what makes an urban, suburban or rural American community

The stark contrast between America’s urban and rural populations was demonstrated in the 2016 presidential election in the United States. It also prompted several researchers (including ourselves) to investigate how Americans’ attitudes and political activities are influenced by the type of society they live in.

To investigate a question like this, researchers must first define “community type,” mainly because there are several ways to determine whether an area is urban, suburban, or rural. In this article, we’ll look at three different ways to assess the kind of community in the United States and compare them to a more arbitrary method: simply asking Americans what type of society they live in.

Overall, we discover that external indicators usually match how people describe their location. However, using self-reported data has its own set of benefits, mainly when distinguishing between different community types — such as urban and suburban — isn’t always straightforward.

Foreground

According to a Pew Research Center survey from 2018, a quarter of Americans live in a metropolitan area, while 43% live in a suburban area, and three-in-ten live in a rural area.

We needed to understand if there were any quantitative metrics — such as government classifications, the density of where people live, or distance from a city center — that could help us move beyond self-reported group form when doing background research for this study.

We ended up comparing our survey results to three external measures: two from the US government and one focused on ZIP code characteristics. We’ll look at how each of these indicators compared to our survey respondents’ views in the sections below.

The National Center for Health Studies’ Urban-Rural Classification Scheme is the first government indicator.

The Urban-Rural Classification Scheme for Counties is published by the NCHS (National Center for Health Statistics) of the Centers for Disease Control and Prevention.

The NCHS first divides counties into metropolitan, micropolitan, and non-metropolitan areas, as specified by the Office of Management and Budget. The size of the metro area is used to categorize counties within metropolitan areas further. Small metro counties have less than 250,000 residents, while medium metro counties have between 250,000 and 999,999 residents. Large metro counties with populations of one million or more are categorized as either “central” or “fringe” metro counties under the NCHS classification scheme. Expansive “core” metro counties are those in a large metro area that includes all or a substantial portion of the metro area’s principal city. The other counties in major metro areas are considered large “fringe” metro counties.

The NCHS scheme is divided into the sizeable central metro, large fringe metro, medium metro, small metro, and non-core metro. This metric was explicitly developed to measure health outcomes. The NCHS, for example, states that residents of large fringe metro counties (similar to suburbs) and residents of non-metropolitan (either micropolitan or non-core) counties frequently have the most disparate health outcomes.

Overall, the NCHS community-style metric corresponded to the self-reported community designations we discovered through our survey. Nearly half of people (45%) who live in an NCHS-designated large central metro area also said they live in an urban area in our survey. On the other end of the scale, the NCHS was much more correct, with 78 percent of people living in an NCHS classified non-core county claiming to live in a rural area.

When the six-category NCHS scheme was broken down into metropolitan, suburban, and rural areas, the measure worked particularly well in rural areas. Around two-thirds of people (68%) listed as living in a rural area based on the NCHS measure said they also live in a rural area. Fewer suburban residents (48%) and urban residents (48%) identified themselves as suburban (45 percent )

Government measure #2: The Rural-Urban Continuum County Classification of the United States Department of Agriculture’s Economic Research Service.

We have looked at the Rural-Urban Continuum Codes from the United States Department of Agriculture’s Economic Research Service (ERS) (RUCC). These codes classify non-metropolitan counties by the degree of urbanization and divide metropolitan counties by the metro region’s size (similar to the NCHS classification).

Counties in metro areas are divided into three categories based on the metro region population; they are less than 250,000, between 250,000 and 999,999, or 1 million or more. Non-metropolitan counties are divided into three categories based on the number of people residing in a Census Bureau-defined urban area: less than 2,500, between 2,500 and 19,999, and 20,000 or more. According to ERS, grouping non-metro counties by the urban population-scale allows for a more comprehensive study of non-metro patterns.

 

This government indicator, like the NCHS, usually corresponded to self-reported group form. In our survey, eight out of ten respondents said they live in a non-metro region with a population of fewer than 2,500 said they live in a rural area.

 

We dissolved the six Rural-Urban Continuum Codes into urban, suburban, and rural to compare with self-reported community form. The RUCC codes categorized rural Americans with a high degree of precision, similar to the NCHS measure. Those classified as residing in a rural area said they live in a rural community 68 percent of the time, while those classified as living in a suburban area said they live in a suburban community 42 percent of the time.

When it came to classifying those living in cities correctly, this coding scheme fell short. Just about a third (32%) of those identified as residing in a metropolitan area (counties in a metro region with a population of 1 million or additional) said they lived in one.

The ZIP code is a unit of measurement.

We decided to see if we could better understand community form by looking at smaller geographical areas, in addition to the county-level indicators available from government sources. We decided to look at the ZIP codes where our survey respondents reside to see if the neighborhood’s characteristics matched their self-description.

The ZIP code distance to the center of the largest principal city in the closest metro area (as determined by the distance to city hall) and the ZIP code’s household density were used as inputs.

These were incorporated into a decision tree, a machine learning technique for determining the best combination of variables to predict a specific outcome — in this case, the type of society in which Americans reside.

Decision tree analysis defines different ways to break a dataset into branches based on each variable’s choices. The algorithm begins by looking for a value among the predictor variables that can be used to divide the dataset into two classes that are the most similar in terms of the outcome variable — in this case, community sort. These subgroups are referred to as nodes, and the decision tree algorithm divides each node into increasingly homogeneous groups until a stopping criterion is met.

Our research yielded eight mostly homogeneous nodes. Based on the bulk classification within the node, we determined which areas would be categorized as urban, suburban, or rural. When there was no substantial majority for a single group form in a node, ZIP code-level data was analyzed to decide the best classification. The following is the classification that resulted:

Americans who live in ZIP codes that are 12 miles or less from the nearest city’s center and have a household density of more than 1,314 per square mile are considered urban.

 

Americans who live in ZIP codes that are 12 miles or less from the nearest city’s center and have a household density of 1,314 or less per square mile are classified as suburban. Folks who live in ZIP codes that are more than 12 miles from the city center and have a household bulk of more than 106 households per square mile are also included in this category.

Americans who live in ZIP codes that are more than 12 miles from the nearest city’s center and have a household density of 106 or less per square mile are considered rural.

This urban area classification mainly refers to those who reside in heavily populated areas near city centers. In contrast, the rural area classification primarily applies to those who live in less densely populated regions farther away from cities. Suburban areas include less densely populated areas surrounding towns and densely populated areas that are slightly farther away from the city center.

We discovered that 56 percent of those counted as living in an urban area self-identified their neighborhood as urban, 34 percent as suburban, and 9 percent as rural, using our decision tree model. 58 percent of those who live in a residential area consider their neighborhood to be suburban. In comparison, about a quarter (24 percent) believe it to be urban, and 17 percent believe it to be rural. And, among those who said they lived in a rural area, about two-thirds (66%) said their neighborhood was rural, while 22 percent said it was suburban and 9% said it was urban.

How did each of the three tests fare?

When we compared all three steps to the results of our survey, we noticed a few trends. Rural Americans were categorized the most accurately by all three approaches, while urban and suburban Americans were classified the least accurately. While all of the measures performed admirably on average, the decision tree was the best at matching self-reports in all three population groups.

According to the decision tree model, nearly two-thirds of those listed as living in a rural area also self-reported as residing in a rural community. Similarly, a sizable proportion of those identified as residing in suburban (58%) and urban (56%) areas said they lived in those types of communities.

According to the NCHS county classification, most of those living in rural areas (68%) reported living in the same type of neighborhood, compared to about half of those living in suburban communities (48%) and 45 percent of those living in urban areas.

 

Using the RUCC county classification, nearly two-thirds (68%) of those identified as rural self-reported as such. Smaller percentages of those identified as residing in suburban (42%) and urban (32%) areas reported living in those types of communities.

final thoughts

Although the decision tree measure outperformed the two government measures by a small margin, there was no simple winner for our purposes. In the end, we determined that the most helpful metric was the self-reported group form.

The benefit of the self-reported group style is that it measures how the respondent feels. Take, for example, a neighborhood on the outskirts of a big city that is legally within the city limits. The houses are far off apart, and there is no food or trade within walking distance. Is it called urban because it is legally within city limits? Is it residential, or is it rural, because the houses are so far apart? We discovered that how respondents felt about their neighborhood aided us in assessing their community in a survey.

Comments are closed.