Jurisdiction articles show indicators computed from traffic fine lookups processed by systems operated by APINITRO's authors. They are not official statistics: they describe what the consulted sources returned during a period, for a sample that depends on who ran the lookups and where.

What is measured

Each indicator is computed over two universes: the fine records observed in a jurisdiction and the lookups that included it. From them we report shares (for example, lookups with at least one open fine), category mixes, amount bands, the median age of open fines and monthly seasonality. The period of each dataset and its generation date appear next to the figures.

What is never published

  • License plates, fine numbers, identity documents, names, session identifiers or customer data.
  • Exact amounts: monetary values are reported only as bands.
  • Figures from cells with fewer than 50 observations; a jurisdiction is published only when it gathers at least 100 fine records and 100 lookups in the period.
  • Per-customer or per-organization metrics, and the names of the source systems.

Rounding and suppression

Shares are rounded to multiples of five points when the denominator is below one thousand observations and to one point otherwise. Durations are rounded to tens of days. Sample sizes are reported as ranges rather than exact values. If publishing an indicator would allow a suppressed cell to be derived, it is suppressed as well; the list of suppressed indicators is part of every dataset.

Periods and versions

Each dataset covers a period of at least six months and carries a methodology version. When it is regenerated for a new period, the article citing it is updated and records its modification date. Earlier versions are not kept public.

Limitations

The sample is not random: people who look up a plate usually suspect it carries fines, so the share of open fines tends to be higher than in the general vehicle fleet. Source coverage varies by jurisdiction and over time; an unresponsive source is counted as a source error, never as an absence of fines. Infraction categories are assigned with manually reviewed text rules and may group heterogeneous descriptions.