Reading two years side by side

Jurisdictions join and leave the dataset, and the fields they publish change as their systems change. Add up two years of raw records and part of the difference is building activity while the rest is coverage.

A number built that way is sensitive to a single jurisdiction's publication schedule, which is not something a reader can see or judge.

Where this site states a year-over-year change, it counts only the jurisdictions represented in both years, so the same places sit on both sides of the comparison. The panel below shows the dataset-level figure and the parts behind it.

This is the only form of growth statement on the site. It is generated in one place so the same wording, and the same caveat, appear every time.

Within that comparable read, the categories behind this site's services moved in both directions. Plumbing fixtures and openings rose; outdoor structures and landscape irrigation eased.

A category moving down is a category with fewer filings, not a category with less work to do. Several of the categories that eased are ones where the wording on applications is least stable.

The comparison is between two single years, so it inherits whatever happened in either of them. A storm season, a change in a filing rule or one large jurisdiction's publication schedule can move a category.

Use it as context for a decision, not as a forecast of one.