Oh, to build a report from numbers rather than narrative

From A Deep Dive into Mass Shooting Data (Rachel Maddow Hardest Hit) by Kevin Downey, Jr.  There is a lot to dislike about this reporting, mostly the gleeful gotcha's and partisan vituperation.  Downey appears overly concerned about the racism of the media.  But . . . he has data.

The whole argument of what constitutes mass killing and mass murderers is plagued definitional issues, poor writing, innumeracy, etc.  

For this article I’m using the same definition used by the Gun Violence Archive: four or more people shot, not including the shooter, in a fluid situation. This can include drive-by shootings, gunfights between gangs, or two or more shooters redecorating a bar. 

Downey seems concerned that 

The race-obsessed liberal press has long pushed the belief that white guys in pickup trucks are cloaking themselves in Trump flags and mowing down former workmates with illegally purchased Terminator street-sweepers with optional bazookas. 

Way too provocative to the point of undermining his argument.  Does the media make out mass shootings as a strongly white male issue?  Yes, I think that is true though I don't recall having seen a solid analytic documentation.

Part of the issue is that of definitional ambiguity.  If we are talking about mass murderers (which may accumulate over time), then the media narrative is wrong.  There have been a handful of studies on this phenomenon and they almost always find mass murders in proportion to population.  If whites are 65% of the population, they are 65% of the mass murderers, etc.  

With mass shootings (occurring in a single instance) the dynamics are far different as are the outcomes.  As Downey documents for March 2021.  

There were 45 mass shootings in March 2021.  Downey looks at the details of all news accounts and can explicitly identify the race of shooters in 44% of the cases.  When he goes further allowing context to infer likely race of shooter he was able to make a determination in 90% of the cases.

The results are:

30.5 black or likely black shootings (67.7%)

7.5 Hispanic or likely Hispanic shootings (16.6%)

2 white or likely white shootings (4.4%)

1 Muslim (2.2%)

4  too little info to make a call (8.8%)

(Reminder: one shooting had two shooters, one black and one Hispanic)

One month is an inadequate data set, inference is not a confidence-inspiring approach, playing the race gotcha game is repugnant.  But this data is not too far out of line with lots of other data.

The mundane and boring nature of mass shootings makes it routine and unmemorable and not reported.  Yet it is in the aggregate a significant absolute number.  White males as the archetype of mass shooters is incorrect (though it gets closer if you raise the body count, i.e. a massing shooting defined as 10 or more people shot.)  

As long as we continue to foster false stereotypes, it occludes clear thinking about root causes and viable alternative solutions.  

While not accepting that the above data is fully accurate, it does raise questions.  Why is the plurality of mass shootings in big cities?  Is race a causal factor (I think not) or is it a dependent factor (I suspect so.)  If only a dependent factor, what are the direct causal factors?  Almost certainly neighborhood disfunction?  What are the alternative viable solutions to neighborhood disfunction?

We only get to these good questions when we move away from stereotyping and narrative spinning.  We need numbers to begin to understand the realities of problems we do need to solve.