 25% | 19% |
| Delaware | 12 | 25% | 25% | 33% | 42% |
| Wyoming | 8 | 25% | 25% | 38% | 25% |
| Vermont | 4 | 25% | 25% | 25% | 25% |
| Kentucky | 50 | 24% | 22% | 20% | 16% |
| Rhode Island | 17 | 24% | 24% | 18% | 29% |
| Minnesota | 70 | 23% | 23% | 23% | 20% |
| Massachusetts | 128 | 22% | 37% | 14% | 16% |
| Michigan | 100 | 22% | 21% | 16% | 19% |
| New Hampshire | 9 | 22% | 11% | 22% | 22% |
| New York | 108 | 21% | 18% | 14% | 23% |
| Kansas | 24 | 21% | 29% | 25% | 29% |
| Hawaii | 14 | 21% | 21% | 21% | 29% |
| Iowa | 35 | 20% | 23% | 17% | 20% |
| Oregon | 48 | 19% | 23% | 27% | 21% |
| Alabama | 77 | 18% | 22% | 18% | 21% |
| South Carolina | 60 | 18% | 22% | 12% | 20% |
| Arkansas | 33 | 18% | 21% | 15% | 30% |
| Illinois | 156 | 15% | 13% | 14% | 13% |
| Mississippi | 29 | 10% | 10% | 7% | 10% |
Method
Fixed definitions and visible denominators
Shut-off coverage means at least one provider-specific research note among grace period, late fee, delinquency timeline, warning protocol, and reconnection cost. The other measures use the named single field. A field containing only an N/A marker is not counted; notes that document an explicit public-information limit retain that qualification.
The analysis uses dataset 2026.08, published August 1, 2026. Read the complete methodology and inspect the downloadable data.