Wide Open Spaces

The analysis of social distancing continues with a comparison of COVID-19 pandemic growth rates depending on how urban or rural a state is.

A previous post raised the question of why some rural states that delayed implementing stay-at-home orders did not seem to suffer the same explosive growth rates as states that locked down sooner. It is plausible to assume, as some states have in crafting their policies, that rural communities are naturally self-distancing. To quantify this effect, we consider the impact of population density on case growth. Specifically, we plot case growth versus the percentage of the population that lives in urban centers:

As with the previous social distancing post, we assess the consequences of government decisions using a snapshot of the five-day growth rate 28 days (4 weeks) after the state crossed the 10 cases per million people threshold.

The graph shows a clear correlation between how dense the population of a state is (% Urban - see definition below) and how fast the pandemic is growing in the state, and in fact this correlation is even stronger than the previously presented data on active social distancing policies (school closures and stay-at-home orders). Some states have been able to benefit both ways. For example, Oklahoma may not have implemented a state-wide stay-at-home order, but the major metropolitan areas (Oklahoma City and Tulsa) did have their own local lockdowns, which has no doubt kept the overall state growth rates down. This type of leaky box strategy may have worked (so far) in rural states that have experienced the pandemic later, but clearly it failed in the Northeast where people fleeing New York City carried the virus to upstate New York and to neighboring states.

It is worth again pointing out the huge outlier that is South Dakota, whose Governor, Kristi Noem, foolishly refused to allow its major metropolitan area, Sioux Falls, to impose a stay-at-home order when there was a major outbreak at a Smithfield food packing plant. Good governance requires not thinking that one size fits all, whether in the city or in the country. Even for those states with more pragmatic approaches, it is also worth pointing out that doing relatively better than the rest of the country is not the same as having the pandemic under control, which requires getting the number of active cases to a steady state that does not exceed the ability of medical systems to care for patients, a threshold that no state has yet reached (although some are close).

Given how charged the politics of the pandemic have been in the US, let's go one more step and layer on the political leanings of each state. Since states' policies are largely driven by their governors, we have labeled each state by whether the governor is a Republican (red) or a Democrat (blue):

For reference we have plotted the trend line between % Urban and 5-day growth rate, as well as the average values for Republican and Democratic states (and DC).

As expected, rural states lean towards electing Republican governors and urban states lean towards Democratic governors. Otherwise, both groups follow the same trends with regard to outcomes, namely the closer people live together, the more interactions they have and hence the more each infected individual infects others. States' outcomes may start to diverge as they discontinue stay-at-home orders, in many cases for political rather than sound epidemiological reasons (e.g., two weeks of case declines), so we will revisit this topic in a couple of weeks.

Pundits like to make the pithy comment that viruses don't respect state lines and therefore we should have uniform policies across the country. This would be true if there were only a small number of cases and we could trace and isolate every contact, but we are long past that point (and evidence continues to mount that we were past that point in January if not December). Instead the best approach right now is to achieve a sustainable case level that doesn't overwhelm our hospitals. This approach necessarily allows some transmission of the virus from city to city and state to state. Further, as much as we would like clear and sound guidance from the federal government, it is reasonable that policy coordination has devolved to the regional level, because, for example, people may be circulating freely in the Tri-State area or on the West Coast, but much less between those regions.

Pundits also like to characterize the pandemic along political lines by saying that the reaction is different depending on whether you are a Democrat or a Republican, but in fact because of the correlation of both growth rates and political leanings with population density, red states do in fact have a different current reality than blue states. These realities will converge when the pandemic fully spreads to rural America, but for now there are only a few lily pads on the pond.

Definition of % Urban

The Census Bureau’s urban-rural classification is fundamentally a delineation of geographical areas, identifying both individual urban areas and the rural areas of the nation. The Census Bureau’s urban areas represent densely developed territory, and encompass residential, commercial, and other non-residential urban land uses. The Census Bureau delineates urban areas after each decennial census by applying specified criteria to decennial census and other data.

The Census Bureau identifies two types of urban areas:

• Urbanized Areas (UAs) of 50,000 or more people;

• Urban Clusters (UCs) of at least 2,500 and less than 50,000 people.

“Rural” encompasses all population, housing, and territory not included within an urban area.