It's common these days to think that ordinary people are typically either liberals or conservatives. Most political surveys ask people to label themselves as overall liberals or overall conservatives, without other options except "moderate." The two major political parties have platforms containing mostly liberal positions on one side and conservative positions on the other. The vast majority of national elected officials agree with the vast majority of their own party's ideologically aligned positions. Some political researchers have even proposed that fundamental psychological forces cause people to naturally cluster in these two ideological camps.
But then there's a long history in political science of findings that the general public in fact doesn't tend to cluster neatly on a single left-right axis (here's a recent example). In particular, people's opinions on "social" and "economic" issues are often liberal on one set and conservative on the other. Yet there's also evidence that liberal-conservative consistency has been increasing in recent years.
I'm a fan of economist Paul Krugman, and a regular reader of his New York Times blog. But last week he waded into these waters with a genuinely goofy post.
Krugman correctly noted "the fact that not many Americans consider themselves libertarian." Indeed, a 2014 Pew poll found that only 14% of the public say that the "libertarian" label describes them well. But then he said things that are, umm, less factual: "There ought in principle, you might think, be people who are pro-gay-marriage and civil rights in general, but opposed to government retirement and health care programs—that is, libertarians—but there are actually very few." He claimed that while there are tons of robust liberals and robust conservatives, libertarian views represent a political box that is "basically empty." Further, when considering the opposite side from libertarians—people who combine conservative social opinions with liberal economic opinions, a category Krugman calls "Hardhats"—he says this box is "maybe even emptier."
In our book, Rob Kurzban and I took a deep dive into public opinion. In one section, we addressed this exact point, comparing General Social Survey data on public opinion on same-sex marriage and income redistribution, splitting it out in a simple way to show that there are in fact plenty of people who mix and match liberal views on one and conservative views on the other. Indeed, when FiveThirtyEight’s head number-cruncher, Nate Silver, responded to Krugman’s post last week, he used exactly these same variables from the same dataset to make the same point.
Silver's version, though, involved a tremendous bit of over-simplification: He threw out everyone in the sample who had a neutral opinion on either issue. This was not a trivial move—it's almost one-third of the sample. If you want to see our more detailed numbers leaving these folks in, you can check out pages 10 to 12 from chapter 1 of my book, a chapter Princeton has made available online.
I've been working on a new paper on these topics, so I happen to have a boatload of freshly prepared data on this. It's also from the General Social Survey, but this time I've combined people's views across a number of religious/lifestyle opinions (whether abortion should be legal in various circumstances, whether premarital sex is morally wrong, and whether marijuana should be legal) and also combined views across a number of economic opinions (various items relating to income redistribution and government help/money for the poor).
Combining survey items produces a clearer picture of people's overall lifestyle and economic positions. One aspect of this clearer picture is that lots of people really are properly considered "moderate" on these broader political dimensions. The bad news, though, is that you have to make fairly arbitrary decisions about where the cut-off points are between left, middle, and right.
For this post, I considered people's responses on these two dimensions for the full range of available survey years (1975 to 2014). I categorized both dimensions into three equal pieces (liberal, moderate, and conservative), which, as I said, involves arbitrary cut-offs, but ones that I think wouldn't greatly offend many people's intuitions when considering the details of who ends up in these categories. Then I simply calculated the percentages that were lifestyle conservatives and economic liberals (a group Krugman calls "Hardhats," but which I’m giving the more conventional "communitarian" label), lifestyle liberals and economic conservatives (libertarian), lifestyle moderates and economic conservatives (no label), and so on.
So what's the bottom line? How ideologically aligned are the public's lifestyle and economic opinions? First let's look at the recent years, 2004 to 2014. If you're like Krugman, and have been lured by all the talk of liberals and conservatives into thinking people really are just liberal or conservative, the picture will come as something of a shock:
In recent years, it turns out that people who have lifestyle and economic opinions that are either both solidly liberal or both solidly conservative total 25.6% of the public (14.4% liberal and 11.2% conservative). While this is a bigger group than the 20.7% of folks who are either solidly libertarian (11.3%) or solidly communitarian (9.4%), it would be nonsense to characterize the libertarian and communitarian regions as "empty." Further, of course, there are lots of people in the middle on one issue set but not the other, as well as overall moderates who are in the middle on both.
While simple labels are often useful, one should never forget that people are complicated. In fact, keep in mind that we're just looking at two sets of issues here—lifestyle and economic. We could add still others (like immigration, affirmative action, death penalty, and so on), and simple views of ideological consistency would break down even further.
So how is our recent decade different from the good ol' pre-polarization days of the past? Here's the picture for the sample from 1975 to 2002:
Sure enough, in days of yore there were fewer liberals and conservatives (22.5%, rather than the current 25.6%) and more communitarians and libertarians (22.2%, rather than the current 20.7%). Indeed, the number of liberals/conservatives was almost identical to the number of communitarians/libertarians. So, yes, left-right consistency has been increasing recently, just not as much as our everyday talk would suggest.
It's also important to recognize that different demographic groups contain very different levels of left-right consistency. Here's one last picture for the recent years (2004 to 2014), contrasting whites with 4-year college degrees (on the left side of the picture) and non-whites without 4-year degrees (on the right side):
These days, it's clear that among whites with 4-year degrees there are markedly more liberals/conservatives than communitarians/libertarians—it’s 32.9% versus 20.1%. On the other hand, for non-whites who lack 4-year degrees, there are fewer liberals/conservatives (20.2%) than communitarians/libertarians (24%).
Demographic characteristics really matter in figuring out people's issue positions. On the right side, among non-whites without 4-year degrees, for example, there just aren't a lot of economic conservatives (and thus not many "conservatives" or "libertarians"). But this doesn't mean these folks are mostly left-leaning when you consider lifestyle issues. In fact, this group has more "communitarians" than "liberals."
For whites with 4-year degrees, in contrast, "communitarians" are exceptionally rare. This helps explain why Krugman said of this combination: "I don’t even know a good catchphrase for it." In fact, there is no widely used label for those who are conservative on social issues but liberal on economic issues. But it's not because such people don't exist; it's because researchers and political elites don't spend much time thinking about them.
In the book, Kurzban and I drill down on the demographic details. People have complex combinations of lifestyle opinions and economic opinions in large part because the two sets of issues have very different intersections with people's real lives. Views on lifestyle issues relate strongly to things like how much time people spend in church versus hooking up and partying. Economic opinions, in contrast, have a lot more to do with the classic divisions of race, class, and gender.
We explore these kinds of features to show how complex opinions emerge. For example, who's most likely to have libertarian views? White men with higher socioeconomic status who don't go to church much. Who's most likely to have communitarian views? Non-white churchgoers with lower socioeconomic status. And so on, and so on. We've posted an interactive tool that shows how different real-life features (gender, education, religion, etc.) relate (on average) to a wide range of issue opinions.
It would makes things a lot simpler if people naturally clustered into consistent, easily described political profiles. And it’s fine to use ideological labels as a way to try to get your head around the maddeningly complex reality. Supposing that the simple labels are the reality, however, is a fantasy.
Related articles:
Sex, Murder, and the Meaning of Life: Why are THEIR political views so blatantly self-interested?
Time.com: What makes a Democrat a Democrat and a Republican a Republican? It’s more complicated than you think.
This View of Life: We like to think people vote against their self-interest. Research shows it’s not true.
New York Times: Your very predictable vote.
