Harvard University

Why there is no single total here. The FBI publishes two breakdowns of the same assaults — by weapon, and by how the officer was deployed — and nationally they do not agree: the assignment columns sum to 1,869,337 and the weapon columns to 1,744,852, and the two match on only 84% of rows. Police Census publishes the weapon breakdown as the FBI states it rather than picking one basis and calling it ‘the total’. The file’s ‘cleared’ column is not shown either: it sums higher than the assaults themselves, so it is not a subset of what is above and we have not established what it counts.

This comes from the FBI’s Law Enforcement Officers Killed and Assaulted programme. Despite the name, the file carries no officers-killed data at all — those columns are empty for every agency and every year, so this page reports assaults only and says nothing about officers killed.

LEOKA identifies agencies by name, not by the ORI used elsewhere on this site. A row was attached only where exactly one agency in that state reduces to the same base name; 78.7% of rows matched nationwide. An agency with nothing here may have reported no assaults, or may be one of the names that did not match — Police Census does not guess which.