New Hampshire: the last 100 land-use decisions we read

decided means we read it in the minutes. The body voted and the outcome in the action column is what passed. 26 of the 100 rows here.

scheduled means it was on a published agenda and had not been voted on when we read it. 74 rows. These are the early ones, and they are the reason to read agendas rather than permit databases, but calling one of them approved would be wrong.

The address column is a hint, and roughly a third of the filled ones are wrong. We hand-checked 40 rows: 24 correct, 4 truncated, 12 wrong. The failures are real addresses from the right document attached to the wrong item on a long agenda, which no blocklist fixes. So every address ships with the verbatim snippet it came from, in the CSV, and 60 of 100 rows on this page have one at all. Read the snippet before you drive anywhere.

The applicant column is only filled when the document says so. 13 of the 100 rows here name one. Unlike the address, nothing is inferred from position: if the agenda or minutes did not print “Applicant”, “Petitioner” or “represented by” next to a name, the column is empty. That is why it is filled on about one row in nine rather than most of them, and it is the trade we chose on purpose. Where it is filled it often names the engineer, surveyor or developer behind the project, which is usually the more useful half of the row.

What this is not. It is not a contact database and there is no parcel owner or phone number in it. It is not a permit feed, and it does not claim to be every decision in New Hampshire — it is every decision in the 27 jurisdictions we currently read there, out of 2530 nationally. When a government changes agenda vendors, our coverage of it breaks until we notice.