Can systemic racism be found in data related to fatal police shootings? Along with anecdotal evidence, many have based widespread protests on the premise that it can be found in the stats.
But have you seen the unvarnished numbers from their sources? Instead of blindly following the interpretations of activist leaders, Facebook warriors, or corporate newsrooms; try crunching the data for yourself and arrive at your own conclusions.
Here’s three resources widely used to make racial bias determinations:
Started in 2015, this database documents police shooting numbers of all types, collecting information from news articles, social media entries, and police reports. It has nine search filters with numerous selections that allow for literally thousands of variations. This is where calculations would begin—with fatality totals for whatever stipulations are applied.
To make accurate comparisons it’s necessary to give full context to the database numbers. The most common of these contextualizations is to determine the rate of occurrence in each demographic by compensating for the difference in population. Pretty basic but I would remind you that in many racial groupings “White” includes hispanics, as well. The census has that number at 76.5%—along with a number that is “White/Not Hispanic” which is 60.4%. These numbers are often confused, but it’s safe to assume most are looking to make a non Hispanic, Black and White comparison. In which case, these numbers show that Whites have 4.6 times the population of Blacks.
The second step to contextualize the data is to determine the frequency of “high stakes” interactions between the demographics and police. More frequent deadly scenarios in shared proximity raises the amount of opportunities, which increases the chances for fatalities. There are several metrics that could be used to determine such frequencies but “Homicide Offender” statistics seem effective in quantifying interactions in “life and death” circumstances—where it’s reasonable to expect fatal force to be more common. Just like with the Census data , “White” numbers include Hispanics and, in this data set, have to be manually deducted if you’re comparing Blacks and Caucasians, which is what most mean when making a Black and White comparison.
Crunching The Numbers…
Now that we’ve identified the three sources, here’s an example of how to calculate the data. We’ll use simplified, made-up numbers and generic demographic category names. So, when you make your own calculations they won’t be biased by anything written here.
Let’s say…last year “Group A” experienced 10 unarmed police shooting fatalities and “Group B” experienced 20. In our list of three resources these numbers would come from the database in section one.
Group A 10
Group B 20
Now we need to inform these numbers by expressing the difference in population using the data from the census numbers. Since the “Group B” population is, hypothetically, 4x as large as “Group A” we enter that variable and compare the difference, giving us the proportional rate of police shooting fatalities.
Group A 10 x 4 = 40
Group B 20
Difference 2:1
This shows that “Group A” is twice as likely to be shot and killed by police. This is where most, especially in the media, stop calculating because the math supports their “unfairness” narrative.
However, desiring the full picture, we’ll also need to account for how many opportunities exist for police shootings to occur in each of the two groups. I realize this is nuanced for some, but we’ve already accounted for the amount of police actions in the first section. So, we also need to account for how often members of those groups are available to be affected by those police actions. In anything, odds dictate that more chances produce higher likelihoods.
To ascertain the frequency of these opportunities in each group we’re using fictional Homicide Offender data. “Group A” commits 2000 annual homicides—“Group B” 1000.
Group A 2000
Group B 1000
But once again we have to return to the Census data in section two to determine the rate—using our simplified factor of four to account for population difference.
Group A 2000 x 4 = 8000
Group B 1000
Difference 8:1
This shows that “Group A” is eight times as likely to be present in a homicide interaction as “Group B”. What that means is “Group A” is extraordinarily more likely to interact with law enforcement in “live or die” scenarios.
What this hypothetical example concludes is: Although Group A is twice as likely to be fatally shot by police, they are also eight times as likely to commit murder—linking them with a much higher probability of police shootings simply because of exponentially more opportunities.
Do the math for yourself. Plug in the values and draw analytical conclusions. But please do all four of the calculations. If you stop at the first two then you’re only solving half the equation. In an era where facts are being called racist we need the rationality of science more than ever. Good luck.
No comments:
Post a Comment