Ok, let’s talk about the data here because the actual disparity is much higher than 3.6x. A thread (1/x).
Data on racial disparities in marijuana arrests or any other arrests at the national level underestimate how much disparity there is. That’s because the nation’s arrests data program (the UCR) is run by the FBI. And the FBI‘s methods artificially reduce disparity. Here’s how...
Up until a few years ago, the FBI UCR program didn’t even collect data on Latinx people arrested. Instead, Latinx people were categorized as “white” in 85%+ cases which means arrest disparities are calculated as Black arrests divided by both white *and Latinx* arrests.
So if Black people are 3.6x more likely to be arrested for marijuana than white + Latinx people, it means Black people are more like 2-3x more likely than Latinx people and 4-6x more likely than white people to be arrested for marijuana.
But the precise disparity is larger than what is reported publicly because the data the FBI publishes doesn’t specify how many white or Black arrests are also Latinx arrests, so it doesn’t allow any organization to calculate the actual disparity levels between the these groups.
This isn’t an accident. It’s a decision by the FBI to report data in a format that makes it impossible to precisely calculate arrests rates for Black vs white people. Instead, the disparities that can be reported make it look like police are less discriminatinatory than they are.
Over the past few years, the UCR has collected data on the number of Latinx people arrested. But since the arrest totals by ethnicity are reported separately from arrest totals by race, we still can’t determine exactly what the Black vs white or Latinx vs white disparities are.
A simple fix that the FBI could implement would be to require each agency to specify how many Latinx arrests were categorized separately as white or Black. But they haven’t done this/haven’t indicated that they’ll do it moving forward.
Until then we won’t be able to accurately calculate arrest rates for Black vs white, Latinx vs white, Black vs Latinx or any other combination. This is basic, we need these data to develop policy and it’s not available to *anyone* other than FBI (who don’t do anything with it).
