COVID: A tale of two Americas

It’s Thanksgiving, also known as DJ Khaled’s birthday, and COVID-19 is back with a vengeance. We’ve all been cautioned by the CDC to celebrate a socially distanced holidays this year, and the data shows most of us are ignoring them so far. I’m not here to lecture anyone about that — but also you’re all terrible people. 

I’m here to talk about what’s different about this so-called “Third Wave”, which has been going on since October. Specifically, there has been a stark Urban-Rural Divide, which has been wildly easy to see in the data. Rural areas, which managed to avoid a lot of the initial damage of COVID, are getting especially hammered now. And the data indicate they may not have the resources to deal with it.


Urban-Rural Divide

Measuring COVID

One of the best ways to measure the ever-evolving conditions of COVID is daily new cases. This metric is unstable, so researchers typically take a moving average over a 5-7 day period instead. Daily new cases allows us to easily visually detect the individual “waves” of COVID, as opposed to Total cases, which is always increasing by construction. The following charts illustrate the differing levels of usefulness of each measure: 


With that out of the way, let’s look at the daily new cases of COVID over time in rural areas vs. urban areas. Before presenting the graph we might expect that testing rates differ in rural areas and urban areas due to factors like access to healthcare. That means comparing confirmed cases between the two isn’t exactly fair because the number of people getting tested isn’t equivalent. I performed a rough correction for this using state-level testing data to account for this:


Here we can see that rural areas were largely spared from the First Wave of COVID and were hit equally as hard as urban areas during the Second Wave. However, during the unprecedented Third Wave there has been a stark difference between the two. New cases per capita has risen much more rapidly in rural areas and is currently roughly 0.8 per 1000 people, compared to 0.5 for urban areas. That means the virus is spreading 60% faster in Rural Areas as compared to Urban Areas right now. Obviously, this seems counterintuitive, as rural areas are less densely populated, and your average rural resident interacts with fewer people. You would imagine a virus to spread more quickly in a metropolis like Chicago than Emporia, Virginia (where nearly 0.6% of the population has died of COVID so far, good for third in the country). 

Measuring the Impact

What’s really concerning about this divide in new COVID cases is another, related divide between rural and urban areas: rural areas have less access to healthcare than their urban counterparts. This implies that rural areas may not have the healthcare infrastructure to handle the influx of patients brought in by the Third Wave, leading to a collapse of systems in certain areas. 

Ideally, we would measure this through metrics like hospitalization rates, patients turned away, etc. However, it’s impossible to collect this data nationally at a county level. At least not in the middle of the pandemic. We can, however, look at the ultimate outcome metric: mortality rates, or how many people are dying of COVID per capita. As per our earlier discussion, we must look at “new mortality rates”, or daily new deaths per capita to see what’s happening. If mortality rates are holding steady despite increasing COVID infection rates, that implies the healthcare system is probably coping with the influx of cases. If not, there is some evidence that the healthcare system is struggling to keep up:

In this chart we see a massive spike in new mortality rate, especially in urban areas, during the First Wave as hospitals struggled to both keep up with the initial influx of patients and treat the novel disease. The Second Wave, which was worse than the first in terms of the expected influx of patients, had a lower new mortality rate “bump” — indicating that hospitals adapted and became better at dealing with the virus. Remarkably, despite the unprecedented gravity of the Third Wave, new mortality rates in urban areas have risen much more slowly than new cases, indicating that healthcare systems are somewhat managing to cope with the influx of patients. That should be taken with a grain of salt — this could be attributed to a “lag effect” where it takes a couple of weeks for people to die of Coronavirus, however. 

However, in rural areas, we see another story. New mortality rates are skyrocketing with the Third Wave. Recall that the virus is spreading 60% faster in rural areas. In contrast, people are dying of COVID at a 120% faster rate in Rural Areas than Urban Areas. Even though we don’t have granular hospitalization data, this signals that rural healthcare systems are struggling to cope with their new patients compared to their urban counterparts. These data conclusions are in-line with prior real world news articles.



It’s notoriously difficult to define “Rural America.” I used a metric developed by the Office of Management and Budget and USDA, which defined roughly 46 million Americans as “rural.” To help contextualize this number,  earlier I found that there were an additional 0.3 daily COVID cases per 1000 people in rural areas than urban areas. That means rural areas alone are collectively coping with an additional 100,000 cases a week than if they had the same infection rate as urban areas. And the gap between the two grows every day. 

It’s somewhat interesting to note that the counties defined as “rural” under my metric collectively voted +32 for Trump in the 2020 presidential election (65% for Trump, 33% for Biden). Them’s landslide margins. It’s pretty well known that the Trump campaign has downplayed the severity of the virus, with Trump himself often neglecting to wear a mask — even going so far as to publicly mock Joe Biden for wearing a mask too often. As such, Trump and his movement have become associated with a chronic downplaying of the virus, which includes not wearing masks in public spaces and avoiding social distancing — both of which would help spread COVID. I also wrote a previous article that there appeared to be pretty serious evidence that the GOP as a whole was not taking COVID seriously — it wouldn’t be unreasonable for their voters to follow suit. This may be partially contributing to the disparity we are seeing here — in contrast, the Urban areas voted +10 for Biden, whose campaign had virtually the opposite messaging regarding the virus.

No matter what, the reality is that COVID is overwhelming Rural America more than Urban America. For a myriad of factors (less population, poorer, etc.), Rural America gets a lot less media coverage than other areas of the country. It’s important that despite this, we pay attention to things like health trends in rural areas, because they can quickly (and silently) get out of hand because of poor healthcare infrastructure — see the opioid epidemic. And if you don’t care because you live in a city, you should because overwhelmed rural hospitals are sending their sickest patients to your hospitals, which means they might not have space for you if you get sick. Right now, our best bet is that measures like Joe Biden’s plan to depoliticize mask-wearing will work so that residents in mostly-red Rural America will take steps to combat the spread of COVID. 

P.S. For my friends who live in Virginia, rural Virginia has been hit especially hard compared to Central Virginia and Northern Virginia. It’s tragic. Two of the top 10 municipalities with the highest mortality rates of the pandemic in the U.S. are located in the Commonwealth, and there are (rural) counties where nearly 10% of the population has had COVID so far. Right now, Southwestern Virginia is being hit especially hard by the Third Wave. I’ve thought about doing a separate write-up about Virginia. I have some in-depth code on the inequalities in the state, but I’m not sure what to do with it. Entertaining suggestions. 


* indicates required

Leave a Reply

Your email address will not be published. Required fields are marked *