Like everything else these days, policing and the fairness of the criminal justice system have become fault lines that divide liberals and conservatives.  A controversy that has long been at the center of this debate is the extent to which efforts should be made to ensure that the racial makeup of local police forces matches that of the populations of the cities they serve.

In an impressive feat of data reportage, FiveThirtyEight’s Batya Ungar-Sargon and Andrew Flowers compare the racial diversity of the police in major U.S. cities with that of their residents. Their analysis is based on numbers maintained by the U.S. Equal Employment Opportunity Commission on the percentage of each major racial group (African Americans, Asian Americans, Latinos, and whites) among those employed in urban police forces. They examine the extent to which these percentages corresponded with the Census Bureau’s figures on the share of each group making up the populations of these cities.

Building a police force that perfectly mirrors a city’s population down to the last officer is impossible. But some cities do come close to this ideal. As shown here, in Los Angeles, New Orleans, and Washington, the shares of cops belonging to each racial group are very similar to the groups’ corresponding percentages in these cities’ populations. However, cities like these are the exception: in almost all large cities, white officers are over-represented on police forces.

To what extent is this variation in diversity due to politics? Research by political scientists demonstrates that many city policies follow the liberal-conservative divide. In recent work, UCLA’s Chris Tausanovitch and MIT’s Christopher Warshaw use survey data to develop a measure of the liberalism and conservatism of the residents of large U.S. cities. They find that cities’ policies on taxes, spending, and a range of other issues are highly correlated with where their residents stand on the liberal-conservative spectrum.  Given that liberals tend to be more committed than conservatives to efforts explicitly designed to achieve racially diverse workplaces and schools, one might expect liberal cities to be more likely to have police forces that match their populations’ demography.

But as shown in the figure above, racial diversity is not found exclusively in the police forces of America’s most liberal cities.  The graph plots the unrepresentativeness of each city’s police force against the political views of city residents. The smoothed green line traces the relationship between these two variables. (Only cities for which data on both measures were available are included in the figure.) While very liberal cities such as San Francisco, Seattle and Washington have police forces that look like their populations, the same is true for conservative cities like Colorado Springs, Jacksonville and Virginia Beach. If anything, it is cities in the middle of the political spectrum that are most likely to have a police diversity problem: those with the most unrepresentative police forces have relatively moderate residents.  This evidence is only suggestive, falling short of statistical significance (if you’re keeping score, a test of a quadratic fit between the two variables has a p-value of .15).

In sum, among America’s big urban areas, liberal cities are no more likely to have diverse police forces than conservative cities. Why might this be the case? One possibility may be the varying degree across cities to which police unions are powerful. In conservative cities, public employee unions are typically weaker and thus less able to block efforts at diversifying the police, including residency requirements and expanded recruitment. Another hypothesis is that military bases tend to contribute to a place’s conservatism, but they also provide a ready supply of diverse individuals who are veterans — a top source of new police officers. The most likely explanation, however, is that until recently police diversity was not high on the agenda in many cities. Voters weren’t paying much attention to the issue, and thus city officials pursued diversity policies for reasons unrelated to their residents’ liberalism or conservatism. That may change over the next few years if the spotlight of controversy continues to shine on the lack of diversity in most big-city police departments.

Diversification of police forces is no magic solution to our current controversies surrounding violent encounters between cops and citizens. For example, as FiveThirtyEight’s Ungar-Sargon reports, evidence is mixed that non-white police officers are involved in violent encounters any less than white officers. But common sense suggests that police forces will be more effective and more respected to the extent that their racial makeup is similar to that of the city residents they serve. For now, this circumstance is just as likely to be true in our most conservative big cities as in our most liberal ones.