(CNN) In Philadelphia, the largest television market in Pennsylvania, the five most-watched cable shows among viewers with at least a four-year college degree include CNN's "Anderson Cooper 360," MSNBC's "Hardball" and reruns of the cosmopolitan situation comedy "Modern Family."
In Erie, the smallest television market in Pennsylvania, the five most-watched cable programs among viewers with at least a four-year college degree include A&E's "American Pickers" (which tracks a duo searching for rare Americana in other people's junk); the History Channel's "Forged in Fire" (in which blacksmiths try to recreate famous historical weapons); and the Food Network's "Diners, Drive-Ins & Dives" (which celebrates down-home back roads cuisine.)
Those starkly contrasting viewing patterns, provided by Brent McGoldrick, co-founder of Deep Root Analytics, a Republican media targeting firm, illuminate one of the central questions about our steadily widening political and social divide: Is the fundamental fissure in American life now demographic or geographic?
The answer, a growing body of evidence suggests, is both. And that may point to a future of even greater distance -- and antagonism -- between a Democratic coalition centered in racially diverse, largely secular, and post-industrial metropolitan centers and a Republican coalition grounded in small-town and rural communities that remain mostly white, Christian and rooted in traditional manufacturing, agriculture and resource extraction.
"I fully expect smaller places to become more conservative and larger places to become more Democratic," says McGoldrick, a long-time Republican consultant. "If you believe geographic sorting/ arbitrage is happening, then where people live and how they live is increasingly important in predicting their views on politics."
Since the early 1990s, the two parties' coalitions of support have steadily separated, both demographically and geographically. That process reached a new peak in the bruising 2016 presidential race between Donald Trump and Hillary Clinton. Geographically, Clinton dominated the nation's biggest places, winning 87 of the nation's 100 largest counties, while Trump carried over 2,600 of the nation's other 3,000 counties, most of them smaller. (He won more counties than any candidate in either party since Ronald Reagan in 1984.)
Demographically, the divides were just as formidable, with Clinton posting big margins among younger and minority voters, Trump romping among blue-collar and older whites, and college-educated whites dividing almost exactly in half between them. The parties' positions in the House of Representatives largely follow these tracks, with Democrats relying mostly on diverse and white-collar urbanized districts, while most of the Republican caucus represents predominantly white and heavily blue-collar seats beyond the metro centers.
One reason small-town and rural areas tilt so much more toward the GOP than urban areas is because their demographic composition leans so much more toward the groups that now most favor Republicans: older, blue-collar and evangelical whites. In an exhaustive recent study, the non-partisan Pew Research Center, for instance, found that non-whites comprised over half the population in the largest urban centers, about one-third in suburban communities, and only about one-fifth in small town and rural places. Whites without a college degree represented about three-in-10 urban residents, exactly four-in-ten in suburbs and nearly six-in-10 in rural places.
But the viewing data that McGoldrick cites point toward another key factor widening the American divide: voters with the same demographic characteristics display very different political and cultural attitudes depending on their geographic location. Each of the electorate's three broadest groupings -- whites without a college degree, whites with a four-year college degree or more and non-whites -- bend steadily toward more conservative views as they move from the most- to the least-populated communities.
That striking trajectory emerges from previously unreleased data Pew provided to me from its survey published in late May comparing attitudes in urban, suburban and small-town America. At my request, Pew calculated the results for an array of key questions among college-educated whites, non-college whites, and non-whites in each of the study's three geographic zones: urban, suburban and rural.
That process produced a consistent pattern that underscored how geography and demography intertwine to shape attitudes.
On the one hand, non-college whites almost always expressed more conservative views than did either non-whites or whites with a college degree living in the same kind of geographic area.
When asked, for instance, whether immigrants had a positive impact on their community, in urban areas 62% of college-educated whites and 51% of non-whites, compared to only 36% of non-college whites said yes. In suburban areas, 56% of college-educated whites and 50% of non-whites, compared to just 32% of blue-collar whites, saw a positive impact. In rural areas, about 40% of both college whites and non-whites saw a positive impact, compared to only about one-fourth of non-college whites.
Likewise, in urban, suburban and rural communities alike the share of college-educated whites and non-whites was greater (often much greater) than the proportion of blue-collar whites who agreed that whites still have advantages over African-Americans; agreed that women still face significant obstacles in society; agreed that society can prosper without people making marriage and child-rearing a priority; and agreed that the growing number of newcomers strengthens, rather than weakens, America. Urban and suburban minorities and college-educated whites were also much more likely than their white blue-collar counterparts to say government should do more to solve problems. (Rural blue- and white-collar whites largely converged on the question.) The sole wrinkle in this general pattern is that in urban areas non-whites were slightly less likely than blue-collar whites to express liberal views on abortion and gay marriage -- a reflection of the deep culturally conservative strains in many African-American and Hispanic churches.
But, just as important, Pew's survey also found that the share of each major demographic group expressing liberal views was almost always greater, often much greater, in larger than smaller places.
For instance, the share of college-educated whites that said that society can prosper without people focusing on marriage and raising children grew from 55% in rural areas, to 59% in suburban communities to 74% in urban centers, Pew found. College-educated whites in urban centers were 27 percentage points more likely than their rural counterparts to consider gay marriage a good thing, and 23 points more likely to support legal abortion. (In each case, suburban whites with college degrees fell in between.) The share of college whites who said government should do more to solve problems rose even more precipitously from about two-fifths in rural places, to just over half in suburbia, to nearly three-fourths in urban centers.
Among non-whites, the share supporting more government activism similarly grew from 62% in rural communities, to 65% in suburbs to 78% in urban centers. The proportion of non-whites who said women still face significant obstacles grew from 56% in rural places to 64% in suburbs to 66% in urban centers. The share of non-whites who said immigrants strengthen society jumped from 56% in rural communities to just over seven-in-ten in both suburban and urban centers. More non-whites in suburban and urban places also took liberal positions on gay marriage and abortion than did their rural counterparts, Pew found.
The same general pattern was evident even among blue-collar whites, the most consistently conservative group. The share of working-class whites that consider gay marriage a good thing grew from 43% in rural places to 49% in suburbs to 65% in urban centers; the share that back legal abortion increased along that same continuum from 43 to 50 to 59%. The portion that believed immigrants strengthened America increased from just under three-in-10 in rural places to over four-in-10 in urban centers (with the suburbs falling in between). And while a 53% majority of rural blue-collar whites (as well as 60% of their suburban counterparts) thought government was doing too much, nearly three-fifths of them in urban centers thought it should do more.
These results all testify to the persistent power of place, and not just social and racial characteristics, in shaping political attitudes. In that way, they reinforce the argument that Ruy Teixeira, a longtime liberal electoral analyst, and author John Judis made in their landmark 2002 book, "The Emerging Democratic Majority." In that book, the two argued that Democrats had a better chance of reaching blue-collar whites who chose to live amid the diversity of urban centers than those who located in more racially and religiously homogenous communities outside the metropolitan core.
In an interview, Teixeira cited three reasons that could explain why voters with the same demographic characteristics are trending toward more conservative positions in smaller geographic areas.
"One is you hang around in an area where certain types of ideas are dominant and you tend to absorb those attitudes," he said. Second, he continued, in small places people are less likely to actually face personal interaction with the sources of so many cultural flashpoints. "There is a well known relationship about ... having certain attitudes about immigration or feminists and not encountering many," he notes.
Finally, he said, these impulses are reinforced by the growing economic gap between thriving larger metropolitan areas and smaller places that are struggling to hold population and jobs. "The fact is that a lot of these white non-college voters who are living in dense areas are living in areas that are working, where economic mobility is feasible, and that takes the edge off of their cultural conservatism," Teixeira says.
McGoldrick likewise sees cultural and economic factors reinforcing each other to deepen the urban/rural split. "The more 'politics' continues to become a series of cultural skirmishes and less about policy, the more the two parties find it advantageous to represent different worldviews," he says. At the same time, he notes, not only are urban centers more welcoming of the nation's propulsive demographic and cultural change, but they are also more likely to consider themselves winners in the economy's ongoing transformation. "My view is the forces driving politics from the New Deal to the end of the 20th century were largely about the role of government plays in society," he says. "I believe, whether or not we realize, we have been in the midst of a different debate, which is the role technology plays in society."
The November midterm election seems likely to further extend this crevice between what I have called the Democratic "coalition of transformation" and the Republican "coalition of restoration." All polls suggest Republicans face enormous risk in white-collar suburbs and urban districts crowded with college-educated whites and minority voters resistant to Trump. But the Democrats' prospects appear much more limited beyond those urban centers. "I think President Trump has effectively given voice to people who see every other voice, institution as aligned against them," McGoldrick says.
Teixeira agrees that Democrats are facing a much steeper hill outside of the largest metro areas, both in 2018 and 2020. But he sees a sliver for daylight for Democrats in the tendency of small-town and rural residents, including most culturally conservative blue-collar whites, to support federal entitlement programs such as Medicare and to view the economy as favoring the rich and powerful. In Pew's data, large majorities of blue-collar whites across rural, suburban and urban communities agreed that the economy favors the powerful; across all three areas, in fact, they were nearly as likely to agree with that sentiment as were minorities and college whites.
"The chink in the armor [for Republicans], such as it is, there is a conflict between these [blue-collar and rural] voters' views of the rich and powerful in general and their views of entitlement programs and the way Republicans really do approach policy," Teixeira says. "If [Democrats] can convince more people that it's a really top priority to help you and your community, they would look the other way on some of their cultural conservative views. But until and unless you can make that case in a way that is convincing to a lot of these people, I don't think they are going to change."