Conservatives Produced By “Low Effort” Thinking: Study

If there is one subject which fascinates academics more than any other, it is why are they so smart and the rest of the populace not. Although there is no research to back this up—which there must be, because according to academic custom research is the only path to knowledge (research says this must be so)—academics do not so much question their own superiority (which is a given) but they wonder why others do.

Our latest peer-reviewed story comes from Scott Eidelman and others who penned the article, “Low-Effort thought promotes political conservatism” in the journal Personality and Social Psychology Bulletin. Eidelman wondered wondered, well he wondered just what the title says, whether “low-effort thought promotes political conservatism.” The unstated, but surely clear, implication is that high-effort thought, such as the kind of deep pondering that courses through the minds of academics like Eidelman, promotes political leftism.

Our researchers conducted three plus one experiments:

In Study 1, alcohol intoxication was measured among bar patrons; as blood alcohol level increased, so did political conservatism (controlling for sex, education, and political identification). In Study 2, participants under cognitive load reported more conservative attitudes than their no-load counterparts. In Study 3, time pressure increased participants’ endorsement of conservative terms. In Study 4 participants considering political terms in a cursory manner endorsed conservative terms more than those asked to cogitate; an indicator of effortful thought (recognition memory) partially mediated the relationship between processing effort and conservatism. Together these data suggest that political conservatism may be a process consequence of low-effort thought; when effortful, deliberate thought is disengaged, endorsement of conservative ideology increases.

To summarize: drunks are more likely to vote for Mitt Romney, as are the stressed and the unreflective.

Conservatism they define as “Emphasis on personal responsibility, acceptance of hierarchy, and preference for the status quo.” Sharp political observers (of any party) will understand that personal responsibility is one thing with which progressives are not comfortable. Better to let those who know better (and know more) decide matters of real importance. By “acceptance of hierarchy” they mean whatever is customary: adherence to the hierarchy progressives wish to impose would not count towards conservatism. And the same for the status quo.

For their first study, they went to a bar and watched people drink and get drunk. Eighty-five folks “participated without remuneration” (they had to pay for their own drinks!). Now, although none of the authors is from there, the bar chosen was in New England. Why is that? One has the idea that this “research” was conducted during a visit to a conference, sponsored perhaps by some Massachusetts college. This is important when we attempt to understand the prevailing attitudes of the citizenry before alcohol loosened their lips.

Hold that in mind (use high-effort thinking) and consider that the imbibers were asked to answer ten questions on a “strongly agree” to “strongly disagree” scale like “Production and trade should be free of government interference” and “Ultimately, privately property should be abolished.”

Eidelman never tells us the ages of the drinkers, though we do learn they are college educated. In New England. Eidelman did not ask people the ten questions then get them drunk and then ask them the questions again. No, sir. Instead, he asked people to puff into a breathalyzer and then answer the questions: their degree of drunkenness was taken as their blood alcohol content (BAC).

To determine whether BAC was related to political conservatism, we regressed the 10-item conservatism index on participants’ self-identification as liberal/conservative, sex (0 = male; 1 = female), level of education, and BAC. Consistent with predictions, BAC was a significant predictor of political conservatism, beta = .21, t(82) = 2.40, p < .02…

Because alcohol limits cognitive capacity and disrupts controlled responding, while leaving automatic thinking largely intact (e.g., Bartholow et al., 2006), these data are consistent with our claim that low-effort thinking promotes political conservatism.

Well, the p-value has it, of course. The “conservatism index” was a sum of the ten questions. Limitations? The only one Eidelman admits is “the possibility of reverse causality remains—political conservatives may drink more alcohol.” And well they might, too, especially in New England.

The real meat is in Eidelman’s admission, “As BAC increased and capacity for deliberative thought decreased, liberal and conservative participants shifted toward conservatism.” Now our researchers interpreted this to mean that mental incapacity induces conservatism. But it’s also possible, and surely more likely, that his casual, non-methodological sample of drunk bar patrons is useless; that is, that it misleads him.

And even more possible is the theory that alcohol reduces inhibitions. A student crawling to a morning Harvard class might well intone that private property should be abolished; she might even be a member of the LGBT-(add letters to suit your tastes) Alliance. But get her down to the bar and pour some Tennessee sippin’ whiskey down her gullet and then you start to hear what she really believes.

The other three studies are all subject to the same criticism. That when drunk, stressed, etc. people tell you what they truly think. Our researchers do admit that “It is quite likely that well-rehearsed, habitual political positions are unaffected by load, alcohol, or distraction” but they do not recognize what this implies for their results.

However, and in a generous tone, they close with

Low-effort thinking promotes political conservatism. This claim provides a counterweight to early psychological perspectives on political ideology that tended to see conservatism in somewhat pathological terms (Adorno et al., 1950).

Admitting to liking private property thus might not be a pathology. Conservatives rejoice.


  1. Low-effort thinking produces junk science and papers that are easily fisked.

    Why are liberals attracted to “disciplines” (if they may be called such) like “gender studies” and “political science” (the mind boggles at the cognitive dissonance in that particular term)? Because they don’t require much thought, just rote regurgitation of shallow bromides, and an overwhelming desire to be perceived as “cool”.

    A sorry lot, really.

  2. I am glad Eidelmman has answered this important question in my life;
    “Why do I tend to be Conservative?”

    I have always basically been lazy, and since engaging in prolonged High-Effort thought can lead to serious mental exhaustion, I have certainly wanted to avoid this state. Therefore, I guess I have subconsciously chosen the Conservative path, taking the easy way out.

  3. As you said, alcohol reduces inhibitions. Alcohol and drugs don’t make you do anything except lose coordination and control of bodily functions. Meet a mean drunk and you’ve met a truly mean person. There has always been the tendency to blame the substance and not the person.

    Now our researchers interpreted this to mean that mental incapacity induces conservatism.

    But of course! Conservatism doesn’t require any thought. Ask any liberal academic intellectual with a superior intellect.

    as blood alcohol level increased, so did political conservatism (controlling for sex, …

    Odd. My experience is increased alcohol level results in loss of control when it comes to sex for a lot of people. But numbers don’t lie.

  4. Actually what is more fascinating is that even drunks have a greater “Emphasis on personal responsibility…” than ivory tower academics. I’m reminded of the William F. Buckley Jr. quote:”I’d rather entrust the government of the United States to the first 400 people listed in the Boston telephone directory than to the faculty of Harvard University”.

  5. The first hint regarding this particular piece of ‘research’ is that it’s ‘peer reviewed’. If it didn’t conform to the a priori assumption that conservatives are somehow mentally deficient it would have never been published.

    It’s a pity that the old adage “in vino veritas” hasn’t prompted them to examine their own ideas. Maybe they just don’t drink enough.

  6. Well, at least Briggs (sorta) read the research article. It’s a start…

    If he read closer, he’d know that the first (and second and third) studies were conducted in Maine because the first-author lived in Maine, and that the relationship in the first study held for those who leaned liberal and conservative. The “alternative” explanation that people lose their inhibitions when drunk, but also when in a hurry, distracted, or instructed to think lightly isn’t much of an alternative account. It takes some work to get past one’s first cognitive response, whatever might motivate the effort. Whether liberals might be faking it is besides the point.

    Also, Briggs misunderstands what “low-effort thought” means, doesn’t understand the difference between basic hypothesis testing and generalizing results, and his interpretation of the dimensions of conservatism the authors discuss is bizarre (but a nice rhetorical device to distract and ridicule)

  7. CDE seems surprised our host pans this scientific study, but hold on! Isn’t that outcome a natural result of Low-Effort thought? And isn’t our professor a “Low-Effort” kind of guy? For an example of High-Effort cogitation just wait until Luis Dias chimes in. Could I get a peer review here from anyone, please?

  8. Two Comments:


    Blood Alcohol Content is not easy to measure. In the jurisdiction I live in the police can give you a BAC test that involves blowing into a portable, handheld unit. If it reads over .08 they can arrest you and charge you with driving under the influence (DUI). However, you are required to take a second test on a much more expensive and sensitive machine one hour after the first test. The second test is the one presented in court to prove the DUI charge.

    A friend had a handheld device, similar to the police device, and I blew into it after consuming less than half a pint of beer. The device reported that I was over the .08 level. My friend said, “Now rinse your mouth out with a glass of water and blow again.” The second time the reading indicated almost no BAC.

    Unless the professors conducting the study controlled for this problem it is not clear how accurate the BAC readings are. (I haven’t read the study and could not link to it.)


    My daughter’s boyfriend asked her if I was ‘on the left or on the right’ (liberal or conservative)?

    When she asked me I said, “That’s a really stupid question.”

    She laughed and said, “That’s what I told him you would say.”

  9. and for an example of the logical fallacy of affirming the consequent (and some pretty sweet irony), see most of the comments above.

    (sorry, but from the suggestion that low-effort thinking leads to political conservatism it does not follow that conservatives use low-effort thinking)

  10. You have to be careful not to do too much high effort thinking. It rapidly wears out your brain cells leading to early brain failure. I conclude this happens a lot in liberals when they talk about “reverse causality” a term I am unfamiliar with. In reverse causality does the effect preceed the cause?

  11. Then there’s this:

    from which I quote:

    The Pew survey adds to a wave of surveys and studies showing that GOP-sympathizers are better informed, more intellectually consistent, more open-minded, more empathetic and more receptive to criticism than their fellow Americans who support the Democratic Party.

    “Republicans fare substantially better than Democrats on several questions in the survey, as is typically the case in surveys about political knowledge,”

  12. Did the academics test potheads? I’m curious as to their leanings. Or crackheads? Whiffers? Heroin addicts?

    There’s more than one way to fry your brain, ya know.

    “High” effort is right on the mark, though. What isn’t clear in the article is how high the researchers were. Did they blow test themselves?

    A good follow up would be to drug test academics and see how many “high” thinkers are feeding from the public trough at our institutions of “higher” learning.

  13. This is so stupid…BAC tests like they administered are notoriously inaccurate! Now Briggs deserves some criticism for misleading us regarding the study yet CDC’s criticism that Briggs jumps to conclusions re the political leanings of the researchers is off base because academics are by far liberal.

  14. Progressives can’t hold their liquor — they all passed out before the quiz could be administered.

  15. I read this differently.

    Being conservative is simple and straight forward (a drunk could do).
    Being Liberal involves mental gymnastics and a convoluted justification system.

    And study 2 proves my point nicely. When you need a workable answer now, use what is simple and quick, and expected to work. Why on earth would you look for a complex answer where a simple one exists?

    I must be conservative.

  16. You all understand that any lack of reliability of the breathalyzer means that it would be *harder* to find a significant relationship between BAC and conservatism, right? and since the researchers found a significant relationship…

    And can William please tell us why participants who were randomly assigned to not think hard about political terms endorsed the conservative terms more than those who did think hard about the same terms (and why a measure of recognition accuracy of the terms mediated the relationship between processing instructions and conservatism?

  17. Is not the opposite of ‘acceptance of hierarch’ acceptance of ‘anarchy’? So, being conservative means being a law-abiding citizen while, according to these shyentishts, being dunk. The more one drinks the more law abiding one becomes. LOL. Who pays these scientists? The North Koreans? If they’re paid by the north Korean government then I have no problems with that, but if they’re paid from my own tax money, then I want my oney back.

  18. There seems to be a basic misunderstanding about the term, “peer review”. A “peer” is someone who is your equal in any relevant characteristics. Or we could say, is someone just like you. So there should be no surprise when a paper, any paper, is accepted by peer review. It has been read by people who think just like the author.

    Incidentally, “sex(0 = male, 1 = female)” has been pretty much my experience too.

  19. ted,

    Your first statement is false. The variability is in the BAC makes it easier to produce a small p-value, especially in a regression context like this, where you have to work hard not to find one. And did you notice the diminutive size of the coefficient? And how that small beta must necessarily mean that predicting actual “conservatism scores” will be of little skill?

    To the first part of your question: I can. This turns out to be the distribution (we only see the mean) of the convenience sample the authors used. I cannot understand your second part, but it is early and I haven’t started drinking yet.

  20. I don’t understand this fourth study where as ted describes it,

    “participants who were randomly assigned to not think hard about political terms endorsed the conservative terms more than those who did think hard about the same terms”

    I’m not politically ‘active’, but I have an opinion on most political questions and it’s an opinion that I’ve developed over a number of years. If someone asks me my view, I have it ready for them. I don’t need time for hard-thinking.

    Maybe I’m unusual in this regard and the rest of the population need to derive their opinions from first principles.

  21. Briggs’ response to Ted is wrong.

    “The variability is in the BAC makes it easier to produce a small p-value…”

    Of course variability is important. But the concern is about the reliability of the measurement, not variability of responses. An unreliable measure has more error, and to quote an intermediate level stats book: “If there is error in measuring X and/or Y, r(xy) underestimates the “true” correlation coefficient that would be obtained if X and Y could be measured without error. This should not be surprising. If our data are contaminated by large amounts of measurement error, they can hardly be expected to reveal strong systematic relationships” (Myers, Well, & Lorch, 2010, p. 457).

    Of course the beta weight is small. It’s competing with participants’ self-reported identification as liberal/conservative. The authors’ never claim the effect is big, only that it is real.

    Briggs doesn’t address the question(s). Instead, he seems to again misunderstand basic hypothesis testing (done with convenience samples or otherwise), and is curiously confused by the most direct evidence in favor of the authors’ causal claim. Was he also confused (or not drunk enough) when he critiqued the studies in his original posting?

    I hate to question other people’s motives in a public forum, but Briggs doesn’t seem to mind doing so, and so one must wonder if the good adjunct professor is intentionally misleading his audience.

  22. The variability is in the BAC makes it easier to produce a small p-value, especially in a regression context like this, where you have to work hard not to find one. And did you notice the diminutive size of the coefficient? And how that small beta must necessarily mean that predicting actual “conservatism scores” will be of little skill?

    If there is an exact relationship between two variables, say, Y = 0.001 * X, then this model can predict without error, i.e., it’s of great skill, even though the coefficient is tiny. Please note that the scale of the independent variable affects the size of the regression coefficient. Mr. Briggs, how about rethinking the first sentence…

  23. CDE,

    In a hurry. But I have to answer:

    Having a small p-value (especially accompanied by a wee regression coefficient—and never mind a linear regression is the wrong model here) does not mean the effect is “real.”

    Take a look at my Start Here tab and you’ll see I advocate Bayesian predictive inference. Hypothesis testing is strictly a suckers game. After you integrate out the parameter uncertainty of the model (all the betas, the sigma, etc.—still assuming the obviously false model is true) you will find the probability of “conservative thought” to be non-predictive; i.e. that knowing whether “conservative” (whatever) essentially gives no skill.

    And then add to the idea of running a regression on a limited-number scale, with its accompanying probability leakage, you have even less a chance the thing is real.

    And then add in the measurement error of the BAC, etc.

    This is a poor study, no matter how you slice it.

  24. My first reaction to the study was that, being conservative, was I supposed to feel somehow insulted? For me personally, conservatism is indeed low effort thinking, especially when I compare it to the considerable thought I invested when I was young and very liberal. In my life, liberalism was a set of opinions about how I thought the world ought to work, and conservatism was my long, slow, painful realization of how the world actually works. I envy those who got there so much earlier than me, benefiting from upbringing or innate intelligence or more penetrating life experiences, but I pity those who will clearly never get there.

    My second reaction to the study was that it seems like such an odd hypothesis to spend time and effort on – it seems, well, so inconsequential. Why not study whether liberal or conservative policies produces better ‘outcomes’ (however they want to define that)? If they were to do such a study, I might actually spend some time engaging in non-low-effort thinking about their results.

  25. I must be really conservative, since I thought I arrived at my conclusions through hard effort. Almost so conservative as to be a skin-head, and I have a hairline to prove it.

  26. >> you will find the probability of “conservative thought” to be non-predictive; i.e. that knowing whether “conservative” (whatever) essentially gives no skill.

    This sounds like complete nonsense and bluff. Variable selection using posterior probabilities is not as simple as described. Trashing p-values incorrectly is not the way to advocate Bayesin predictive inference. Your anti-academic agenda has clouded your mind.

  27. OK, I get it. People with nothing to or think about do tend to be liberal. Busy people with lots to do and lots of responsibilities to think about tend to be conservative.

    Certainly fits observable reality.

  28. OK, I get it. People with nothing to do or think about tend to be liberal. Busy people with lots to do and lots of responsibilities to think about tend to be conservative.

    Certainly fits observable reality.

Leave a Comment

Your email address will not be published. Required fields are marked *