Search Episodes
Listen, Share, & Support
Listen to the latest episode
Subscribe via iTunes
Subscribe via RSS
Become a fan
Follow on Twitter

Support Us:

Please consider making a donation to help make this podcast possible. Any contribution, great or small, helps tremendously!

Subscribe to E-Mail Updates

Related Readings
  • Answers for Aristotle: How Science and Philosophy Can Lead Us to A More Meaningful Life
    Answers for Aristotle: How Science and Philosophy Can Lead Us to A More Meaningful Life
    by Massimo Pigliucci
  • Nonsense on Stilts: How to Tell Science from Bunk
    Nonsense on Stilts: How to Tell Science from Bunk
    by Massimo Pigliucci
  • Denying Evolution: Creationism, Scientism, and the Nature of Science
    Denying Evolution: Creationism, Scientism, and the Nature of Science
    by Massimo Pigliucci

RS 187 - Jason Weeden on "Do people vote based on self-interest?"

Release date: June 25th, 2017

Jason Weeden

What determines which policies a person votes for? Is it their personality, their upbringing, blind loyalty to their political party? Or is it self-interest -- people voting for policies that will benefit themselves and the groups they belong to? This episode features psychologist Jason Weeden, arguing that self-interest is a much bigger determinant of voter behavior than most political scientists think it is. Jason and Julia talk about why researchers disagree over this question, and what "self-interest" even means.

Jason's Blog: We the Pleeple

Jason's Book: "The Hidden Agenda of the Political Mind"

Jason's Pick: "Dragons of Eden" by Carl Sagan

Edited by Brent Silk

Music by Miracles of Modern Science


Full Transcripts 

Reader Comments (11)

The claim that ringbearers should support *legal* restrictions on abortion, birth control etc.. out of self-interest just doesn't track.

The interest of ringbearers is to minimize extramarital temptation for people like them and their spouses by ensuring the widest possible cultural gulf between them and freewheelers, e.g., your wife won't be tempted by those people engaging in abhorent behaviors like abortion, promiscuous use of birth control etc..

In other words Misses Flanders won't ever be tempted by disco Stu. She might, however, be tempted by cool bible camp counselor Stu.

What is in ringbearer's self-interest is to loudly and publicly extol monogomy and demonize abortion, birth control etc.. thereby regulating temptation in their own community. The votes are merely a counterproductive byproduct and, indeed, if they ever succeeded in bringing the freewheelers closer to their own values that would actually work against their interests.


Given the fact that even upon reflection its far from clear if voting for these things actually is in the interest of ringbearers it makes it less plausible that self-interest is really playing a causal role in voting patterns. If it is so hard to even figure out what is in someone's self-interest how do we expect voters to do this unconsciously while believing they aren't voting out of self-interest?
June 26, 2017 | Unregistered CommenterPeter Gerdes
As a broader, methodological point, it seems to me that Dr. Weeden fails to make the prima facia case that 'self-interest' as he defines it is a useful theoretical tool.

The notion of self-interest Dr. Weeden uses isn't the one we tend to use in everyday speech or find of moral relevance so the notion of self-interest he appeals to isn't the one that is of intrinsic interest. So imagine instead he had simply talked about whether people's votes tend to correlate with some construct SI he defined. In that case I don't think he made a compelling case that this is a helpful or useful way to think about voting.

In other words, what can I figure out using his notion of self-interest that was previously obscure or hard to see in models of voter behavior?
June 26, 2017 | Unregistered CommenterPeter Gerdes
Long time listener to the podcast, rationalist, statistician, and deeply disturbed by the lack of criticism of the methodological claims in this podcast. Jason has quite overtly hobbled all methodological controls against spurious causation, (calling them "shit"), substituted standardized definitions for his own clearly biased definitions, and then presented his causal model and claimed validity. This is absurd.

Julia, you were right to press Jason about causailty, because he apparently has no deep understanding of causal concerns. Direction of causailty between correlated variables, which is the only answer he gave, is only one of many concerns that should be controlled for. Your question was in its natural form more specifically focused on latent confounding causes, and he completely ignored that line of inquiry. If you measure lifestyle, which is certainly influenced by personality, and voting preference, which is widely considered to be influenced by personality, the only way to attribute causation between personality and lifestyle is to include both. Excluding that variable as "shit" is not nuanced reasoning, inherently limits the kinds of claims that can be cogently made, and in this case appears simply to be structural confirmation bias.

He clearly states his critical assumption that "you can be much more sure about demographics being causal than you can about things like personality", which in light of all psychometric research of the last few decades seems absurd. Psychometric measurements of personality tend to be around 50% heritable (R^2 of model predicting children's measurement given parents' measurements is around 50%), and neurological bases for many measurements are being worked out. We know rather precisely how psychometric traits vary across demographic groups across time because the model is specific and individual. Group characteristics are clearly causally driven by individual characteristics, and emergent social behavior is driven causally by our mental substrates. The only possible defense I can imagine Jason having for his claim about the primacy of demographics is that he is a dualist (some ancestor-deity model would suffice).

Jason excludes questions about values, but sees it fit to include questions like "are you personally being held back in a job because white people are being favored?" It is preposterous to think that that perception is objectively about self-interest and not heavily colored by deeper beliefs or values. Hiding the ball by asking supposedly objective value questions instead of subjective value questions is not a methodology that can in any sense differentiate between self-interest and beliefs.

Later in the podcast, Jason shows what a vague, all-encompassing notion of interest he has, in the "taking out the trash" example. While simultaneously claiming that self-association doesn't matter, he includes responses about the interests of the people the respondents associate with. Such a metric *does not* distinguish between *self-interest* and association, and Jason's claims that he has distinguished between those cannot be taken seriously.

Julia, in your podcast with Dani Rodrik, you pressed him to admit that he doesn't care about predictive validity, which was a critically important point to make. This interview is incomplete without the points that:
- the latent causes seen most significant by the field are completely ignored, so causal attribution between them and the new proposed measures are statistically impossible
- the assumption about primacy of demographics over personality is clearly contrary to consensus and to a scientific worldview
- self-interest is measured in a way that clearly includes the conceptions of value that Jason claims to be avoiding
- self-interest is defined in a way that clearly includes the self-identificaion questions that Jason claims to be avoiding

As far as I can tell, all of his research that he mentions here is a game of cups and balls, hiding causal structures and obscuring definitions.
June 26, 2017 | Unregistered CommenterMichael
I don't know about anyone, but I think of every thoughtfully considered desire of mine as being in my interest, although the personal benefit that I anticipate might be very indirect. As Julia pointed, there is no natural boundary between self-interest and the interests of others. For example, I consider justice for oppressed groups that I don't belong to to be in my interest because I believe that my life would be better in a more just world even if my relative position in that world were diminished. So I think that instead of asking whether people advocate for their self-interest, a better question would be, How broadly (in time, space, social distance, etc.) do different people define their self-interest?
June 26, 2017 | Unregistered CommenterEli Damon
Feminists allied with Islamists, unions marching with illegal immigrants who take their jobs, wealthy liberal Jews in favor of taxing the rich and welcoming antisemitic Muslim refugees, Asians and white academics supporting Affirmative Action that discriminates against them, white trash voting to cut welfare, vegans altering their whole diet for the sake of farm animals, doctors advocating mandatory vaccination, defense contractors backing Ron Paul...
"That may seem downright counterintuitive: Why would workers for companies that profit from war back an anti-pork candidate? It’s a matter of ideology, military analyst Loren Thompson explained to Defense News. “There’s a strong libertarian streak among many in the sector,” he said. “Just because people work in the defense industry doesn’t mean that they always vote their economic interests.”"
June 28, 2017 | Unregistered CommenterMax
I would have liked to hear if and how Weeden's model has been directly tested against competing models. Toward the end of the podcast the conversation got close to this question, but it was never made explicit. I would be shocked if self interest didn't play a role in voting preferences, just as I would be shocked if political upbringing and moral beliefs didn't play a role in voting preferences as well. Weeden states that:

"What we're talking about is on average, in general, is it statistically more likely that people come down on the side of what looks like their self-interest."

If he's talking about statistical significance, this isn't interesting at all - of course it's significant. At other points he says that the effect is "substantial". Doe he means "substantial, relative to alternative explanations"? Or "substantial, relative to a world in which it was equally likely for people to vote for vs. against their self interest"? I don't know this field, my guess is that the alternative causal models for voting preferences (moral beliefs, upbringing...) are also supported by "substantial" observed effects.

So, it would be interesting to hear what kind of possible statistical outcomes Weeden would treat as conflicting with his model, or of giving more support to an opposing model than to his. He seems to be saying that opposing models are wrongly specified (they have "shit" in them) - so what would a correctly specified opposing model look like? And has he pit his own against these? Are there specific scenarios in which these different models make conflicting predictions that would allow us to say one is better than the other?

Also, to follow up on Michael's post above, Weeden mentions at one point using "stepwise regression" to specify his models. Yikes - I hope there was more to it than that! Computers can't come up with good models for estimating causal relationships just by selecting on p-values or adjusted R-squared or AIC or whatever. Its fine to make variable choice automatic if you're only interested predicting outcomes. But if you're making causal claims about your variables, everything you put in your model must be carefully justified on its own terms.
June 29, 2017 | Unregistered CommenterBen Prytherch
Thanks for the information you brought to us. They are very interesting and new. Look forward to reading more useful and new articles from you!
September 22, 2017 | Unregistered Commenterwings io
Thank you for sharing the podcast. The talk is very helpful for me.
October 6, 2017 | Unregistered CommenterJulia
Hypothesizing about why ring bearers' self-interest includes legal restrictions on reproductive freedom is not difficult.
Women who support those freedoms believe that that their economic fulfillment depends on them. Ring bearers, including women ring bearers, oppose them for exactly the same reason. Economic liberty benefits most those who can take advantage of it.

In ring bearer households, one wage must support the entire family. Freewheeling households are more likely to have two wages (or single freewheeling women support only themselves with their one wage). Freewheeling households are also likely to put off childbirth longer, allowing them to save more before the cut in income. A freewheeling parent who has spent substantial time in the workforce before leaving it can generally return to it at a higher level than one who leaves shortly after joining it.

When the labor force expands because women who would otherwise be caring for children have entered it, the price that labor commands will fall. A household with two earners will command more resources than a household with one. In other words, the ring bearer wage earner has less income and faces higher prices than they would without freewheeling women in the workforce.. For households that cannot command the resources as one with twice the earning capacity, restricting women's ability to enter the workforce makes a lot of sense.

How do ring bearers compare socioeconomically to free wheelers? My cognitive biases tell me the former tend to be at a lower level than the latter. If so, then ring bearers tend to marry/mate with/reproduce with other ring bearers and free wheelers with other free wheelers. In such a world, the entry of women into the workforce benefits free wheelers and harms ring bearers.
November 14, 2017 | Unregistered CommenterJake
I always vote in my own self interest. To that end, I never vote for myself, thus sparing myself the hassle and frustration of public service.

Abortion does seem to relate mostly to the question about when does personhood begin. Religious people have a much higher tendency to view a fertilized egg (zygote) as a person than a person well educated in science. Also, someone who opposes abortion might naturally allow abortion in cases of problematic pregnancy such as rape or an abnormal embryo. Whilst that person would see abortion as unnecessary for mere accidental pregnancy to either a married or single person, that person could have enough residual doubt about their anti abortion position to allow for abortion in cases where someone could expect a high likelihood of a negative outcome in continuing the pregnancy.

People do vote with their pocket books. People dependent on welfare programs vote for more welfare. People with higher incomes vote for tax cuts. Someone on welfare might also vote for tax cuts to stimulate the economy, and this could benefit them since it would help them find a job and eliminate their dependence on welfare. Someone with a tremendous amount of money might vote for more welfare since they have so much money they have fare more money than they need to survive, procreate, and provide for their family.
November 30, 2017 | Unregistered CommenterJameson
Moron, thief, whore.
April 10, 2018 | Unregistered CommenterJane

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
All HTML will be escaped. Hyperlinks will be created for URLs automatically.