In order to gauge government transparency across New Hampshire, each municipality received the same eight requests.
We asked for AI use policies, AI chat logs, ethics and conflict-of-interest rules, highest-value requests for proposals, leadership and constituent emails, a performance evaluation for the town or city manager, a performance evaluation for the police chief and police settlement agreements.
Each school district received four: leadership and constituent emails, a superintendent performance evaluation, parental objections to instructional materials and parental bill of rights records. That is 184 municipal and 92 district requests. Four additional requests were re-filed or split across agencies, for a total of 280 records requests.
We graded each request on a scale from 0 to 10 based on our communication with the agency. A 10 indicates records were provided speedily and electronically. A 1 indicates the agency rejected records that are considered public. The middle grades represent middle-ground outcomes: records released after significant delays, records available only for in-person review, denials for requests requiring a balancing test and more.
Our Rubric
Each request was scored from 0 to 10 based on how the agency handled it. The transparency score is the average across all requests for that agency.
| Score | Critreia |
|---|---|
| 10 | Immediate electronic access, little to no redaction |
| 9 | Timely access with minimal redaction |
| 8 | Delayed access after back-and-forth |
| 7 | Access, but required in-person pickup or residency |
| 6 | Access with excessive redaction |
| 5 | In-person pickup required, documents not retrieved |
| 4 | Process confusion, documents not provided |
| 3 | Rejection on records requiring balancing test |
| 2 | Acknowledged but never fulfilled |
| 1 | Rejected records routinely considered public |
| 0 | No response at all |
A municipality or school district’s scorecard score is the average across its requests, rounded to two decimal places.
A subset of requests fell outside what the rubric could judge automatically. We scored those by hand or excluded them from the average. The per-request breakdown on each scorecard notes which were excluded and why.
School districts that serve more than one town in our sample are each paired with a single municipality we also assessed. Towns and school districts share the same scoring rubric.