What ideological journalists say about what the data say about police shootings.

A striking example of biased reporting. From What the data say about police shootings How do racial biases play into deadly encounters with the police? Researchers wrestle with incomplete data to reach answers. by Lynne Peeples.

This comes across more as a gun control advocacy article than it does necessarily an anti-police article but most of all it is a piece trying to advance the theory of racial bias.

Peeples has two real headline issues. First is that police departments collect data in disparate fashions. Data is sparse, fragmentary, possibly unrepresentative. Second is that research on officer involved shooting is recent, inconclusive, and plagued by differences in research intent, and by differences in definitions and measurement.

Both these things have been known for decades. It is not new news. Strategies and actions to address both issues have been going on for years.

The main problem with Peeples' reporting is what is unsaid in the article. There is great focus on the inadequacy of data and the sparsity of quality research. It is not explicit from the reporting that most research in the past five years has had mixed conclusions as to whether there is disparate difference in civilian deaths by race from police shootings. Nor does Peeple mention that the more rigorous the research methodology, the more likely it is to find that there is no difference. That there is no racial bias.

In general there are four key items which Peeples does not address which should be in her reporting were she actually a qualified journalist.

The first issue not mentioned is widely known in the research community. The propensity towards violent crime manifests differently across races. Very roughly, African-Americans are 13% of the population and commit 50% of the murders. Hispanics are 15% of the population and commit 25% of the murders. Whites are 70% of the population and commit 25% of the murders. If one takes propensity to commit murder as a proxy for probability of interacting violently with the police, then one would expect 50% of police shootings to involve African-Americans. Peeples skirts this issue completely. She does not acknowledge it at all. She instead focuses on rates of general population, not rates by propensity toward violent encounters. When you compare apples to apples (based on propensity towards violence using murder as a proxy) - the disparate rates of police shooting by race disappear. It is analogous to comparing lethal car accidents of children and adults and determining that adults must be very dangerous drivers indeed (when in fact you are not comparing relevant populations.)

The second omitted issue is that if one is going to focus on racial bias motivation, and given that we have multi-racial police departments, one would want to look at whether African-American officers are less likely to have a lethal encounter with African-American civilians than their white counterparts. If racial bias is a motivation, then one might assume that white police officers are more likely to shoot African-American civilians than African-American officers. Peeples does not ask this question. Presumably because all the findings so far are that there is no difference in probability. Police officers of any race have the same rates of lethal encounter. Which of course suggests that racial bias is not the root issue. Some departments are clearly much more effective in training their officers to deescalate encounters than others. It involves real training, not racial bias training.

The third omitted issue is the source of police deaths by race. Peeples looks at number of police killed but does not look at the race of who killed them. The research on this much murkier than the above two questions but again the data so far seems to indicate that police are more likely to be killed by African-American civilians than by civilians of other races.

The fourth omitted issue is all the findings to date which indicate we do not have a police racial bias issue, we have a police violence deescalation training issue.

The last is perhaps the most egregious though they are all egregious. If racial bias is not a root cause, then providing racial bias training will not reduce the number of lethal police encounter deaths. Peeples is misdirecting us from the real root causes of officer-involved shootings. Shame on her.

And shame on Nature magazine for running such a shoddy article which so clearly smacks of ideology over real world problem solving.