North Carolina's Polarization Problem

and its evolution through time

American Political Polarization

Considering Space

By all accounts, the United States is dealing with an increasingly vitriolic, polarized political system. Political polarization has grabbed the attention of media outlets across the political spectrum: ran a headline in September 2017 warning that “rabid partisanship could destroy American democracy” while Breitbart reported that a “poll document[ed] polarization of country into two opposing world views.”

Perhaps to no one’s surprise, Pew Research Center found that since 1994, the partisan gap between Democrats and Republicans nationwide has jumped from 15 points to 36 points, a record high since the survey was first conducted.

Investigations such as the surveys undertaken by Pew Research Center provide valuable data that assess the attitude of the electorate, but offer only a bird’s eye view of American politics by privileging reported attitudes and non-spatial statistics over outcomes of elections and smaller-scale spatial data.

Taking a more geographic approach, John Agnew and Luca Muscará (two political geographers) assert that, “not only is American politics increasingly polarized ideologically, but the country itself is increasingly geographically polarized politically, albeit not at the state level of sections or macroregions, but at the scale of counties, suburbs, and municipalities.”

Let's take a look two elections in North Carolina – the 2000 and 2016 presidential elections – to examine the evolution of spatial-political polarization in the state.

2000 Presidential Election

George W. Bush* (R) & Al Gore (D)

2000 Total Traditional Landslide Counties: 51
(40 Republican wins, 11 Democratic wins)

2000 Total Mega-Landslide Counties: 6
(all Republican wins)

A helpful metric for considering polarization is the landslide victory, which is defined by Bill Bishop in his book The Big Sort as an election where one candidate wins by more than 20 points (FiveThirtyEight also considers elections with margins higher than 20 points to be a landslide.)

2000 Total Traditional Landslide Counties: 51
(40 Republican wins, 11 Democratic wins)

2000 Total Mega-Landslide Counties: 6
(all Republican wins)

The problem with such a definition, however, is that it conceals the heterogeneity of landslide counties. While some margins sit just above 20 points, others soar beyond 60 (a point also highlighted by the above FiveThirtyEight piece).

It's useful to add an additional category: the mega-landslide, or an election where the margin of victory exceeds 40 points. In the above map of the 2000 presidential election, there are 51 traditional landslide counties (78% of which were Republican, and the other 22% were Democratic). Of the 51 landslide counties, 6 are mega-landslides. Since 2000, both the numbers of traditional landslide counties and mega-landslide counties have increased dramatically.

2016 Presidential Election

Donald Trump* (R) & Hillary Clinton (D)

2016 Total Traditional Landslide Counties: 63
(52 Republican wins, 11 Democratic wins)

2016 Total Mega-Landslide Counties: 26
(24 Republican wins, 2 Democratic wins)

From 2000 to 2016, the total number of landslide counties jumped from 51 to 63, resulting in an increase of approximately 20%. More notably, the number of mega-landslide counties leapt 433%, from 6 to 26, and heavily favor the Republican candidate (this trend is clearly demonstrated in the graphic below). Nearly 83% of all the landslide counties (traditional and mega-landslides) were won by Donald Trump, and an overwhelming majority of mega-landslides were red as well (92%).

2016 Total Traditional Landslide Counties: 63
(52 Republican wins, 11 Democratic wins)

2000 Total Mega-Landslide Counties: 26
(24 Republican wins, 2 Democratic wins)

Such a trend could be explained by any number of forces: collective movement to the right on the part of the Republican party, a shift of Democratic voters out of traditionally conservative areas to urban cores (which are perhaps not coincidentally Democratic strongholds), successful efforts to get out the Republican vote, or successful efforts to stifle the Democratic vote in the state (check out what the Atlantic has to say on that here).

Also of interest is the distribution of people among those counties, and simply showing the election results betrays another layer of complexity. In the chart below, you'll find the percentage of the total population of each county (based on July 2015 estimates) in the state of North Carolina compared to the proportion of each county type. For example, Democratic Mega-Landslide counties comprised only 2% of North Carolina's counties, but those counties hold 4.35% of the state's total population. It's evident then that despite few Democratic county-level victories, Democratic counties tend to be more vote rich and represent more constituents than their Republican counterparts.

That elections are becoming less competitive should certainly raise flags among party leaders and the public. Spatial-political polarization and the resulting reality that leaves us in increasingly homogenous political communities is a problematic prospect, given the tendency of like-minded groups to grow more reticent to consider alternative points of view with passing time. Check out the third chapter of Bishop's The Big Sort called "The Psychology of the Tribe" for additional explanation + more resources to explore.

Sources & Data


Agnew, John, and Luca Muscarà. 2012. Making Political Geography. John Agnew + Luca Muscarà. 2012. Plymouth: Rowman & Littlefield Publishers, Inc.

Drutman, Lee. "Hyperpartisanship could destroy U.S. democracy." September 5, 2017.

Hook, Janet. Political Divisions in U.S. Are Widening, Long-Lasting, Poll Shows." Wall Street Journal. September 6, 2017.

Pew Research Center. The Partisan Divide on Political Values Grows Even Wider." Pew Research Center. October 5, 2017.

Other Reading

Bishop, Bill. 2009. The Big Sort: Why the Clustering of Like-Minded America is Tearing Us Apart. Mariner Books.

Gimpel, James, and Iris Hui. 2015. "Seeking politically compatible neighbors? The role of neighborhood partisan composition in residential sorting." Political Geography 48: 130-142.

Holbrook, Thomas. 2016. Altered States: Changing Populations, Changing Parties, and the Transformation of the American Political Landscape. Oxford University Press.

Election + Map Data

I pulled the 2016 election data from Dave Leip's Election Atlas, the 2000 election data from American University's School of Public Affairs, and the shapefiles from the U.S. Census Bureau.

Thanks for checking out my work.

Feel free to reach out with questions or comments at
For more on what I'm up to, check out my portfolio.