In the previous post, I described my strategy for finding hints about obstacles to voting by Black people, and I presented some results from large urban areas. Today I’m presenting the same kinds of results for medium-sized urban areas.

In my categorization, these areas had at least 50 Black respondents who were United States citizens at least 18 years of age in the November 2018 CPS. As a reminder, I’m looking to see how registration rates translate into voting rates for the Black community versus White voters in the same area, with simple controls for age, gender, income, and education.

The chart above shows that in quite a few medium-sized urban areas, higher registration in the Black community did not lead to as much voting as expected (the red lines). The biggest gaps were in Pine Bluff, Columbia, Dover, Montgomery, Nashville, Charlotte, San Francisco, Raleigh, and Riverside.

Among these areas, several had much higher registration rates that did not result in higher voting rates, at least in comparison to the expectations given my controls. These were Pine Bluff, Montgomery, Cleveland, Columbia, and Dover.

The Pine Bluff area is particularly interesting. In the 2010 Census, the city itself had only about 49,000 residents, about 76% of whom were Black (roughly 37,000 people). So it was not a very big urban area, but it did have a lot of Black respondents to the CPS. Jefferson County, of which Pine Bluff is the county seat, had about 77,000 people, of whom 55% were Black (roughly 42,000 people). This means the other towns in the county were mostly non-Black, but Pine Bluff was the center of political power. In fact, most of the elected officials of the city and the county were and are Black.

With a dominant position as far as ethnic representation, did Black people not feel the need to vote as urgently as in other areas? They might have felt that their rights and interests would be better protected than in other areas where elected officials might have been more likely to show prejudice. Similar dynamics might be at work in Montgomery, where about 55% of the residents of both the city and county were Black in the 2010 Census. In Cleveland, the share was 53%. It was 42% in Columbia and Dover, and lower everywhere else.

That leaves Charlotte, Nashville, Raleigh, San Francisco, and Riverside. All of these areas had smaller Black communities, and they all had lower voting by Black respondents than expected, given my controls. The difference was particularly stark in Nashville and Charlotte. These look like red flags to me. Moreover, the fact that both Raleigh and Charlotte – the two biggest metropolitan areas in North Carolina – show up in this group is rather concerning. But the state changed its voting laws before the 2020 election, so a comparison to November 2020 CPS data (when they become available) may be quite revealing.

In San Francisco, the dropoff in voting was rather small, in fact with a similar slope to the data from Pine Bluff. Perhaps Black people, who made up just 6% of the population in the 2010 Census, also felt less urgency to vote in such a liberal community, with a long history of upholding civil rights. Some 400 miles south in the Riverside area, however, the steeper dip looked like the ones in Nashville and Charlotte. Las Vegas (shown in blue above, as registration was slightly lower among Black respondents) also had a large gap between registration and voting outcomes, almost all of which was due to lower-than-expected voting.

These results are still more suggestive than definitive, but they do make me want to research potential barriers to voting that might affect the Black communities in these five areas: Nashville, Charlotte, Raleigh, Riverside, and Las Vegas. In the next post, I’ll look at smaller towns and rural areas.