Methodology

Everything on this site is computed from the City of Chicago's public 311 Service Requests dataset, re-aggregated nightly. No raw records are stored; the city's API computes the medians.

What "days" means — read this one

Every number is the median time from a request being created to being closed in the 311 system. Closed does not always mean fixed: the city Inspector General's February 2026 audit found requests marked completed while the underlying work was not done. Treat these numbers as how long the city takes to consider your request handled — a meaningful but imperfect proxy for service.

What's excluded, and why

• Duplicate requests (~5%). The city closes them instantly against the original; counting them would flatter neighborhoods that report the most.
• Information-only calls (about a third of the dataset) — not service requests, and geocoded to the 311 call center.
• Canceled requests — closure without service isn't a response time.
• Small samples. Map areas under 20 requests for the selected type and period show gray; report-card rows under 30 are omitted.

Comparisons

Wait times are only ever compared within the same request type — a graffiti day and a tree-trim day are different units. "X× slower" compares a neighborhood's median to the citywide median for that type and period; verdicts between 0.67× and 1.5× read "about average." Neighborhoods are the 77 community areas (stable since the 1920s) rather than wards, which are redrawn each decade and are drawn for politics, not geography. Departments come from the city’s own owner_department field on each request.

Known limitations

• Medians of closed requests understate problems in areas with old open backlogs — that's what the "still open" column (count · average age) is for.
• Research on 311 systems (Liu, Bhandaram & Garg 2023) shows under-reporting itself varies by neighborhood, so closure-time gaps likely understate true condition-duration gaps.
• The city occasionally renames or retires request types; when a type disappears from the nightly build the previous data is kept and the discrepancy investigated rather than silently published.

Questions or corrections: see about.