Social Scientists Know Lots

For me, the year’s best reading is usually the many thoughtful answers to The Edge‘s annual question.  This year they ask:

What have you changed your mind about?  Why?

I’d love to hang out with these folks, so maybe I should audition.  My answer:

Social scientists know lots.

As a physics student and computer science researcher, I assimilated the usual “hard science” perception that “social science” is an oxymoron — no one knows much about it, so your opinion is as good as anyone’s.  When I finally decided I needed social science credentials, to turn my institution hobby into a career, I focused on experimental economics, the only sort a hard scientist could trust, and Caltech, with impeccable hard science credentials.  But I was soon thoroughly convinced: social scientists know tons.

Why then do so many people think otherwise?  Many say it is because social scientists are stupid, or the social world is too complex or uncontrollable.  Better answers are that social expertise conflicts with our overconfidence about familiar experience, or with our democratic ideology that everyone’s political opinions should get equal weight.  But the best answer, I think is that most public talk by social experts reflects little social science.  That is, what social experts say in legal or congressional testimony, or in newspapers or magazines, mostly reflects what they and we want and expect to hear, instead of what expert evidence reveals.

For an analogy, consider the lives of a distant isolated group, like Al Qaeda or devil worshipers.  If most of us had strong emotional preconceptions, or if many advocates wanted us to have certain opinions about them, then we might not be able to believe much of what we heard about this group.  That is, we might not have cheap and effective ways to distinguish honest reports from those driven by other agendas.  This would not imply that no one knows anything about this distant group; insiders there surely know, for example.

Similarly, social scientists have data and theory giving powerful insight into a great many social issues, at least to those with open minds.  Open minded social scientists talking privately can make great intellectual progress.  But powerful forces are eager to distort the messages social scientists give the public on important topics. Academics with deserved reputations for careful accurate work on obscure academic topics tend to adopt different standards when writing editorials or advising politicians.  Even if most academics would not do this, those selected for such roles usually do.

This effect is a good reason for “intellectual travel,” to see many topics for yourself up close.  Also, a mechanism that could cut through this fog and tell the public what honest social scientists really think might have great social value, at least if the public could be shamed into listening to them.  This is one of my great hopes for prediction markets.

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • Hendrik Boom

    Even if most academics would not do this, those selected for such roles usually do.

    I think this is the crucial point.

    When Reagan became president, with his — to say the least — unorthodox economic theories, they were ridiculed by the mainstream economic theorists. But I noticed that over the years, as he proceeded to follow those theories to the detriment of the United States, his theories started to seem to have more support among economic theorists. I suspect that those economic academics that supported government policies ended up getting better financial support. Grant support seems to be essential in the academic world. I doubt that granting agencies are as independent of political influence as they try to be.

  • eric falkenstein

    I think social scientists are at fault. They aren’t saying what the Mainstream Media makes them say, look at any sociology textbook, and see it emphasizes the same crap that makes ‘hard’ scientists roll their eyes: relativism, probabilities imply an ‘its all grey’ viewpoint, Marxist economic determinism and alienation.

    Consider Al Qaeda, where early after 9/11 the most prominent experts said their behavior was based on economics. Violence is a reaction to the callous indifference of Western Society. The US puts sanction on Iraq, starving Iraqis, killing babies–> young men kill themselves to kill Americans. Then we learn most radical Muslims are well educated, relatively prosperous. So are we to give the social scientists credit for getting it right, only after the tendentious journalists fact-checked them and brought the correct views to the fore? Even today, are their ‘true and interesting’ views promoted with any confidence?

  • Stuart Armstrong

    But I was soon thoroughly convinced: social scientists know tons.

    Just out of curiosity, what was it that most convinced you of this?

  • Mason

    Thanks for the post,

    Before people start beating on sociologists, I’d like to point out that economist are just as guilty of ignoring their own work (see Mankwin’s tall tax). As long as you say “should” your open to easy criticism.

    While being right the first time all the time is great, it’s unlikely, and knowing when to change your mind is a valuable skill.

    I used to think that stopping global warming should be our most important environmental concern, but after following Arnold’s postings over that past month or so, I now think we should focus on something with a clearer cause and effect relationship.

  • Stuart Armstrong

    Why then do so many people think otherwise?

    Probably because social scientists are saying things that directly contradict people’s own experience and their own painfuly constructed ‘wordly wisdom’. Tell someone that they cannot exceed the speed of light, and they will not be offended by their previous ignorance. Tell someone they are wrong in the way they bring up their children/deal with economic issues/interact with their friends, and they will be deeply offended.

    From that, it’s just a short step to “social scientists sure must be dumb to have these dumb ideas”.

    Semi-corrolary: social scientists deal with issues where there are many known points of view (spare the rod, spoil the child; all children need is love; firm boundaries are essential; a proper role-model is vital), so it is easy to dismiss a well researched paper as “just another opinion, one of many”.

  • http://dl4.jottit.com/contact Richard Hollerith

    Questions for Robin Hanson:

    But powerful forces are eager to distort the messages social scientists give the public

    It is your belief then that most of the messages are correct — it is just that before they reach the public, they tend to become distorted by intermediaries who are
    not professional social scientists?

    Let us restrict the question to statements in the social scientists area of expertise published in prestigious peer-reviewed journals.

    Most of the messages given by academic psychologists? Sociologists? Cultural anthropologists?

  • Nominull

    Is there any way for me to avail myself of these insights without having to pursue a career in the social sciences?

  • http://hanson.gmu.edu Robin Hanson

    Stuart, all the specific powerful insights.

    Richard, no, I don’t mean to say messages start pure.

  • Matthew

    Robin,

    What am I to make of the fact that the majority of social scientists are hard-leftists in the economics sphere? It does not inspire confidence in your thesis, to say the least. . .

  • Carl Shulman

    Matthew,

    Most non-economist social scientists may know many things about their own fields, but not about economics.

  • Matthew

    Carl,

    Perhaps they know many things, but it makes their prescriptions for positive social change pretty worthless if they do not understand economic thinking.

    I am not a reductionist, but I find any social analysis that ignores economic considerations is pretty much like biology that ignores Darwin.

  • Paul Gowder

    Oh Matthew, do you really think it’s irrational on the face of it to hold left-wing economic views? And so irrational that it proves one up as completely clueless about all social science? Because I think that says more about you than it does about social scientists. Shall I find some hard-leftists in the economic sphere who know more than you do about economics? It won’t be difficult. And I can confidently say this even knowing nothing about what you know about economics, since I can think of at least two Nobel Prize winners off the top of my head.

  • billswift

    Maybe not completely clueless about everything, but if they are obviously deluded about something that you know something about, but have still written about it, you are perfectly justified in avoiding them in the future. There is too much to learn to waste time on things with authors you know are unreliable.

  • Cihan Baran

    As an engineering major, I want to believe in what you say but I feel like I can’t. I am still with the “hard science” perception. Social science is all fun and nice but I think the real deal is hard science.

    Also, I believe that hard sciences require more intelligence than social sciences.

  • Paul Gowder

    Bill: the problem is that I suspect a lot of those judging social scientists have a very low threshold both for “obviously deluded” and for “know something about.” Imagine, if you will, someone who has taken a couple microecon classes contemptuously dismissing, say, John Roemer (one of the easiest 10 or 20 examples of extremely good economists who lean to the severe left of the political spectrum) for being “obviously deluded” or failing to “understand economic thinking. I’m not saying that you, or Matthew, are falling prey to that sort of stupidity… but the edge of the cliff is there… don’t slip… There are a lot of fools who have taken a couple of microecon classes, or majored in it in college, who think that the best economic theory unequivocally dictates a certain kind of right-wing or libertarian politics on distributive issues. But it just ain’t so.

  • Caledonian

    There are a lot of fools who have taken a couple of microecon classes, or majored in it in college, who think that the best economic theory unequivocally dictates a certain kind of right-wing or libertarian politics on distributive issues. But it just ain’t so.

    Could you provide an example of a large (let’s say, greater than ten million people) civilization that functioned according to hard-leftist principles? Or even a modified version of leftist economics?

  • poke

    If someone is skeptical of social science, to convince them of its worth you’d need an example of a theory in social science that is successful on both the terms of the physical sciences and the social sciences, rather than one that’s successful only on its own terms. Are there any?

  • http://drzeuss.blogspot.com Dr. Zeuss

    Robin, could we get a post of your favorite cool things social scientists have figured out?

  • K. Larson

    P. Gowder- you mention that there are a great many “fools” who believe that economics ineluctably implies a libertarian/right wing policy framework. Obviously these fools ignore Stiglitz, Roemer, Frank, and Krugman- all of whom are formidably intelligent. However, there seems to be a LARGE number of economists (many of them equally festooned with Nobels and Clarks) who DO believe that their field of study implies a libertarian approach.

    I can’t reliably critique Kasparov-Karpov chess games, or intelligently weigh the relative merits of string theory and brane theory in physics, likewise, my BA in economics leaves me totally unqualified to adjudicate between Friedman and Stiglitz or Mankiw and Krugman. The proof will be in the pudding, as they say, and the pudding is very deep and very thick in this case. Even if I possessed the academic tools and time to render a totally informed judgment, I know ENOUGH economics to know that the opportunity costs of such an undertaking would be cost-prohibitive.

    So what are we rationalists to do? We could pick the person with the most shiny, smarty-pants medals, but Keynes, Freud, and Lord Kelvin had lots of those, and they seem to have made some mistakes (amongst a great many triumphs). We could pick whoever we think is acting in good faith, though this seems to always wind up miraculously confirming our priors, no matter what they are. Picking one guy to listen to seems rather fraught.

    Rather than focusing on outliers, shouldn’t we select a body of individuals who DO have the relevant tools and time (how about the AEA?) and adopt whatever beliefs have majority support, weighting our confidence in said beliefs in proportion to their level of acceptance? You’re certain to never be triumphantly vindicated, but it seems your average quantity or error will be minimized.

    What do AEA surveys say about distribution issues?

    Apropos of nothing in particular, it would appear that overcoming bias has been blocked by the Honorable, Great, and Always Correct Chinese Communist Party. I can now only access this site through a proxy server. This sort of thing often only lasts a day or so, and is probably only an artifact of China’s idiosyncratic approach to the internet. Still, I thought you might get a kick out of knowing.

  • http://omniorthogonal.blogspot.com mtraven

    Why do social scientists lean left? Maybe it has something to do with the assumption of the right, bluntly articulated by Margaret Thatcher:

    And, you know, there’s no such thing as society. There are individual men and women and there are families.

    It is not surprising that social scientists would not be attracted to an ideology that denies the existence of their object of study, and gravitate to the opposite view.

    Someone above wrote:

    Social science is all fun and nice but I think the real deal is hard science.

    Drunk; streetlight; car keys.

    Also, I believe that hard sciences require more intelligence than social sciences.

    Quite possibly. It also takes more intelligence to do hard science than plumbing, but if my pipes are backed up I’m going to call a plumber, not a physicist.

  • Paul Gowder

    K. Larson: that’s fine. I don’t think that you’d be irrational to pick the majority vote (as it were) and conclude that libertarianish principles were implied (if and only if one accepts certain utilitarian normative claims about “efficiency” as the end of the economic system, and so forth). I’d disagree with you, but I think it’s within the scope of reasonable debate.

    But the view I was objecting to goes further. Matthew and those who agreed with him suggested that it would be irrational to do the opposite — that it would be so patently absurd to endorse left-wing principles and also claim that you know anything about economics. I was objecting to that claim, not to yours. And that claim is pretty easy to refute, by pointing to people who unequivocally know a lot about economics but also endorse left-wing principles.

    I’m perfectly happy to say that both views are plausible, and to cite as support for that plausibility the fact that both views have many really smart adherents who know economics, even though a certain kind of Bayesian (one who is not me) would come out on balance on the libertarianish side.

    Caledonian: Nope. Could you provide an example of a large civilization that functioned according to libertarian principles?

    Poke: why would you need that? Social science is, in a lot of ways, harder than physical science. For one thing, you can’t do a lot of experimentation. You can do some experiments — there are awesome things going on in experimental economics — but a political scientist can’t just start a war in order to see what happens.

  • londenio

    I also studied and worked in the hard sciences (chemistry) and now am an academic in the softer sciences (I do quantitative Marketing). I also changed my mind in the same direction as Robin. I used to be very dismissive of the social sciences, thinking they were populated by people who could not make it into the hard sciences. I was wrong.

    The problems of the social sciences are very difficult and some of the people tackling these problems are outstandingly intelligent. Moreover, the intelligence needed to understand and frame the problems of the social sciences goes beyond the mere mathematical ability that helps a person get a phd in physics, for instance. The intelligence needed for success in the social sciences involves many more dimensions, on top of the mathematical ability.

    I think that some of these misconceptions come from the fact that you need a high intelligence to obtain a phd in physics whereas you do not **necessarily** need a high intelligence to get a phd in a subject of the social sciences. Yet, the people who are at the cutting edge of research, bringing new ideas that help us understand social and economic phenomena are as intelligent as any physicist or chemist at the cutting edge of research, and perhaps more intelligent. And probably more culturally broad.

    Note: I do not have data, or clear examples to back up all this. It is just my opinion. Many of you will disagree. I am ok with that.

  • Caledonian

    Caledonian: Nope. Could you provide an example of a large civilization that functioned according to libertarian principles?

    Not a pure system, no. But I cannot think of a single large, functional economy that wasn’t based on capitalistic exchange.

    Can you offer such an example?

  • Michel

    “social scientists know tons”

    Like what?

  • Paul Gowder

    Caledonian: no, but I don’t have to. The vast majority of left-wing social scientists don’t want to take away the whole market economy, so you’re attacking a strawman.

  • Nadim Shafiq

    Paul G.: Can you provide sources for the statement “The vast majority of left-wing social scientists don’t want to take away the whole market economy”? Thanks.

  • Caledonian

    What happened to those far-left economists you were talking about? Perhaps you say “far-left”, you mean something totally different than everyone else.

  • Doug S.

    It seems to me that MMORPGs could make amazing laboratories for economists. An MMORPG has a mostly self-contained economy that can be manipulated in all sorts of ways that real-world economies cannot be, and they exhibit characteristics of real-world economies, such as inflation. Robin, you’re an economist: if someone tried to do a Ph.D thesis on the macroeconomics of World of Warcraft, how would other academics tend to react?

  • http://hanson.gmu.edu Robin Hanson

    Doug, the reason it is hard to experiment on people is expense: people care what happens to them, and so must be paid to accept random things happening. This is just as true online as offline.

  • Paul Gowder

    Nadim, Caledonian, most of the left-wing social scientists in my experience favor a market economy with substantial redistribution. There are some out-and-out Marxists (including some out-and-out Marxist economists), but most don’t want to abolish private property and so forth. But this is very casual empiricism.

  • Caledonian

    Nadim, Caledonian, most of the left-wing social scientists in my experience favor a market economy with substantial redistribution.

    So they’ve abandoned the idea of founding an economy on their economic ideas and just want to graft their programs onto a supporting capitalistic base?

    What does this movement from communistic principles suggest to you? (Better question: what does it suggest to a reasonable person?)

  • Paul Gowder

    I don’t know Caledonian, you’re a reasonable person, and I’m obviously not, so why keep me in suspense?

  • http://www.iSteve.blogspot.com Steve Sailer

    I’m not a social scientist, but I can play one on the Internet.

    In defense of social science, it’s actually quite easy to make social science predictions that come true. Here’s one: 20 years from now, the public schools in Compton, CA will have lower average test scores than the public schools in Beverly Hills, CA.

    Anybody want to bet against me?

    I’ve got a million social science predictions, all just as depressing and boring as that one.

    So, the social sciences have discovered lots and lots of stuff, it’s just stuff nobody wants to hear.

  • http://www.iSteve.blogspot.com Steve Sailer

    Here’s a social science prediction bet that we won’t have to wait 20 years to see if I’m right.

    Take all the public schools in California.

    Give me the racial makeup of two schools picked at random. Don’t tell me the name of the the schools or where they are located or anything else. I’ll tell you which school has higher average test scores on the state exam. We’ll do it for 100 random pairs of schools and I’ll be right for at least 85 pairs.

    Anybody want to bet?

    If making accurate predictions is the essence of science, then this is science. But people don’t like their social science to be that depressing and boring.

  • http://profile.typekey.com/Unholysmoke/ Ben Jones

    Steve,

    20 years from now, the public schools in Compton, CA will have lower average test scores than the public schools in Beverly Hills, CA.

    Your claim is about as scientific as me saying ‘the sun will rise tomorrow, and that’s physics’ (and not saying anything else).

    Just because you’re saying something about the world that’s probably right, doesn’t make it science. I’m sure you would get 85% right, but without empirical data to support your claim, it’s just pattern recognition and guesswork. I could predict the rising of the sun every day until the end of the world, but unless I think about gravitation and make some measurements, I’m not doing The Science Thing.

    What you say plays into the hands of those who would claim that social science isn’t science at all.

  • charlie

    I skipped the comments section since this should be a no brainer. An effective social science is the end of any individuals creativity, conscience or worth. Only by destroying their (they being: governments, academics, corporations or your family) models do we as people effect change in our living conditions. I’m talking history people, recent history, how about the polling failure in New Hampshire for example. Always lie to pollsters, sleep with inappropriate people, be unexpected and then try to take over the world. I do love my physics and computer science but never ever confuse them with real life.

  • Doug S.

    Doug, the reason it is hard to experiment on people is expense: people care what happens to them, and so must be paid to accept random things happening. This is just as true online as offline.

    Seems to me that some people will pay to have random things happen; consider the popularity of gambling. 😉 An MMORPG is a game, and “random” things are expected to happen at the whim of the game designers. They’re called “events.” Remember Corrupted Blood? That was an accident, but it was still the kind of thing that many players find interesting. In order to perform experiments using an MMORPG as a laboratory, one simply has to secure the cooperation of the game designers, as they would (presumably) veto anything that would make the game significantly worse for players. Furthermore, MMORPGs even offer the ability to run a controlled experiment; popular MMORPGs consist of several independent “servers” that are theoretically identical except for the players and their actions; one can designate one server to be an experimental group and another server to be a control group.

    I suppose there are plenty of obstacles involved, but getting a game designer to agree to change an aspect of an MMORPG should be easier (and less dangerous) than getting the Federal Reserve to do the same thing. 😉 If economics is important (and I think it is) then wouldn’t the ability to do controlled experiments on the scale of actual economies be worth investing a small fraction of the money spent on, say, building better particle accelerators?

  • Douglas Knight

    Doug S,
    check out
    http://terranova.blogs.com/
    If that doesn’t answer questions about how much economists care about such experiments, how hard they are to do, etc, you could ask the people there.

  • elfvillage

    I find it altogether delightful that, after reading to the bottom of the comments, not one person has managed to provide an actual example of something social scientists know.

  • Pingback: Posner on pundits | Javier Aparicio / Public Economics

  • Pingback: Overcoming Bias : From Eternity To Here

  • Pingback: Thou shalt not sit with statisticians nor commit a social science - Statistical Modeling, Causal Inference, and Social Science Statistical Modeling, Causal Inference, and Social Science