Maybe “The Big Sort” Never Happened

by John Sides on March 20, 2012 · 14 comments

in Campaigns and elections,Public opinion

Many readers will remember the book The Big Sort by Bill Bishop.  It argues that Americans are increasingly clustered in like-minded political communities.  If one categorizes a county by how its residents voted in presidential elections, as of 2004 nearly half (48%) of Americans lived in “landslide” countries where one presidential candidate got at least 60% of the vote.  In 1976, that number was 27%.

A new article (currently and graciously ungated) by political scientists Samuel Abrams and Morris Fiorina challenges this account, however.  Abrams and Fiorina argue that presidential voting is not a reliable indicator of partisanship, as voting may depend on idiosyncratic features of candidates.  Better, they argue, is party registration, which more reliably measures people’s underlying partisan preference (if any).

When landslide counties are identified using party registration and this same 60/40 threshold, the trend is the complete opposite of a Big Sort.  The fraction living in such counties was 50% in 1976; in 2008, it was 15%.  This same conclusion emerges using thresholds lower than 60/40.

Abrams and Fiorina conclude:

Do the preceding analyses prove that political residential segregation is not occurring? No. That is not our position…The simple fact is that it will take much more detailed research to settle questions about geographic sorting one way or the other. In particular, to examine the subject of residential polarization in a systematic manner requires data at a much lower level than the county level. One of us lives in New York County, New York, where neighborhoods range from the Upper East Side and SoHo to Harlem and Washington Heights. The other lives in San Mateo County, California, where neighborhoods range from the Woodside estates of Silicon Valley billionaires to the Redwood City bungalows of Mexican immigrants. No county-level figures can capture the disparate political textures of these areas, as well as thousands of others in the United States.

There’s another interesting point.  Abrams and Fiorina argue that, even if political segregation in neighborhoods has increased, it may not have negative consequences.  Citing the work of Robert Putnam and others, they argue that a decreasing amount of interaction—political and otherwise—takes place among neighbors.  And so they conclude:

In sum, neighborhoods are not important centers of contemporary American life. Americans today do not know their neighbors very well, do not talk to their neighbors very much, and talk to their neighbors about politics even less. And they do not see themselves as swimming in a sea of like-minded people who have intimidated or cast out anyone who believed otherwise; they are aware that their neighbors differ politically. Even if geographic political sorting were ongoing, its effects would be limited by the preceding facts about contemporary neighborhood life.

I know Bill Bishop from my time in Austin, and he’s also a regular reader of this blog.  I’ve invited him to comment if he sees fit.  Other comments welcome, of course.

{ 14 comments }

James Evans March 20, 2012 at 9:41 am

Meh. Party ID as a superior indicator of partisanship compared to presidential voting? I don’t buy it.

Take the case of white Southern Democrats, who, like most people, have the party ID they were raised with. As we all know, this lineage of white Southern Democrats comes from pre-Civil Rights and since 1964 this group has shifted to become much more Republican in voting preferences. However, for a lot of white Southern Democrats, party ID has remained stable (mostly because of tradition I suppose).

So according to Abrams and Fiorina, a white Southern Democrat who has voted Republican in the last, say, 5 presidential elections is–when it comes to partisanship–more Democratic than Republican. This doesn’t add up to me.

Mo Fiorina March 20, 2012 at 2:01 pm

Mr. Evans, how many white southern Democrats who came of age pre-1964 are still alive? Whites in the south today don’t hesitate to register Republican (Warren Miller documented southern whites moving away from the Democrats in a 1991 article more than 20 years ago). Besides, our sample contains only one deep south state (LA). The trends we document are occurring outside the south.

David Shor January 25, 2013 at 12:03 am

“Mr. Evans, how many white southern Democrats who came of age pre-1964 are still alive?”

Take a quick look at what percentage of voters in Liberty County or Dixie County Florida are registered Democrats and how Obama did.

This isn’t really limited to Florida. There are plenty of precincts in Eastern North Carolina where 80% of voters are Democrats but no Democratic candidate has cracked 30 in a generation.

Stephen Wildstrom January 3, 2013 at 9:26 am

This thinking about white southerners repeats an error common in political analysis. We see it most often in statements like “older voters are getting more conservative.” Actually, the political views of cohorts don’t change much over time, but the cohorts age. In other words, older people aren’t getting conservative, conservatives, particularly those most influenced by the Reagan revolution, are getting old. Only a small percentage of southern voters remember the pre-civil rights Democratic Party. I’m 65 (and a student of Warren Miller’s from a couple of generations ago) and that makes me a member of the youngest cohort that can remember much about life before the 1964 civil Rights Act.

Eric McGhee March 20, 2012 at 1:19 pm

To me, the key here is independent identification. The reality of independent leaners is such a strong finding in political science that I think it blinds the discipline to the things that independent identifiers *can* tell us about American politics. Independent leaners act like partisans, but it’s weak partisans, not strong ones, that they most closely resemble. Thus, if one wants to know something about geographic polarization, it’s important to know how many people in a community register as independents. And that number has been growing tremendously over the same period of time that the geography of presidential voting has polarized. To my mind, this squares well with Fiorina, et al (and many other political scientists) who argue that it’s the elites who have polarized, not the voters.

For more details on independent registration trends, see my paper with Dan Krimm (gated): http://www.palgrave-journals.com/polity/journal/v41/n3/abs/pol20096a.html

Dave March 20, 2012 at 6:09 pm

Campaign 2004 and campaign 2008 were markedly different in voter enthusiasm … with a giant gain in enthusiasm among Democrats in ’08. Gallup shows a 64% enthusiasm edge (‘more enthusiastic’ minus ‘less enthusiastic’) for Democrats that year, while Republicans sat at a minus 4% edge. (Gallup reports a combined Republicans + Republican leaners – with same for Democrats – so this edge includes all but ‘true’ Independents.)

Could this alone have created the different findings in the 2 studies? We might guess that such an enthusiasm difference could have muted what otherwise would have been a larger number of landslide Republican counties.

Equally, has anyone looked below the county level for self-sortation. I ask because Arapahoe County Colorado includes both whole school districts where two-thirds of students are on free or reduced price lunches AND communities that are premiere upper crust golf course neighborhoods. Combining those two elements may show that county NOT to be a landslide county — but I promise you there are landslide precincts going both ways.

Henry March 20, 2012 at 9:01 pm

I don’t like the “Big Sort” hypothesis, because I think 2004-levels of polarisation have been the historical norm and that 1976 and elections around it only seem unpolarised because they reflected a regional realignment. Just look at elections in the late 19th/early 20th century, notably the1894 House elections, where all but six states had 80%+ representation by Republicans or by Democrats.

Jeff Morgan January 3, 2013 at 3:05 am

I think this is a great point. However mid-70′s political flux should affect voting and party ID similarly. Anyway, the first quoted block anticipates our reaction.

“Do the preceding analyses prove that political residential segregation is not occurring? No. That is not our position…The simple fact is that it will take much more detailed research to settle questions about geographic sorting one way or the other. “

Benjy Cook March 21, 2012 at 12:38 pm

What I don’t understand is how a person’s attachment (or lack thereof) to a political party in America’s 2 party system reflect their politics. Compared to Parliamentary Democracies, this link is tenuous as best. If you’re a political person, you vote for issues, and in America, that means voting for specific politicians – without regards to party lines.

Steve Sailer January 2, 2013 at 7:38 pm

There’s a distinction between expressive voting and functional voting. New Yorkers, for example, despite being largely registered Democrats, haven’t voted Democratic in the last 5 mayoral elections ever since a black mayor got a lone term a couple of decades ago. That’s functional voting: keeping city hall white hands is a high priority for a lot of New Yorkers. But, many registered Democratic white New Yorkers who don’t vote Democratic at the mayoral level would be loath to admit they don’t want a black mayor. It’s just, well, you know what happened when Dinkins got in. Never again.

On the other hand, New Yorkers voted overwhelmingly for a black Democrat for President: that’s expressive voting that shows how morally superior they are to all those white racist Republicans who voted for Romney.

I’d say that Presidential voting, being more expressive than functional, is more useful for measuring self-image.

William January 3, 2013 at 12:21 am

Hmm… While I understand the distinctions you are making, I’m not sure it is a fair to compare Dinkins to President Obama. Other than race, I would argue that Obama is a higher quality candidate. Do you think white voters in NYC would not vote Cory Booker in as Mayor?

Aaron Investigates January 3, 2013 at 6:29 am

Fabius Maximus commented on this topic and happened to link this article.

Forgive me, but what exactly are you trying to either question or point out?

County figures? People clustering together with similar political views? Presidential elections having different dynamics? Something completely different which might be posted in one or two sentences??

Thanks.

zbicyclist January 3, 2013 at 10:19 am

I don’t even know my party registration. Illinois has open primaries, so I go in and ask for whichever ballot I want (only 1 to a customer, even in Chicago). I think there are other states that also follow this scheme, making registration an unimportant factor.

DaveL January 4, 2013 at 2:48 pm

Massachusetts allows independents to pick any party’s ballot. If you are a registered member of a party you can only pick that party’s ballot. Thus there is a clear tactical advantage to being an independent.

Another common meaning of independent is “I’m really a (Republican/Democrat/whatever) but I like to display that I’m open-minded.”

Comments on this entry are closed.

Previous post:

Next post: