There are many ways to measure rural areas. Traditionally, researchers used nightlights as the universal measure of urbanity. This method has some problems. For example, nightlights technically shows economic activity through night time electrification. This is normally not a problem because cities are where the economic activity is. Sometimes this is not always a 1 for 1. For example, due to oil and gas activity, parts of the Dakotas lights up. The average America would disagree if you called western North Dakota urban.
To account for this, researchers have developed increasingly more complex spatial models of urban area. In 2018, the Malaria Atlas Project released a surface that shows the travel time to the nearest city for each 5 km by 5 km place on earth as of 2015. (It was a refresh of an older dataset created by the EU.)
I am interested in how the various urbanity measures can be used to predict outcomes. Last year, I co-wrote a report looking at how they can be used to predict health and employment indicators in three countries in East Africa. In my personal time, I have been doing research into the geographic distribution of donations to Biden and Trump in the 2020 election. I am interested marrying the two and looking at how the urbanity measures that I am used to in my work can be used to predict which candidate received the most donations in each county and eventually who won the county. So, I have been doing a series of extractions at the county level.
I made the map, below, from those extractions. Because it was midnight, I composed a slightly trolly tweet about Iowa not being very rural and set it to auto-post at 9 AM.
People felt, for a variety of reasons, that this map is unfair to the west. That is true. Clark County, NV, King County, WA, Maricopa County, AZ, and Los Angeles County, CA all have much higher average travel times values because they contain a city and a massive area. To make up for this, I symbolized the raw surface data with the same exact breaks as the county map.
The pixel map is more striking than the county map because you can see the fade into very rural areas in the west and the pockets of inaccessible area in the east. My original point still stands. Midwestern states like Iowa, Ohio, and Indiana all have quick access to cities when compared to Appalachia. Even though they grow corn and soy beans, I still consider the coal country of West Virginia, western Kentucky, and central Pennsylvania to be more rural because they are much more remote.