How much of politics can be lined up along one dimension?

Jason Weeden and Robert Kurzban report:

Despite occasional statements to the contrary, most political scientists have long known — going back at least to Philip Converse’s work in the 1960s, and probably farther to Walter Lippmann’s in the 1910s/1920s — that many Americans do not in fact show substantial ideological consistency across policy views, except among limited groups…The 20% of the adult population who are white voters with bachelor’s degrees show some degree of coherence when it comes to views on same-sex marriage and income redistribution.  But, when it comes to the 40% of the adult public who have one or none of these characteristics — including, for example, African Americans and Latinos without bachelor’s degrees and nonvoting whites without bachelor’s degrees — there is no tendency whatsoever for people who lean in a given direction on one of these issues to lean in the same direction on the other.  For the remaining 40% of the adult public, who have two but not three of these features (e.g., white voters without bachelor’s degrees), ideological coherence is barely measurable.

That is from their new book The Hidden Agenda of the Political Mind: How Self-Interest Shapes Our Opinions and Why We Won’t Admit It, interesting throughout.


Generally speaking, what we think of as ideology in America is mostly something that interests white men with 3-digit IQs.

Not so sure about the white men part, but yes to the 3-digit IQ grouping.

All of what passes for 'warring tribes' takes place in this small group. I have long exposure, through school, work, and inclination, to this group, but, unlike most of y'all, I live most of my real life among regular people. Too many people in the high IQ bubble think it is 'tout le monde'- some though genuine obliviousness, others through instant dismissal of the great mass of humanity. Either way, it's a sometimes embarrassing blind spot.

Steve, you (like a lot of lefties, ironically) claim to represent the interests of those outside the bubble, but (again like a lot of lefties) I wonder how often you even interact with ordinary people.

The people who post all the ideological posts in my Facebook feed are certainly not the 3-digit IQs.

Well it is true, there aren't many liberals out there who break the 90 IQ threshold.

The electorate of the city of San Francisco would like to have a word with you.

Like the employees of Google, Apple, and Facebook? Or do you mean the street kids?

Half the population has an IQ below 100 (i.e., Median IQ = 100).

Not surprising then that many folks have a very full agenda just dealing with their daily lives, much less the fine points of political ideology. Only a minority even bother to consistently vote at all in elections they are eligible to vote in. Politics is an unwanted intrusion into ordinary life.

No left or right at the top. Just a tightly knit family running the show--and it is a show.

How am I incoherent because I support gay marriage rights but oppose income redistribution? The two issues seem orthogonal at worst, and to the extent they align on the same dimension, it's "I don't want the government interfering."

Wouldn't your views be considered coherent according to the logic of this book. It's not clear from the quoted passage.

You favor one form of state power intrusion but disfavor another. There's at least some contradiction.

Well, no. A lot of people think of marriage as just an enforceable voluntary contract. It is the denying of the right of contract that is the intrusion. It is the people who are denying the right to marriage for homosexuals who are advocating for an intrusion of state power.

Without reading the book, however, I would have to assume that TedF is, as Sam points out, wrong about what the authors are calling incoherent. I hope I am not wrong about that assumption- it would speak poor of the authors otherwise.

I don't think that's right. State involvement in marriage is about subsidy and regulation. Extending marriage (the legal thing which the state controls, not the social thing which people control) to other people is an expansion of state power. It imposes costs and requirements on people not involved in the contract. For example, it grants the ability to use "Married, filing jointly" on your taxes.

It was wrong (or at least not libertarian) for the state to get involved and override the will of the people when it came to what they wanted subsidized and what they considered pro-social. I'm one of those crazies who thinks the state shouldn't be involved so much in marriage in the first place. And extending it to a realm where the justification for subsidy no longer exists or is considerably lesser was just kowtowing to a wealthy and powerful group. Also, the fact that you can't discuss this in the open is strong evidence that this was about power and not logic or fairness.

If it was just about right of contract, that wouldn't bother me at all.

I'm not sure how to characterize the argument that giving a special legal privilege to one group is less government intervention than giving that same privilege to everyone... but it's plainly wrong.

I disagree Thomas, I think it's less of a government intervention to give welfare/benefits/whateveryouwannacallit to those who really deserve it than to just give it to everyone.

The picking and choosing is the very heart of the failure of government. I think that a consideration of 1st amendment jurisprudence is useful here: choosing the acceptable topics to debate is destructive of free speech just like banning one side of a debate. In that sense the government that enforces unequally is far more impactful and damaging then the government that enforces over broadly in general.

Thomas, that's profoundly wrong. We didn't come close to extending the same privilege to everybody. We're still picking and choosing. We just picked and chose to give a big pile of cash and privilege to a rich and powerful interest group.

I'm an atheist and I have various concerns about plural marriage, but it's manifestly more deserving of subsidy and the legal regime of marriage than gay marriage is. (1) It's about having children, which is the whole reason other people should be interested in helping pay for it, and (2) it's asymmetric in that a wife exchanges sexual exclusivity and child-bearing for a husband's financial support and needs some community protections to make that exchange a safe one and therefore one that happens. When marriage is not about children and not asymmetric, it ceases to be deserving of our cash and oversight.

Plural marriage is an extreme example (and notably common in an unpopular group), but we don't allow just any cohabitant to pay less tax or share health insurance either, do we? No, we subsidize some and not others on the basis of the particulars of their behavior. This is entirely reasonable even to the purest libertarian. Libertarians don't think everyone should go to prison because, hey, we send murderers there...

What a bunch of sophistic gibberish. Marriage is premised on asymmetry? Have you ever heard of working mothers? And under your model I suppose men don't need to be sexually exclusive? We live in the 21st century man, get a grip.

Everyone except anarchists and Kim Jong Un favors some forms of state power intrusion but disfavors others.

You don't want it interfering but you want it to establish homosexual marriage?

Gay marriage = income redistribution

Married people are allowed all kinds of economic benefits paid for from the commons.

Were it not for that bug/feature gays would not be interested in marriage.

Gays are interested in "marriage" because of a desire to damage the moral fabric of society. The tax thing is just a rationalization libertarians use to rationalize their support, few actual gays care.

My, reading comprehension is so challenging.

The authors divide the population into 8 groups. In one of these groups, knowing how a person feels about gay rights allows you to predict with some accuracy how the person feels about income redistribution.

In the other 7 groups, knowing how a person feels about gay rights tells you nothing about how that person feels about income redistribution.

At no point in the above extract do the authors suggest that a person is incoherent for whatever combination of beliefs they have.

Ideological consistency is of use only to those introspective enough to torture themselves about it or in public debates enough to get called on it regularly. Also, there is no logical reason why gay marriage and income redistribution should be the defining issues of a right vs. left split. I get that it's social conservatism versus idealized progressivism, but the two have nothing to do with each other, and libertarians, the largest ideology not explicitly allied to one party or the other have a worldview that squares that circle just fine.

Well said. These topics always remind me of the Worlds Smallest Political Quiz and similar ideas tacking on an additional 'authoritarian-libertarian' dimension to the traditional 'left-right' one, but I'm always left wondering just how many other dimensions could be added to the picture:

centralized vs decentralized
concentrated vs distributed
collectivist vs individual
populist vs constitutional

I'd love to see a giant PCA depicting these other dimensions.

I've found that the more closely people follow politics, the more the more "consistent" they get in their ideology. I'd love to see the survey above with cross-tabs on how political they consider themselves. I guess that sort of goes to Steve's point above.

Yes, there are reasons that people who find politics interesting tend to take more ideologically consistent positions than those who only pay attention the week before an election. They may not be good reasons, but they are reasons that can't be wholly hand-waved away.

Is it all possible that the parties are actually re-defining themselves to capture more donations and votes of the politically active, than the other-way around? That possibility has not been considered in the comments, as far as I can tell. If you were trying to structure a new party, would you try to capture a large share of the voting, white, college-educated crowd, who votes and donates a lot?

'that many Americans do not in fact show substantial ideological consistency across policy views'

Atheists in office? 'Atheism Top Negative for Potential Candidates' - 53% of Americans would be less likely to vote for an atheist -

Support for the military? - 'Military Service Top Positive' - 43% of Americans being more likely to vote for someone with military service - same poll as above.

The amusing thing, of course, is that ever fewer members of Congress serve in the military - 'The number of veterans in the 113th Congress reflects the trend of steady decline in recent decades in the number of Members who have served in the military. For example, 64% of the members of the 97th Congress (1981-1982) were veterans; and in the 92nd
Congress (1971-1972), 73% of the Members were veterans.'\C%3F%0A

That reflects the applicant pool. No one born after 1953 has ever been subject to military conscription. That aside, there has been considerable temporal variation in the devotion of resources to the military and the mobilization of the populace. Expenditures peaked at 14.5% of domestic product during the Korean war, then declined almost monotonically to about 6% in 1978, then increased to 7.6% through 1986, then declined over the next 14 years to 3.7% then increased over the succeeding seven years to about 5.7%, then declined again. There are cohorts born in the 1920s wherein 80% of the men served in the military. If you're looking at men born during the years running from 1930 to 1938, about 65% had some sort of military service. If you're looking at the cohorts born from about 1939 through 1951, about 45% had some sort of service. If you're looking at the post 1953 cohorts, it falls to under 15%. Also, a higher proportion of the Congress consists of women, which cuts into the share as well.

Post-war expenditures peaked at 14.5%. During the period running from 1940 to 1946, about a third of domestic product was devoted to military uses.

So does this counter the What the Matter with Kansas theory? Where do Scott Alexander's tribalism model, or Arnold Kling's 3 Languages model fit into this story? Do these models only apply to "elites", while the proles are self interested? Is ideology itself self serving for powerful people?

Ted has a good point about the orthogonality of some of the issues. You could tell a coherent moral story for any permutation of income redistribution and gay marriage.

"What the Matter with Kansas theory" is countered by a glance at some descriptive statistics of American voting habits. Thomas Frank's account of what happened to the "white working class" is simply wrong.

Perhaps missing in the analysis: just because there's no correlation between "support for gay marriage" and "support for income redistribution" or between "opposition to gay marriage" and "opposition to income redistribution" within a particular subgroup that doesn't mean there aren't still many members of that subgroup who are "ideologically consistent".

Imagine a subgroup where half support and half oppose same-sex marriage. Now let's suppose that each member of the group flipped a fair count to determine his or her position on income redistribution. Assuming the group is large enough, we would see no correlation between "support/support" or "oppose/oppose". That is to say, a member of this group who supports same-sex marriage is no more likely to support income redistribution than one who opposes same-sex marriage. A member who opposes same-sex marriage is no more likely to oppose income redistribution than one who supports same-sex marriage.

On the other hand, fully 1/2 of this group is ideologically consistent and falls into either the "support/support" or "oppose/oppose" camp. So it's not accurate to say that college-educated whites are the only folks who are ideologues. They are no doubt overrepresented among the set of ideologues, but they do not make up the sum total of that set. Or, necessarily, even a majority.

Exactly what part of consistency is related to voting for a political party?

This is an incredibly unenlightening passage, since it doesn't even explain what "coherence" would mean between one's views on gay marriage and one's opinion on income redistribution.

If I assume that what it means by coherence is choosing both left wing or both right wing positions, then it sounds like the white, educated voters are the idiots, and everyone else is thinking for themselves.

So are ballot initiatives an attempt to un-bundle. If I know that most black Americans will always vote Democratic but I am concerned about same sex marriage I start a petition to get a same sex marriage ban on the ballot un-bundling the Democratic party bundle.

I would think so, yes.

Thomas Jackson reviewed this is amren:

The book says some interesting, bordering on un-PC stuff.

Good to see psychologists finally catching up with what's been well-documented in political science for decades if not centuries.

Culturally conservative and economically progressive, that's the kind of world normal people want. It isn't inconsistent unless you've already set up a pie in the sky fantasy world with it's own boogeyman where right thinking people are obsessed with one ideological vector (freedom) to the exclusion of practical reality.

Political coherence *emerges* in college. College professors are overwhelmingly liberal. Draw your own conclusions.

"Political coherence *emerges* in college."

Umm, no.

Generally speaking, by your definition everyone who didn't attend college is politically incoherent. Be sure to mention that fact to the person working on your car, or plumbing or delivering your mail. Just to let them know where they stand.

Did you intentionally write a reply that attempts to prove Thomas' point?

Well there are various definitions of emergent, so if he was being snarky then kudos to him.

So ask 10,000 people to pick a position on a scale from strongly agree to strongly disagree on 100 questions covering a wide range of political issues, run the singular value decomposition and see how the singular values come out. The top three or four singular values (and their associated vectors) are probably going to explain most of the variation.

Democrats routinely get 90%+ of the Black vote these days. I find this to be extremely odd given that when you actually survey Black voters and ask them what they think about certain issues, you'll find that many of them are as socially conservative as mainstream Republicans. In fact, Black Democrats are *more* likely to go to Church on a regular basis than White Republicans and are out of step with non-black Democrats on a variety of moral issues.

Voting behavior isn't just based on ideology or even policy preferences. There's a cultural component to it as well.

Not all policy positions are created equal.

Probably more like 70% of voters lean Democratic or Republican in their actual voting patterns (almost all formally affiliated voters and a modest share of unaffiliated voters). Political affiliation is driven for any individual by their most salient few political issues. You build majority coalitions (pre-election in a two party system), by assembling positions on the most salient political issues for nearly a majority of the voting public that don't directly conflict with each other.

A rich gay philanthropist is more likely to affiliate based on gay rights, which intimately impacts his right to exist in society, than his economic self-interest, even though that matters to him too. A lot of politically conservative views aren't enough to make you comfortable in a political caucus that wants to declare the United States to be a Christian nation if you aren't a Christian (otherwise, Republicans would sweep the Muslim vote and get a much larger share of the Jewish and non-religious vote). A party that cheers on killer cops and wants to gut employment discrimination laws and fly confederate flags at their meetings isn't going to win a lot of the black vote.

Republicans have focused on fear of minorities, fear of cultural change, fear that the strength of the American family is declining and economic insecurity to mobilize their base. This is more symbolic than it is substantive.

Of course, the initial pull that puts someone in the coalition takes a life of its own once your in. If you get involved, you develop a respect for the values of your allies on issues of low salience to you, and you also start to vote out of habit.

"Republicans have focused on fear of minorities, fear of cultural change, fear that the strength of the American family is declining and economic insecurity to mobilize their base. This is more symbolic than it is substantive."

It's amusing that this statement can be modified 180 degrees and yet just as many people would consider it correct:

"Democrats have focused on fears of minorities, fear of cultural change, fear that the strength of the American family is declining and economic insecurity to mobilize their base. This is more symbolic than it is substantive."


Wow from half your post I can already guess at the content and seriousness of your blog content (that isn't a good thing). How ironic this reply is on a post about single dimensions.

You must be a Republican...

That's not culture, that's raw identity politics.

Democrats go to blacks and announce "We're going to do stuff for black people!" This is unsurprisingly popular with black people.

"The 20% of the adult population who are white voters with bachelor’s degrees show some degree of coherence when it comes to views on same-sex marriage and income redistribution."

Alternative: the opinions of white voters with bachelor's degrees is taken as the standard of coherence.

Being the most coherent and the standard of coherency are not the same things.

What exactly is a coherent POV regarding same-sex marriage and income redistribution? I am quite certain that most liberals would be in favor of both, but a libertarian could make the case that being in favor of same-sex marriage and opposed to income redistribution are both small-government points of view and hence coherent.

Comments for this post are closed