http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0141854
"There is no relationship between county-level racial bias in police shootings and crime rates (even race-specific crime rates), meaning that the racial bias observed in police shootings in this data set is not explainable as a response to local-level crime rates."
Are you familiar with what Bayesian implies?
From the same link:
"Understanding the source of racial bias in police shootings is difficult to do from county-level data, as the ecological inference fallacy can potentially obscure any results [39]. County-level data are far too coarse to use to reliably tease apart the conditions that drive racial bias in police shootings; "
"More detailed analysis of the data is needed. In the current analysis, county-specific risk ratios are estimated. In the future, these estimates could be extended in a Bayesian framework to include estimates unique to police departments, as clustered into counties, as clustered into states."
"Ecological regression on county-level characteristics is plagued by difficulties theoretically [39, 51]; issues with data quality make it even harder to use county-level data. In the analysis of county-level predictors of racial bias in police shootings conducted in this paper, some of the data were low quality. Notably, the crime data may be biased by the reporting practices of the police, and Florida, Alabama, and Illinois failed to fully release data, which led to the use Bayesian imputation for counties in these states."
"is driven by race-specific crime rates (at least as measured by the proxies of assault- and weapons-related arrest rates in 2012)."