Bryan Caplan  

Conspiracy Theory: The Hidden Agenda of the Political Mind

NYPD's Reverse Washington Monu... Be skeptical, be very skeptica...
Does self-interest explain individuals' political views?  Surprisingly, political science's standard answer is No. While self-interest occasionally plays a role, it poorly predicts both issue positions and voting behavior. 

Unlike most economists, I strongly endorse political scientists' consensus.  Their research doesn't just look solid.  I've also personally played with the data for over a thousand hours, confirming that their basic approach is correct.  When I teach this material, I make my graduate students hunt for counter-examples - exceptional cases where self-interest is highly predictive of political views.  Most return from this quest empty-handed, or nearly so.

Jason Weedon and Robert Kurzban's new The Hidden Agenda of the Political Mind: How Self-Interest Shapes Our Opinions and Why We Won't Admit It (Princeton University Press, 2014) frontally attacks the academic consensus against political self-interestSince they charmingly paint me as a leading voice in this consensus, it is in my self-interest for their book to be widely-read.  Unfortunately, Weedon and Kurzban are basically high-brow conspiracy theorists.  They trumpet a strong, incredible thesis, then "interpret" virtually every fact to fit it. 

This is easier than it sounds because they quickly water their thesis down to near-tautology.  After criticizing political scientists for ignoring the power of self-interest, Weedon and Kurzban officially abandon the very word they put on cover of their book!  Page 38:
[I]t's probably best to jettison the term "self-interest" altogether... So, we'll refer to "inclusive interests."  Something is in a person's "inclusive interests" when it advances their or their family members' everyday, typical goals.
If that's not loose enough for you, they stretch the definition again two pages later:
And so our earlier notion of "inclusive interests" needs to be expanded further.  People's everyday, real-life endeavors are enhanced by various kinds of material and nonmaterial gains, over shorter-term and longer-term horizons, received by themselves, their family members, and their friends, allies, and social networks.
Who are your "allies"?  Who's in your "social networks"?  Millions of people.  Weedon and Kurzban appear to count entire races (including "white") and broad religious categories (including "Christian") as allies and social networks.  What about the logic of collective action - the hard fact that helping out amorphous social networks is almost always, selfishly speaking, a waste of time?  Though I'm confident the authors grasp this elementary point, they write as if they don't:
If most African Americans support policies that attenuate the negative effects of racial discrimination, and most African Americans benefit from their policies, and few African Americans are harmed by these policies, then we don't see why supporting one's group would be self-sacrificial.  We view most examples of things that advance "groups" as basically equivalent to things that advance the individual interests of lots of members of those groups.
Structurally, this is no different from denying that blood donation involves self-sacrifice.  After all, don't most blood donors benefit from the availability of blood in case they need it?  Never mind the Prisoner's Dilemma and the Fallacy of Composition.

Given The Hidden Agenda's expansive definition of "interests," it's not surprising that they find a ton of evidence supporting their position.  Single atheists are much more pro-choice than married church-goers?  Of course, single atheists have an interest in easy access to abortion.  Harvard grads are overwhelmingly Democratic?  Of course, because rich workers have an interest in lowering the relative status of rich capitalists.  The book goes on and on like this, a parade of just-so stories of selfishness.

In future posts, I'll investigate some of Weedon and Kurzban's empirical results in detail.  For now, I'll point out the big conceptual flaws.

1. They never clearly state what would count as evidence against their view.  Instead, they dismiss the top competing theories on conceptual grounds.  Most notably: As far as I can tell, they never include a simple measure of left-right ideology as an explanatory variable.  Why not?  Because if their theory is right, ideology is merely an index of interests.
It could be the case, for example, that many people choose to call themselves "liberal" or "conservative" (or "libertarian" or something else or none of the above) based on a kind of summation of their particular policy views.
So what measure of non-interest-based ideology would Weedon and Kurzban accept?  They provide not a clue.

2. The Hidden Agenda never statistically "races" its thesis against any competing view.  When social scientists want to empirically test the view that ideology is merely an index of interests, for example, they normally run a regression of issue positions on (a) detailed measures of interests, and (b) ideology.  Then they see if (b) remains important controlling for (a).  Weedon and Kurzban do nothing like this, ever.  And anyone who knows the data knows why: Ideology usually wins the race by a landslide. 

To be fair, The Hidden Agenda measures interests in some novel ways.  It's conceivable that their novel measures could actually prevail in a statistical race.  But Weedon and Kurzban studiously refuse to allow a contest.  This approach would be understandable, if not excusable, if the scholarly consensus were already on their side.  But as they freely admit, the opposite it true.  When a champion refuses to race, it's suspicious; when a challenger refuses to race, however, it's ludicrous.

3. The Hidden Agenda almost entirely eschews one of their side's standard defensive maneuvers: Retreating from objective interest ("People do whatever is actually in their interest") to subjective interest ("People do whatever they think is in their interest").  To stick with the strong version of their claim, though, they tacitly embrace the absurd view that measuring the effects of public policies is child's play.  FYI: It's not.  To know whether a higher income tax is in your interest, for example, it's s vital to know the elasticities of labor supply and labor demand.  Measuring these elasticities is notoriously difficult.  Ignoring this issue lets the authors blithely interpret education and intelligence as proxies for varying interests, never considering the possibility that education and intelligence might lead to more reasonable beliefs about policies' effects.

4. Suppose a person volunteers at a soup kitchen once a week.  Every year, the volunteers get one free lunch at McDonald's.  A dogmatic believer in the power of self-interest could say, "They volunteer for the Big Macs."  A reasonable person, however, would focus on magnitudes: Selfishly speaking, a free lunch at McDonald's is worth vastly less than 52 days of toil.  In the real world, of course, weighing magnitudes is rarely so easy.  But Weeden and Kurzban seem oblivious to this issue.  If rich pro-choice people vote Democratic, The Hidden Agenda summarily concludes that their abortion interests must outweigh their financial interests.  How could a tiny reduction in the availability of abortion be worth thousands of dollars per year in extra taxes?  Inquiring minds want to know, but Weeden and Kurzban don't tell us.

The Hidden Agenda's final chapter does present two hard cases for their theory: the environment and the military.  The challenge: The views of people with no personal connection to the energy or defense industries still vary widely.
These constituencies [folks tied to the energy and defense industries] are too small, however, to explain the outsized role environmental and military issues play in political debates.  Why, in short, do so many people only distally affected care so much about these issues?  And why have they split out the way they have in terms of competing views?  Fights over global warming, renewable energy, interventions in Middle Eastern conflicts, and related areas affect people's everyday lives, but the connections, it seems to us, are unusually remote, and, further, it would have been hard to predict ex ante which people would have wound up on which side.

When it comes to spending on the environment, the correlations between people's positions and their demographic traits are modest in size, and don't lend themselves to easy interpretation.  People who favor higher environmental spending tend to have more education, not to be regular churchgoers, to be younger and - perhaps most surprisingly - to have no children.
If only Weeden and Kurzban had applied the same skeptical filter to every issue they analyze!  Then they might have been appropriately puzzled by a multitude of other facts they document, starting with:

1. Well-educated, secular whites' high support for redistribution (pp.178-82).

2. Men's above-average support for legal abortion (pp.62-3).

3. The elderly's relatively low support for Social Security (p.137).
The Hidden Agenda of the Political Mind is a book-length application of the motte-and-bailey fallacy.  Scott Alexander provides the background.  In a medieval castle...
...there would be a field of desirable and economically productive land called a bailey, and a big ugly tower in the middle called the motte. If you were a medieval lord, you would do most of your economic activity in the bailey and get rich. If an enemy approached, you would retreat to the motte and rain down arrows on the enemy until they gave up and went away. Then you would go back to the bailey, which is the place you wanted to be all along.

So the motte-and-bailey doctrine is when you make a bold, controversial statement. Then when somebody challenges you, you claim you were just making an obvious, uncontroversial statement, so you are clearly right and they are silly for challenging you. Then when the argument is over you go back to making the bold, controversial statement.

Weeden and Kurzban's put their bailey right on the cover: "How self-interest shapes our opinions and why we won't admit it."  When they confront the data, however, this position is utterly indefensible, so they flee to the motte of "inclusive interests," which they interpret to mean virtually anything.  Once they peruse the data, they return to the bailey, claiming victory for their daring, contrarian position.

The Hidden Agenda isn't all bad.  It has a few nuggets of insight.  It also contains many candidate nuggets - novel claims that might prevail in a statistical race against conventional theories.  And they treat me very well.  Overall, though, this book is an attempt to replace decades of careful and curious social science with near-tautologies and just-so stories.  Contrary to its fans, the only "important" thing about this book is that it might destroy a valuable body of knowledge.

COMMENTS (11 to date)
Thomas writes:

"Well-educated, secular whites' high support for redistribution (pp.178-82)."

This example (among others) suggests that you take a narrow view of self-interest (e.g., that which enriches one). Self-interest means more than that. It's what makes one feel good. Well-educated, secular whites have been indoctrinated to feel good about redistribution, which is why they tend to vote Democrat rather than Republican.

matt h writes:

Isn't the term self-interest problematic in the way that it's being used here. If I define self-interest as anything that gives me positive utils, then almost everything I do is in my self-interest. If I define self interest as in, something that increases my personal wealth, power, or both. Then many things will fail that test.

If X makes me feel good it's in my self interest, but that interest can only be objectively measured by the individual. If X reduces my wealth it or power it may it still maybe a selfish act. Giving money to my children makes me happy, in that sense it's self interested, but the term is meaningless if you define it that way.

Mark V Anderson writes:

Exactly, Matt h. I think that's what Bryan was referring to above as a tautology. If one defines self-interest as doing what makes one feel good (including all charity and voting against one's financial interest), then it is pretty trivial to say people vote their self-interest. It would be worth writing a book to show the opposite to be true.

I assume the consensus that people do not vote their self-interest defines self-interest as an increase in financial or other objective measure of well being. If the authors use another definition, then the book is just silly.

Will Wilkinson writes:

I said some similar things in my short review. Strange, frustrating book.

cassander writes:

I find people are very good at conflating "what is good for me" with "what I think is good for everyone else" particularly when it comes to matters of tribal identity.

Nathan W writes:

Quite a lot of voting models will assume self interest, even though the field of political science (especially the theory/philosophy segments, as compared to realist/liberalist paradigms which treat us as either hawks or cooperators) is quite comprehensively equipped to treat people as far more complex critters than a conditioned dog, for example.

Glenn writes:

This review is unnecessarily critical. Except that the public adoption of political ideology has tremendous signaling value, and most of the "contradictions" you find in observed ideological patterns become easily explained by self-interest.

In the case of rich, well-educated whites, being liberal signals that you are both rich and well-educated, and the cost of doing so is relatively small (however costly liberal policies may be for the rich, the marginal impact of one household's preferences on the overall order is surely tiny).

Glenn writes:

Read "except" as "accept" in the above :)

Jason Weeden writes:

FYI, my reply is here:

Wilkinson: Some of this applies to your piece as well, of course. But the real reply to you will come out on This View of Life, I think very soon -- it's been held up there for weeks as they're preparing their site re-launch.

Two comments:...
1. Thanks for providing the name, "motte and bailey fallacy", of this rhetorical strategy. The most prominent practitioner of the modern age is (was) Steven J. Gould, who would announce a major revision of evolutionary theory and then retreat into qualifications and redefinitions until his major revision became an addition of a punctuation mark to a footnote in someone's commentary on Darwin.
2. "Self interest" is such a useful concept it's mistake to undermine it. Perhaps we can usefully add modifiers like "immediate", "narrow", "visible (material)" to the term, and so admit the purchase of movie tickets and lottery tickets and expressive voting into the category of "rationally self-interested" behavior. Still, some behavior is not self-interested (e.g., catching the flu, falling downstairs).

Andrew writes:

Bryan, thank you, you've altered my prior... I am considerably more convinced of the rational self interest of voters. that's not what I expected really but there it is. Specifically, nourishing a positive self image is one of the most utility enhancing things a human can do, and is probably what the brain spends most of its time doing.

In a wide variety of controlled experiments individuals behave selfishly, generally as expected, so its puzzling that they don't vote for self serving policies. But these experiments share a feature/flaw: the individual believes they can influence the outcome.

As the authors point out: voters can be characterized as voting for policies they think will generate positive outcomes for their group. This is very selfless! But maybe they do this because they think their vote wont make a difference. In presidential votes this is true for most voters. In this case, the vote is not an attempt to affect outcome, but an announcement of values, and a self flattering one at that.

This seems a perfectly self serving act to me

Comments for this entry have been closed
Return to top