The curve is flattening for wealthy white communities, not for poor non-white communities

A month ago, I started looking at COVID-19 data from cities in southern California and distinguishing the rates of laboratory confirmed cases by community characteristics such as race, income, and age. At the time I found that confirmed cases of COVID-19 are most prevalent in wealthy and white communities. But that finding no longer applies.

This was the relationship between the laboratory confirmed COVID-19 case rate and a city’s median household income for cities in southern California as of March 30, 2020.

Cities with median annual household incomes above $110,000 had a COVID-19 case rate of 34.3 cases per 100,000 people, which at the time was over four times greater than the case rate for cities with median annual household incomes below $50,000. To me this gave an indication of which groups of people had access to testing, which was rarer at the time.

Four weeks later, the relationship between a city’s household income and its COVID-19 case rate has been eliminated.

While the case rate for the wealthiest cities has almost tripled from 34.3 to 101.8 cases per 100,000 residents, the case rate for the poorest cities has gone up 18 times, from 7.5 to 134.8 cases per 100,000 residents.

On March 30, there was a strong relationship between cities with high percentages of non-Hispanic white residents and cities with high case rates for COVID-19.

Coronavirus rates for the whitest cities were four times greater than the rates for cities where non-Hispanic whites made up less than half of the population.

This has now changed. As of April 25, there is no clear relationship between a city’s COVID-19 rate and its percentage of non-Hispanic white residents.

Now cities with very few non-Hispanic white residents (e.g. Bell, Carson, Inglewood, Santa Ana, Coachella, etc.) have a significantly higher case as a group than do cities with high percentages of non-Hispanic white residents (e.g. Calabasas, Hermosa Beach, Laguna Beach, Rancho Mirage, Encinitas, etc.).

When graphed over time, the flattening of the curve is quite apparent for wealthy, white cities after an initial surge in case rates. However, the curve does not appear to be flattening for cities with low median annual household income or low percentages of non-Hispanic white residents.

Up until about a week ago, wealthy, white cities had a higher rate of laboratory confirmed COVID-19 cases per 100,000 than did poor cities with few white people. But the change in the slope of the curve (the growth rate) started to flip around a few weeks before that, in early April.

One of the striking things about these charts to me is that it shows that the mitigation efforts and stay-at-home orders declared by southern California health agencies have been most effective in wealthy and white communities and not tremendously effective in poorer communities with few white residents. The daily growth rate a month ago was over 5 new confirmed cases per 100,000 per day for wealthy communities, but now it has decreased to 1.5 new confirmed cases per 100,000 per day. But the daily growth rate for poor communities has increased from around 3 new confirmed cases to (at least for the last five days) 8 or more new confirmed cases per 100,000 per day.

It is difficult to say how much of this effect is real due to many issues regarding coronavirus testing. The Los Angeles County antibody study found that there were 28 to 55 times more people who have had COVID-19 than have been determined through a census of laboratory confirmed data. If the people tested so far are not a representative sample of the people who have COVID-19, then it would be difficult to conclude anything policy-wise. But if these results can tell us anything, they are telling us we need a more effective curve-flattening policy inclusive of all communities.

A note on the data: COVID-19 case numbers come from the county health departments of Los Angeles, Orange, Riverside, San Bernardino, San Diego, and Ventura Counties. Population figures come from California Department of Finance estimates for January 1, 2019. Demographic data is from U.S. Census Bureau Quick Facts, which used July 1, 2018 data at the time I first compiled it. Cities include every incorporated city in Los Angeles, Orange, Riverside, San Bernardino, San Diego, and Ventura Counties, except for the cities of Los Angeles and San Diego (because their large diverse populations cannot be easily characterized by one number), and cities with populations under 5,000. Rates for groupings of cities are weighted by population, not just an average of all cities within the group.

I am not actually a train. Or a roadrunner. But I am originally from New Mexico.