How is it that the United States of America is one of the countries hardest hit by COVID-19? We get a strong indication by studying the resurgence of the virus in early June. Looking at the counties with the most significant outbreaks, one factor immediately jumps out: inequality. The inequality that has been a part of this land since before it was a country. This inequality is shaped by race. It stems from specific choices in our history: the enslavement of African people, genocide of Native people, and the continued exploitation of people who immigrate across the US border.
Our social and economic systems are intertwined with this inequality. It is an inequality that forces people into situations where the virus is more likely to spread. A respiratory virus spreads most easily among people together for long periods in close quarters (e.g. jails, meat packing plants, dormitories). The effects tend to be most severe in people who already have health issues. Many of those health issues are also a consequence of inequality (asthma, diabetes, etc.).
This situation is not an innocent accident. There have always been less desirable places to live, and less desirable jobs. There are systems in place that use race to bias who ends up in those less desirable places and jobs (cf. redlining, reservations, work documentation). Those systems have a deep history.
It is possible that we can formulate a strategy for unraveling those systems. The COVID-19 data makes plain where the consequences of inequality have been most severe. Addressing the specific circumstances in those most-impacted places can reveal techniques that will increase equality generally.
County-level COVID data
On June 14, 2020 I began collecting and monitoring the COVID data from the NYT. In the midst of rising alarm about cases going back up, my motivation was personal. I wanted to know how bad things were looking in my own county. There was little reporting on per capita case growth, which could lead to some especially alarmist headlines (i.e. significant increases, but of small values). By combining the county level data with known estimated populations, I found this for the two weeks ending June 14, 2020:
There is a dot for every county in the US. Alameda County looked okay on the scale of things. We averaged 4.1 new cases per day per 100K people during those two weeks. This was still more than twice the target for safely reopening (57.4 cases per 100K per 2 weeks vs. the target of 25), and it was an increase, as the news reports suggested. But Alameda County’s problems seemed minimal, given some of the other points on the plot:
What were these less populated counties, with huge rates of infection? I began to search for information about why these might be the sites of major outbreaks. Extreme outliers are a way to find factors that are present, but less obvious, in all of your data SCP. The cause that emerged was clear. These counties all labor under the ongoing effects of American racial exploitation.
I wasn’t looking for this in the data, but I was doing this analysis right in the midst of the protests over the killing of George Floyd. It seems not a coincidence that the largest protests for social justice ever (?) occurred during this pandemic.
Systemic racial inequality, such a fabric of the history of the United States, persists. It shows up in the COVID-19 data, and clearly has direct consequences for the spread and impact of the disease. This is a strong indication of why we, as a country, are having trouble dealing with this virus. We will need to confront this at some level to make progress.
In the next section, I review the specifics of the most-impacted counties. We need to understand how historical decisions have shaped the details of these counties. It allows us to see where and how remediation can not only limit the spread of the virus, but also make our nation more just.
How the fates of specific counties depend on American history
There are three large historical forces that show up in the extreme counties. The first two predate the United States as a nation, despite the impact still felt today:
- Slavery in the American South
- Genocide of Indigenous Americans
The third is the modern version of having “non-Americans” do the dirty work of extracting value from American land:
- Exploitation of immigrants in agriculture and food processing
Each of these forces is truly connected to the historical ideology that justifies our economic systems of exploitation. Piketty argues that every society must have an ideology to justify the economic inequalities that exist within it. In the United States (and much of the rest of the world) it has become clear that the term “White supremacy” describes a large portion of this ideology.
Categories and counties
We turn now to the circumstances of individual counties. This is really a shallow look, but in this case, that has some damning implications. These factors are visible on the surface if one knows to look. The details I found are the first things of note about these places, as determined by Google searches and Wikipedia. That does not represent “truth”. We should go to these places and do genuine research to connect their histories to their present day circumstances. But it does reveal something: the facts that our schematized information systems think are relevant about these places. A story emerges when we place these amplified facts in historical context.
Category 1: Slavery in the American South
Prisons
The most extreme county on June 14, 2020 was Lee County, Arkansas. It added 525 cases in the first two weeks of June, with a population of only 8,857 people. Here are some of the first facts that come up when searching:
The county is named after Confederate General Robert E. Lee himself. (This first fact I learned about the first county I looked at almost seems enough.) “When Lee County was created in 1873, much of the county consisted of plantations that had been worked by slaves before the Civil War and continued to be worked by tenant farmers after the conclusion of the war” (Encyc of Arkansas). The current population is 57.24% Black or African American. It’s home to the East Arkansas Regional Unit of the Arkansas Department of Correction. This prison holds at least 1,500 people. Arkansas prisons are, it turns out, notorious for their inhumanity. For example, there is evidence that they still use prisoners to pick cotton.
Two other high-impact counties in the South also contain prisons. Their per capita spikes were due to outbreaks in these “correctional facilities.”
Greensville, Virginia, home of Greensville Correctional Center, and the the Virginia execution chamber, had an outbreak of at least 246 cases. It is the largest prison in the state, with 2,918 people incarcerated.
Lake County, Tennessee, home of the Northwest Correctional Complex. It had an outbreak of at least 394 cases. (The number eventually grew to 805 out of 3,971 people tested through July 31).
There is significant research on how mass incarceration of African-American men is the continuation of segregation. Not only does it deny freedom while in prison, it limits options afterwards by marking even those released with the legal status of “convicted felons.”" As Michelle Alexander puts it in The New Jim Crow: “we have not ended racial caste in America; we have merely redesigned it.”
It is obvious that prisons are ripe for spreading COVID. This is inhumane for the prisoners, but also for the people working there. Guards and other prison staff are exposed. This means outbreaks are not just a problem for the inmates, but also for the surrounding communities. And jobs in prisons generally fall into the category of not desirable, again putting at disproportionate risk people with little choice in their employment.
There is much more that could be said on the relationship between prisons, race, inequality and COVID. But the conclusion for now is that prisons are perhaps the worst place to be in the time of COVID. There is a lesson there.
Military
The next county on the list also has an outbreak that cannot be disentangled from the history of slavery.
Chattahoochee County, Georgia — First item from Wikipedia: “The original courthouse, built in 1854 by slaves, is preserved at the tourist attraction of Westville in Columbus, Georgia.” But the outbreak itself is tied to this fact: “Since 1918 most of the land in Chattahoochee County has been part of the Fort Benning military reservation.” At least 142 recruits tested positive during this time period. The Department of Defense has decided to limit release about case counts at military installations, ostensibly to not give away information about our military readiness. It also conceals from the American people whether we are treating our soldiers humanely (perhaps an oxymoron.)
The demographics of the US Army reflect the systemic inequity of other undesirable jobs. About 30% of enlisted Army recruits are Black, vs 12.7% of the US population. The composition of our “all volunteer” army reflects the availability of different occupations to different groups.
Fort Benning is named after Henry L. Benning, a Confederate general and “ardent secessionist, bitterly opposing abolition and the emancipation of slaves.”
These are the kinds of facts that show up again and again when looking at what is going on in these counties. The taking down of statues and renaming military bases make sense when you see the impacts on the communities in named to honor those believed in preserving slavery. While those are not the actions that will directly stop the spread, they do represent action against the ideology that has led to the current state of catastrophe. There is a relationship between the protests for social justice and the COVID outbreaks.
Institutions that bear the marks of slavery persist in the South. These institutions continue to disadvantage and exploit the descendants of people who were enslaved. The consequences can be seen in the COVID data.
Category 2: Genocide of Indigenous Americans
The next category only shows up in per capita data: Native American reservations.
The populations of reservations are relatively small. (A testament of the magnitude of the genocide.) They do not show up at the top of lists when looking for absolute numbers COVID infections. But per capita they are among the hardest hit. (A similar result showed up in The Guardian’s analysis of police shootings. Native Americans were even more disportionately likely to be killed by police than Black Americans.)
There was one county in this category when I made my original list: Buffalo County, South Dakota. First Wikipedia fact: “The Crow Creek Indian Reservation, inhabited by the Crow Creek Sioux Tribe makes up the majority of Buffalo County.” Again, I wasn’t searching for this connection to America’s genocidal past. It emerged from the data, as I was trying to understand why a single county in the middle of South Dakota would be one of the places most impacted by COVID. One level deeper of searching suggests why reservations would be high risk. The website for the Crow Creek Community says “Even though Fort Thompson, Big Bend, and the Crow Creek Community are set in one of the poorest areas in the nation, there are many positive indications.”
Fort Thompson is the largest settlement in the area. It is one of the poorest places in the US, perhaps the world. From a 2007 article: “Often compared to Haiti, Crow Creek is the poorest American Indian reservation in America and suffers from 80-90 percent unemployment, serious medical issues including a high incidence of diabetes, and substandard overcrowded housing.” Wikipedia for Fort Thompson: “The latest census figures show that 21% of houses do not have a kitchen or plumbing. The housing stock is largely overcrowded, with cases of 15 to 20 people living in a modest house.” These are perfect conditions for spreading COVID.
And who was this Thompson that the Fort was named after? In 1861, Clark W. Thompson was given the political appointment of Superintendent of Indian Affairs for the Northern Superintendency. He was in charge of disbursing funds to tribes, as well as their relocation. There are several accounts of him keeping money for himself that was due the Dakota, Ojibwe and Ho-Chunk people, as well as profiting off reselling their belongings. He was also responsible for sending them to this land in Buffalo County, the land that became Crow Creek Reservation. “The land he selected was dry, lacked game for hunting, and was not well suited for people accustomed to a woodland terrain. For six weeks after their arrival at Crow Creek, three or four people died every day from starvation or disease.” (Details from this essay on Clark W. Thompson.)
This systematic taking of wealth and health was U.S. government policy. It has resulted in one of the poorest places on Earth residing within the borders of the United States — a place most of us don’t usually see. 1,962 people live there. 105 of them have gotten COVID. 3 have died. These numbers may seem small, but this is equivalent to 85,000 people getting COVID in Alameda County and 2,400 of them dying. (Instead of the actual numbers 11,000 cases, 181 deaths. Buffalo County has 8x the rate of cases and 13x the rate of deaths.) The COVID data makes the Crow Creek Reservation visible from my computer in California. It should also point us to how to direct reparations and where to contribute immediate relief.
Update since June 14: the effect on reservations is now known to not be an isolated phenomenon. A July 31 Washington Post article recognized that that Native Americans are disproportionately impacted by COVID. This shows up in the per capita data as well. Until mid-June, the most impacted counties in Arizona were Apache, Navajo, and Cococino. According to Wikipedia, the populations of these counties are respectively 76.9%, 47.7%, 28.5% Native American. And they geographically overlap Navajo Nation, Fort Apache Indian Reservation, Havasupai Nation, Hualapai Nation, and Hopi Nation.
The chart above (for all Arizona counties) shows the disproportionate per capita impact on Navajo, Apache, and Coconino counties. It also reveals that while the pace of infection has not slowed much in those counties, Santa Cruz and Yuma counties have overtaken them as the leaders in Arizona. These counties are on the Mexico border, and connect to our third category.
The United States of America started off by claiming the resources of the people already living on this continent. This practice was enshrined into law when the Indian Removal Act was passed by the 21st Congress and signed by President Andrew Jackson in 1830. This act by our government is the direct cause of the systemic poverty experienced by the people living in Buffalo County, South Dakota today. The consequences can be seen in the COVID data.
Category 3: Exploitation of immigrant labor
One of the earliest types of outbreak that rose to awareness were those in meat processing plants. Even the CDC wrote a report on it, which included these details: “Among animal slaughtering and processing workers from the 21 states included in this report whose race/ethnicity were known, approximately 39% were white, 30% were Hispanic, 25% were black, and 6% were Asian.** However, among 9,919 workers with COVID-19 with race/ethnicity reported, approximately 56% were Hispanic, 19% were black, 13% were white, and 12% were Asian, suggesting that Hispanic and Asian workers might be disproportionately affected by COVID-19 in this workplace setting.”
There are many articles on the inhumane working conditions in meat packing plants. What stands out to me are the references to jobs that “most Americans won’t do”. There is a long history of alienating those of us who eat meat from the means of its industrial production. And the most horrific parts are generally done by people who cannot get any other work. After the Civil War, many of the African-Americans who came North ended up working in slaughterhouses. A well-written history of this in Kansas City is here, which contains this observation:
Reflecting the stereotypes of the day, plant managers remarked that African Americans were well suited to “the most disagreeable” elements of meatpacking labor. A spokesman for one large packing interest explained that company policy directed black workers to the most unpleasant jobs on the killing floor, where “the heat is intense and the smell uncongenial to men of more sensitive disposition.”
Many immigrants also worked in the Kansas City stockyards at the turn of the 20th Century. That article references “Germans, Swedes, Russians, Poles, Greeks, Croatians, Slovaks, Lithuanians and Japanese.”
This funneling of some of the worst jobs in America to those who have no other option continues today. Here is a 2017 Bloomberg article about the dangers and horrors faced by the undocumented workers who clean the plants on the graveyard shift after the processing day is done. Reading about the disregard for safety in the industry as a whole, in hindsight it is not surprising that little would have been done to ensure safety from a virus. The article also comments on the racialized nature of who does this work:
In Upton Sinclair’s 1906 novel The Jungle, the protagonist is injured in a meat plant and, like Martha, summarily fired. Sinclair likened the “unspeakable” conditions for European immigrants in Chicago’s meatpacking plants to slavery, only “there was no difference in color between master and slave,” he wrote. A century later, the racial component is back. Almost 30 percent of the nation’s half-million meat and poultry workers are foreign-born noncitizens, compared with 10 percent of manufacturing workers overall, according to the U.S. Bureau of Labor Statistics. More than a third of meatpacking workers are Hispanic.
And here is a 2017 New Yorker article about a chicken processing company that “built its business by recruiting immigrant workers from Guatemala, who endure conditions few Americans would put up with.” The fact that these workers are technically not permitted to work in the U.S. gives the meat companies tremendous leverage: “When these workers have fought for higher pay and better conditions, the company has used their immigration status to get rid of vocal workers, avoid paying for injuries, and quash dissent.” Also, they targeted people here under asylum. “Guatemalans can’t go back home. They’re here as political refugees.” The difference between this and true slavery may only be a matter of degree.
With that historical context, we can look at the specific meat packing county that showed up in the early June data.
Buena Vista County, Iowa was the site of an outbreak at a Tyson pork processing plant. The data point to it being detected on May 23, and lasting until June 13. 591 workers tested positive out of 2300 tested. Looking for more information about Buena Vista uncovers this New York Times article from 2017 about immigrant workers, and the changing nature of Storm Lake (the town in Buena Vista County containing the Tyson plant). Obviously written before COVID, it gives a glimpse at the lives of the people who work at these plants. It contains a number of relevant sentences:
“this four-square-mile patch has absorbed successive waves of immigrants and refugees — from Asia, from Mexico and Central America, and from Africa. They fill most of the grueling, low-paid jobs at the pork, egg and turkey plants;”
“the new owner did not want to hire union supporters. Instead, the company began actively recruiting in Mexico and in immigrant communities in Texas and California.”
“workers can be hard to come by. Standing in the same spot for eight hours or more at a time, in near-freezing temperatures, slashing at carcasses that swing by at a fast pace, can numb body and soul. The poultry industry also ranks among the most dangerous in the United States, according to a new report by the National Employment Law Project.”
“Even if pay were raised to $20 or $25 an hour, Mr. Smith said, “I don’t think you could get white guys.”
But what I found most shocking somehow is that this one of counties represented by Steve King, the most openly racist member of congress. He was born in Storm Lake. He’s the man who openly asked why white supremacy was a negative term, as reported in this 2019 New York Times article. This prompted even his Republican colleagues to censure him. And though he lost his primary on June 3rd (coincidentally?) right in the midst of this outbreak, he still represents much of Northwest Iowa until the end of 2020. In his own words: “It is troubling that the levels of illegal immigration continue to rise and that the Rule of Law is not being enforced. I believe we only encourage illegal immigration by discussing amnesty for the 12-20 million illegal immigrants living in the United States today.” [https://steveking.house.gov/issues/immigration]
We could call this ironic, given the importance of undocumented workers to the industries in his district. A more accurate view might be that he fully understands their role , but is particularly uninterested in treating them as human beings. In fact, we see how the tenuous status of people in the United States without documentation actually helps the businesses in his state make money.
When asked about the impact of the outbreaks, he seemed more focused on the farmers and hogs than the workers. Though he did say the employees needed to be given confidence (slightly different from saying they actually need to be safe). “So, the first thing is to get the plants open. Can’t do that without the employees’ confidence. That means testing it means PPE. And any other policy that we can bring. And that’s what will help lift this accumulating load off.” This was in a TV interview, way back on April 29th. He described the accumulating load of animals that can’t be processed: “the nightmare that I witnessed today with hogs, hogs that would stretch from Minneapolis to Dallas-Fort Worth is also a nightmare we’re likely to see with cattle if we don’t get this moved.”
It’s hard not to assume he’s more worried about the farmers and livestock than the workers.
Again we have a member of our congress, a leader of our nation. Who advocates for the systemic harms that lead to racialized economic inequity. These are not bad apples, or random events. This is the intent.
There are systems in place to exploit the poorest people, allowing our industries to keep the most heinous working conditions. These systems are maintained by our government. They inhumanely punish people who have come to America seeking a better life. The consequences can be seen in the COVID data.
Conclusions
The 6 counties most affected by COVID in early June 2020 reveal three institutions that contribute to high infection rates. The commonality of these institutions is that they maintain racialized inequality. This racialized inequality has formed much of the core of the United States economy. There is a throughline from appropriating land from the Native Americans, to cultivating that land with the forced labor of people enslaved and brought from Africa, to our current agricultural practices. Which still depend on people born outside of the United States to do the back-breaking work of picking the fruit and slaughtering the animals. This history of exploitation and unjust accumulation has left many people without the freedom to escape the coronavirus. This has consequences for everyone.
It looks like resolving the COVID pandemic will require addressing historical inequities. This is discouraging because there are many forces that have resisted this for centuries. But it is also an opportunity. The data allows us to focus on where the inequities have had the biggest impacts, and where remediation and reparations would have the most leverage.
What’s next?
Much has happened since I started this analysis. I have since found this contemporaneous article which confirms my categorizations, though with less historical context. Many more counties have had spikes in cases. I compiled a list over the rest of June, it is below, for completeness.
The narrative around the virus has changed a bit. As we enter August it is being increasingly recognized how large an impact it is having on the Latinx community. This also shows up in the more recent data. Counties on the border with Mexico have high case counts. Also those with large agricultural populations, far from the border. Such as Yakima, Washington.
These are the very populations targeted by our current administration (and congressmen like Steve King) as being the problem with America. It is unlikely that we will have an effective response as long as this kind of exploitative ideology has power.
Note (October 14, 2020): The spreading of COVID has evolved in the months since this analysis was done. As case rates reach new highs in Wisconsin, Montana, Utah and the Dakotas it is clear that we also need to look at reasons why people who have the freedom to reduce the spread of the virus choose not to.
Appendices
Other counties with similar stories
There were many other counties in June with significant case rates. Many of them fit the same patterns as the six most extreme detailed above. Some may have their own stories to tell.
Other counties in June 2020 with high per capita cases:
- Lowndes, Alabama — Wikipedia article (plantations, sharecroppers; poll tax; lynchings; Stokely Carmichael; 75% Black), Bloomberg piece from June 2020
- Nobles, Minnesota — pork [lant, earlier outbreak (late April); still one of the highest total per capita case counties
- McKinley, New Mexico — 75% Native American
- Duplin, North Carolina — more hogs than any other county in the United States
- Texas, Oklahoma — earlier outbreak
- San Juan, Utah — 50% Native American
- Trousdale, Tennessee — settled down, but worst county per capita in the country; Wikipedia article on private prison
- Lake, Tennessee — correctional complex
- Dakota, Nebraska — early May outbreak?
- Union County, Oregon — some stuff about opening early
- Anderson, Texas — five prisons; spiked because of a change in TX reporting (including prison cases in county total)
- East Carroll (Parish), Louisiana — Wikipedia article; “An area of cotton plantations in the antebellum era”, “The most unequal place in America”
- McDonald County, Missouri — meatpacking
- Issaqueena, Mississippi — popped up June 25. Tiny county; could be an anomaly. But: in 1860, 92.5% of Issaquena County’s total population were enslaved people, the highest concentration anywhere in the United States (Wikipedia).
A few other references with related conclusions
NYT posts an article Aug 6 analyzing “how did it happen?” It barely mentions the role of inequality. When it does, it attributes to our libertarian traditions — “a tradition of prioritizing individualism over government restrictions. That tradition is one reason the United States suffers from an unequal health care system that has long produced worse medical outcomes…” I need a bit more explanation of how individualism and a lack of government restrictions create unequal health care.
WaPO has an Aug 6 article about the challenges facing immigrants in meat packing plants.
Link between prisons in the South and slavery (History of Angola prison): Washington Post
Mainstream catching on: CNN
Kamala Harris VP acceptance DNC speech August 19, 2020:
And while this virus touches us all, let’s be honest, it is not an equal opportunity offender. Black, Latino and Indigenous people are suffering and dying disproportionately.
This is not a coincidence. It is the effect of structural racism.
Of inequities in education and technology, health care and housing, job security and transportation.
The injustice in reproductive and maternal health care. In the excessive use of force by police. And in our broader criminal justice system.
This virus has no eyes, and yet it knows exactly how we see each other—and how we treat each other.
And let’s be clear—there is no vaccine for racism. We’ve gotta do the work.
September 2, 2020: COVID leading cause of line-of-duty death among police officers, because of those who work in prisons: Washington Post