Residents Count

Washington, DC · 2008 to 2026

The documented downgrading of DC crime reports doesn't show up in the public data

Investigations found officials reclassified crimes under pressure. The published totals can't confirm it: the documented edits were too few to see, and one test that seems to catch them is a statistical trap.

Between December 2025 and March 2026, a House Oversight Committee report, a Justice Department review and the Metropolitan Police Department's own internal affairs investigation found that officials reclassified violent assaults, robberies and thefts under pressure from department leadership in 2023 to 2025. The chief resigned in December 2025. In July 2026, the DC Inspector General found that MPD had no independent review of how crimes were classified from about 2013 until reforms in 2026.

This report asks whether any of that left a mark in the public data: MPD's monthly counts of nine serious offenses, published back to 2008.

The 2013 loss of review

When independent review ended around 2013, 6 of 12 measures of the offense mix shifted. But the same test run at dates when nothing happened fires just as often, up to 7 of 12. DC's offense mix shifts often, so 2013 can't be singled out.

The 2023 to 2025 downgrades

The documented downgrades point in specific directions: robberies logged as thefts, weapons assaults logged below assault with a dangerous weapon, and thefts moved to lesser categories. None of the four shifts those paths predict appears at August 2023, when the documented pressure began.

That is what the scale would lead you to expect. In the places it audited, the internal affairs report documents about 7 to 15 theft reports a month leaving the published data. Against each district's monthly thefts, those edits would lower the count by 3% to 17%, below the 14% to 18% the public data can reliably detect.

Documented downgrades against what the public data can detectDrop in monthly theft counts, three audited places and periods

Expected drop: documented edits a month divided by the district's thefts a month. Smallest detectable drop: the size that comparison districts reach by chance 5% of the time. Source: MPD reported incidents via DC Open Data; MPD internal affairs report IS #26-000051 (March 5, 2026).

A result that looks like confirmation, and isn't

A simple before-and-after comparison does show a large drop in 7th District thefts in 2024, exactly where and when the investigation documents downgrades. It would look like independent confirmation. It isn't. The 7th District's thefts had jumped 38% in 2023, the steepest rise in the city that year, and the 2024 drop is mostly that spike reversing. A test of whether the district had been moving with the others beforehand fails (p = 0.016), so the comparison can't be trusted.

What this does not show

Part of the series Policing in DC, 5 reports.