Methodology
How the data is built
Every figure on this site is extracted from public records and linked back to the document it came from. Here is exactly where the data comes from, how it's processed, how meetings are scored, how often it updates, and what it can't tell you.
Sources
Two public sources, used for different things:
- Utah Public Finance Website (transparent.utah.gov) — annual departmental totals, vendor-level payment totals, and employee compensation reported by each city to the state.
- City council meeting packets — the agendas, staff reports, and check registers each city publishes on its own website or agenda system (BoardDocs, CivicClerk, Granicus, RainPos, and others).
The extraction pipeline
- Discover. For each city we read the published meeting list and record pointers (URLs) to each agenda and minutes document. We store pointers, not copies.
- Parse. Council-approved check registers are parsed into individual payments; the state feeds are parsed into vendor and salary tables. Ledger accounts are mapped to funds (General, Water, Sewer, Capital, and so on).
- Link. Every payment, score, and figure carries a link to the source document, so any claim can be verified at its origin.
What we index per city
| Dataset | What it is |
|---|---|
| Payments | Council-approved warrants from the check register, by vendor and fund. |
| Vendors | Per-vendor annual totals from the state's vendor feed. |
| Salaries | Employee compensation reported to the state. |
| Budget & reports | Adopted budgets and audited financial reports (ACFR). |
| Meetings | Indexed council meetings with agenda and minutes pointers. |
| Scores | Procedural-transparency scores for reviewed meetings. |
| GRAMA | Ready-to-send public-records request templates. |
How meetings are scored
Scored meetings get a 0–10 procedural-transparency rating. It measures whether a decision was made in a way a resident could follow, not whether the decision was good. Three things drive the score:
- Agenda focus — does the agenda say what will actually be decided, or hide it behind vague language?
- Consent bundling — how many separate decisions are stacked into a single consent vote with no individual discussion?
- Scrutiny of major items — do large contracts and ordinances get their own line and their own discussion?
A caveat on "Not Applicable" spending
About half of the spending the state reports has no named vendor — it's filed as "Not Applicable" (payroll, statutory redaction, or under-threshold aggregation). We report named-vendor totals honestly and never present the redacted portion as if it were itemized. We wrote up what that label means, and what cities could do to reduce it, in this analysis.
How often it updates
The dashboards are refreshed on a regular automated schedule that re-pulls each city's data from the public sources and rebuilds the site. Each city page shows its data coverage range.
Limitations & corrections
The data is only as complete and timely as the public sources. State feeds lag, cities post minutes weeks after a meeting, and aggregates can hide detail. Where a city's reporting is incomplete, we say so rather than fill the gap with guesses. If a figure looks wrong, check its linked source first, then send a correction through the contact form on the home page.