A thorny history of race-based statistics: DiManno

On the risks of not collecting race-based statistics:

Now it matters in a reverse context: Blacks killed by Toronto cops. How many? Who? Why? And so it’s the cops who dread disclosure. Yet those are facts we damn well should know.

We’ve put ourselves into a moral and intellectual bind. The reason: We don’t trust facts. More crucially, we don’t trust how facts can be manipulated. Or we don’t trust the potential for what’s known in court as an adverse inference — the reason why judges often kick out evidence deemed highly prejudicial with little probative value.

Surely, in a matter of such gravity — minorities killed in interactions with police — we can discard the extremes of compulsory blinkering. I’d say drop them entirely and live with the consequences. Because we can’t control how information will be computed by another person’s brain — whether they’ll misunderstand, whether they’ll take offense where none was intended, whether they’ll hurl accusations of racism (or sexism, Islamophobia, anti-Semitism, etc.) — we tread delicately, self-censoriously. So we end up with this: institutional timidity.

Among the numerous task forces into Toronto police and race relations was an audit released way back in 1992, which included the following recommendation: That the Toronto Police Services Board “reconsider its policy as to maintaining statistics which identify race and consider a policy which permits the maintenance of such statistics for the purpose of measuring or evaluating police activity… A civilian function should be created to maintain, compile and analyze such data. Statistics should be kept at a level of detail which allows for valid statistical conclusion.’’

The board rejected it.

A further recommendation suggested a formula for addressing racial subtexts — in policing, not the criminal constituency: “A series of indicators be developed using a statistical base to evaluate the level of bias, if any, in policing activities for the Force as a whole as well as to provide internal comparisons within the Force between differing operating units.’’

In essence, the audit pleaded for profiling — of police.

But profiling is such a dirty word. Just like “race-based” statistics. Except when they’re not.

A thorny history of race-based statistics | Toronto Star.