I Can See Your Vote from Here!
The Strange Reverse Politics of the Mississippi Alluvial Plain
Any casual student of the American democratic experiment would be forgiven for believing that our country’s wacky electoral college system is an accurate description of state identity. Every four years, the TV networks depict the electoral college map with clearly defined blocks of bright blue and red. Although this shorthand, first utilized in 1976, provides a useful graphic for understanding election results, it also provides a false narrative about the connection between the popular vote and the electoral college.
This is not to say that the individual states do not have defined histories, symbolism, songs, stories, and pride. But our modern political system is more a rural vs. urban divide rather than red state blue state divide.
A study in the journal Political Behavior from early in 2020 analyzed Gallup poll data from 2000 to 2018, and found that an individual’s probability of identifying as a strong Democrat drops by 12 percentage points if they live in a far rural area but increases by 11 points if they live in a densely packed community. Andrew Reeves, co-author of the paper and associate professor of political science at Washington University in St. Louis, comments, “In rural, less populated areas, residents are more likely to know one another and talk with their neighbors. Those interpersonal relationships are highly influential and can create a social pressure to conform.” In contrast, heavily populated cities are traditionally more open to liberal ideas and more accommodating to unconventional behaviors and beliefs.
This is not news to any educated voter from middle America. Presidential elections are a numbers game, pitting Democratic urban centers against wider rural Republican strongholds. Whichever side gets the most votes claims the “political identity” of the state, at least for another four years.
Illinois serves as a classic example. In the graph below, the county-by-county results from the 2020 presidential election are compared to population density and non-White population according to 2019 US Census population estimates. For the most part, the images line up. Heavily populated urban centers, such as Cook County (Chicago) trended Democratic, whereas sparsely populated counties (much of the rest of the state), trended Republican. Illinois became a blue state only because the urban numbers won out.
Georgia, the attention-seeking child of the 2020 election, showed a similar pattern, with the greater Atlanta area, Augusta, Columbus, Savannah, and Macon trending blue. However, Georgia also showed the exception to the rural-urban political model. Due to the so-called Black Belt in the American South, areas of high non-White population (particularly African American) trended Democratic.
The Black Belt was first named after the rich, dark, cotton-growing soil stretching from Mississippi to the Carolinas that millions of enslaved Africans were forced to work, but in the 20th century it references a wide area of rural counties with a high percentage of African-American residents. The legacy of this region lives on today, geologically, historically, politically, and culturally. According to Selma civil rights attorney J. L. Chestnut in his 1990 autobiography, “for a hundred years, the Black Belt dominated state politics and the big landowners dominated the Black Belt.” It was the center of the Civil Rights movement but it has the most consistently segregated schools. Years of voter suppression, racism, and poverty have affected health outcomes. According to an analysis by the Washington Post, some counties have actually shifted slightly less Democratic. And like the Black Belt in other parts of the South, this part of Georgia is slowly losing population, capping the number of new votes either party can win.
Based on all this, I thought I finally understood the geographical nature of presidential politics. Then my oldest son dropped some new knowledge on me. “Do you realize,” he said, “that you can see political battlegrounds from outer space?” He showed me the blue red map of the 2020 presidential election of the Mississippi Alluvial Plain and compared it to a satellite view of the area. The lighter shaded area matched the blue counties nearly perfectly. “It’s farmland,” he said. “Mississippi is the opposite of the Midwest. Farmland goes blue, not red.”
My own analysis confirmed he was dead right. The Mississippi Alluvial Plain is tan colored on a satellite map due to the deposits of clay, silt, sand, and gravel left behind by the Mississippi river and its tributaries. But a closer examination of the region shows clearly delineated farmland. And this farmland creates a blue corridor in western Mississippi and a slice of eastern Arkansas and Louisiana.
Based on this anomaly, I couldn’t help but wonder, is Mississippi truly the opposite of Illinois? Does a greater percentage of farmland lead to more Democratic votes? And if Mississippi is ever to become the next Georgia, at least in terms of swing state presidential elections, will the same issues that drive Democratic votes in Georgia drive them in Mississippi?
A visual examination of seven states along the Mississippi river (Louisiana, Mississippi, Arkansas, Tennessee, Missouri, Kentucky, and Illinois) shows that in all these states, higher non-White populations are found in urban populations, and these correlate with Democratic wins in the 2020 presidential election. However, in Mississippi, where all counties have population densities less than 500 people per square mile, the urban/rural divide falls apart.
If the urban/rural battleground no longer applies in Mississippi, what exactly drives their presidential politics? Instead of historically or sociologically, perhaps we can tease out an answer mathematically. The Pearson Correlation Coefficient can be used to calculate linear correlations between demographic, economic, and political variables. More specifically, a correlation analysis of demographic and economic factors of all the counties in the Mississippi Alluvial Plain (plus Georgia, which is in the Black Belt but not near the Mississippi River) shows what actually drives populations to vote more Democratic (blue) or Republican (red).
A rule of thumb for interpreting the correlation coefficient is:
- 0.9 to 1.0 (−0.9 to −1.0) = Very high positive (negative) correlation
- 0.7 to 0.9 (−0.7 to −0.9) = High positive (negative) correlation
- 0.5 to 0.7 (−0.5 to −0.7) = Moderate positive (negative) correlation
- 0.0 to 0.5 (−0.3 to −0.5) = Low positive (negative) correlation
- 0.0 to 0.3 (0.0 to −0.3) = Negligible correlation
To be sure, there are caveats to this methodology. First, the analysis looks at each county as a variable, not each voter. Second, as every scientist ever will tell you, CORRELATION DOES NOT EQUAL CAUSATION. Having said that, the table below is pretty illuminating.
Each correlation coefficient (and its associated shade of blue or red) indicate the likelihood that the variable in the Y-axis relates to increased Democratic votes in the 2020 presidential election for that state. I tested four categories of variables:
- Demographic factors: people over age 65; percent White, Non-Hispanic; percent Black/African American; percent Hispanic/Latino; percent foreign born
- Economic Factors: percent with Bachelor’s degrees; percent in poverty; percent who are veterans
- Empowerment Factors: percent of women-owned firms; percent of minority-owned firms; percent Black/African American-owned farm operations (2017)
- Geographic Factors: percent of total acreage that is farmland (2017); relative population density (compared to Cook County, IL)
All states showed that a higher African American population correlated with more Democratic votes, and a higher White population with more Republican votes, particularly in highly segregated Louisiana and Mississippi. Hispanic/Latino populations had a low to moderate Democratic correlation in Tennessee, Kentucky, and Illinois. Old folks and veterans typically skewed Republican.
Higher Bachelor’s degrees correlated with Democratic votes in Arkansas, Tennessee, Missouri, Kentucky, and Illinois, but not in Louisiana, Mississippi, or Georgia. Higher poverty levels trended Democratic in Louisiana, Mississippi, and Arkansas, but trended Republican in Missouri and Kentucky. And in nearly all the states, having more women-owned firms, minority-owned firms, and black-owned farms correlated with more Democratic votes.
But the wildest trend related to Mississippi geography. Increased farmland corresponded to more Democratic votes, but increased population density (urban centers) had no relation to Democratic votes. This was the exact opposite of the other states where high density urban centers skewed blue. As my son pointed out, in Mississippi, farmland goes blue not red.
This means that any future Democratic politician hoping to turn Mississippi into the next Georgia swing state will have to use a completely different playbook. Democratic politicians can’t speak to Mississippi Democrats like they are all from Atlanta. Strangely enough, they might have to speak to them like they are from the middle of Illinois.
