Dan Klein on Truth, Bias, and Disagreement
Mar 16 2009

Dan Klein, of George Mason University, talks with EconTalk host Russ Roberts on truth in economics, bias, and groupthink in academic life. Along the way they discuss the Food and Drug Administration (and the drug approval process), the culture of academic life and the roles of empirical evidence and prediction markets in adjudicating academic disagreement. The conversation closes with a discussion of Econ Journal Watch--the watchdog journal Klein founded and edits--and an invitation to listeners to join a discussion of The Theory of Moral Sentiments by Adam Smith.

Russ Roberts (and Robin Hanson) on Truth and Economics
EconTalk host Russ Roberts talks about the role of empirical evidence and bias in economics and why economists disagree. Roberts talks about how his interviews with various economists at EconTalk have forced him to reassess the role of empirical evidence...
Dan Klein on The Theory of Moral Sentiments, Episode 6--A Discussion of Parts VI and VII, and Summary
This is the sixth and concluding podcast in the EconTalk Book Club discussion of The Theory of Moral Sentiments by Adam Smith. In this episode, Dan Klein of George Mason University and EconTalk host Russ Roberts discuss Parts VI and...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Russ Waddell
Mar 16 2009 at 9:32am

Thomas Schnelling’s model seems to agree with Professor Klein’s explanation of departmental tendency toward uniformity. The relatively small population in this case would reinforce the ideological position quickly and strongly.

As I look at graduate programs for econ, future dinner table conversations with my social democrat fiance are certainly a factor.

Mar 16 2009 at 12:15pm

I enjoyed both the Hanson podcast and this one, but I still can’t convince myself that a reasonable, open-minded person who looks at the history of the past few hundred years can fail to come to the conclusion that state solutions are more likely to be tyrannical and inefficient that those left to individuals, or if you insist, “the market.”

Item 1: From the reign of terror of the French Revolution to all the tyrannies of the of the 20th century, the more repressive and bloody they were, the more centralized was their power structure; item 2: Any comparison of originally similar places–north Korea vs. South Korea, Taiwan and Hong Kong vs. the PRC, East Germany vs. West Germany, surely demonstrates beyond doubt that market economies are friendlier to the realization of human potentialities than state-dominated economies, and, yes, that the most market-oriented economies are better on this score than those less so.

I still need some convincing that such a conclusion flows from bias and not the overwhelming evidence of hundreds of years of history.

Mar 16 2009 at 12:59pm

I think a really good reason the “overwhelming” historical evidence leans in favor of emerich’s view is interesting. Perhaps the historical school appeals more to the masses. Conflict tradition, in sociological theory, certainly gets heard the most because it screams the loudest. Empirical evidence is a separate issue and not as laden with emotion as the history books. I agree history is convincing. However, interpreting empirical evidence is not as fascinating. Can’t just curl up before bedtime on a cloudy night with an empirical analysis of X, Y, Z. It takes too much rigor. Much easier to read “All Things Negative All the Time.”

Based on my undergraduate studies, historical analysis has one side but empirical analysis has another. The dialectical disconnect between the two leaves us with endless debates; this is not a bad thing. Open debates can be good. Open discussions, by definition, though require open minds. I believe it was Friedman who said something like interventionists want freedom of speech for themselves but when another group claims an alternative explanation then freedom of speech becomes less important.

As a lawyer I USED to work for would say, “Your assumption is offensive to me.” Difficult to listen to another point of view.

Just a few loosely associated statements. I have not listened to the p/cast; I try to see what the audience says first. Nonetheless, I’ll enjoy the continuing discussion with Hanson.

From deep in the heart of Texas, thanks Liberty Fund, Russ Roberts, and all involved in the production of these podcasts. Students down here are learning from you. Keep it up. I’m sure the endowments will come at some point.

Mar 16 2009 at 1:27pm


Empirical evidence is a separate issue and not as laden with emotion as the history books. I agree that history is convincing. However, interpreting empirical evidence is not as fascinating as the history but it can be equally convincing, if not more convincing.

Side note: I’d like to hear a podcast on the economics of textbook publishing.

Greg Ransom
Mar 16 2009 at 1:38pm

The CENTRAL causal explanatory element in economic science is not formalizable — entrepreneurial learning and judgment and error in the context of changing relative prices and local production conditions. The KEY empirical element in all of economics outside of entrepreneurial learning are ever changing heterogeneous non-permanent productive resources — an empirical element which is also NON-FORMALIZABLE. In other words, the central causal elements of economics science as an explanatory enterprise CANNOT be put into mathematics.

But look at the filter/incentive structure of the economics profession — it’s all mathematics. The Ph.D process is nothing but a mathematics hoop jump. The easiest way to teach is to teach mathematics. The way to get published is to produce mathematics and “data tests”. All sorts of people have shown how this corrupts economics — Ronald Coase, David Colander, Friedrich Hayek, and on an on.


Economics is condemned to pseudo-science, and I see no way out. The incentive structure and filter system have given us lock-in on a formalist / mathematical pseudo-science that is not going away.

Alan Forrester
Mar 16 2009 at 1:44pm

I think a bit more thought about epistemology might be helpful. In particular I recommend “Criticism and the Growth of Knowledge” edited by Imre Lakatos and Alan Musgrave, which has papers written by Kuhn and some of his proponents and critics:


I think the best essay in this book is the essay by Imre Lakatos. Another very good discussion of epistemology an be found in David Deutsch’s book “The Fabric of Reality”.

Greg Ransom
Mar 16 2009 at 2:10pm

There is an important difference between conversation and argument. The conversational ideal upheld by Russ and Daniel is important. Conversations are uniquely helpful because they potentially take us closer to shared understanding. Arguments often are not helpful because not enough shared understanding exists to allow for productive interaction. On these topics I might recommend:

“Reasoning Together: Temptations, Dangers, and Cautions” by C Campolo & D Turner, _Argumentation_, 2002.

Abstract: “In the appropriate contexts reasoning is a powerful tool for producing intersubjective agreement about what counts as the best answer to a question that generates inquiry; sometimes employing arguments can lead to agreement about what is the right answer. In this paper we hope to show, however, that unabashed optimism about the power of argument is misplaced. Such optimism rests on an implausible picture of the power of articulation. Sentences cashed out as reasons to believe another sentence is true cannot bridge large gaps in substantive understanding. A failure to realize this fact, moreover, leads to an uncautious and unreflective optimism about the power of argument that ultimately threatens the very reasons-giving process. What is needed, then, is a much more modest sense of the role of argument.

Also — “Argument and deliberation: A plea for understanding” by Larry Wright _The Journal of Philosophy_, 1995

Lee Kelly
Mar 16 2009 at 4:37pm

The human eye has a blind spot where the blood vessels pass through the retina. Few notice, but by closing one eye, and looking at the right kind of image, some things seem to disappear. More interesting is that whatever disappears is replaced by something else. For example, it is possible to see patterns or lines where there is only blank paper, or vice versa. But this does not mean that such reports by our senses are false.

Evidence, our senses and data, do not report, suggest, or entail anything. Only an interpretation of them can be true or false. for example, if you account for your blind spot when interpreting your senses, then there is no error, i.e. your senses are “reporting” a true description of the world relative to your interpretation.

It is crucial to understand that evidence itself does not logically entail any particular interpretation. A sensory perception is not a proposition, and cannot serve as a premise in a logical argument until it is interpreted. But how to do this has nothing to do with the sensory perception itself. The same applies to data gathered by a statistician, or evidence taken from a scientific experiment. A particular interpretation can then be interpreted itself, and so on.

Evidence cannot serve to confirm, support, or verify anything, neither wholly or partically, certainly or probably, except relative to a particular interpretation. The depth of this problem is rarely appreciated, and while it pervades science in general, it is especially acute on the social sciences. What Kuhn called a paradigm is like an unstated, perhaps unconscious, set of underlying propositions or values with which one interprets events or evidence.

None of this should be a major problem. If we subscribe to some principle of fallibility, then that our interpretation of evidence might be mistaken is nothing earth-shaking (though it may put to rest some more naive epistemologies). Generally, we decide which fragment of our interpretation of the world we think is responsible for a mistake, and then tentatively refute and revise that fragment. If problems persist then we may begin to consider that something more fundamental in our understanding of the world is leading us astray, and identify other problematic areas of our thought.

Mar 16 2009 at 8:39pm

Great podcast.
I’d like to tease out a few points regarding the discussion of tradition vs. innovation and why the tension seems weighted towards tradition. I’m sure others have stated these points better somewhere else. However in case they haven’t I humbly offer my attempts:

1. Resource allocation (and opportunity cost): recall the podcast with Eric Raymond discussing the open source movement. When an argument concerning two branches of the software comes up the entire community is very quick to try to work out an agreement between the two sides because all parties realize that to continue with two versions is to divide resources. Similarly, I wonder how much implied cost the standard bearers of the main stream intellectual social sciences, humanities, etc. feel when toying with the idea of engaging with the minority opposition against advancing what they feel to be the agreed upon variant.

2. Crackpot refutation costs: My guess is that the implied cost (of engaging with the minority opposition) seems very large. Engaging the opposition almost certainly appears to be an extremely time consuming activity because even if the opposition’s work hold a few good ideas, addressing those ideas implies the addressing of countless other non-mainstream ideas, many of which will not turn out to be valuable and which are very far off the intellectual reservation to begin with. These costs might even appear larger due to the certainly implied and possibly true assumption that the mainstream attracts, in general, better students who turn out to be better professors, who will churn out have fewer technical flaws in their arguments and experiments. Semmelweis almost certainly represented a large cost to engage with in exactly this way.

The very interesting comparison for academia is the open source and wiki environments, touched on in previous podcasts, where there are strong feedback mechanisms on the governing administrators by the masses in terms of topicality, efficiency, and responsiveness.
It is helpful to remember that the academy has its history in the religious institutions of the middle second millennia. ‘Priesthood’ environments like academe’s fields within social science and the humanities are often very slow to reverse themselves as opposed to more market oriented environments because of the very disconnected feedback loop of their selection and retention/promotion criteria.

So an interesting question to consider:
Do both wiki administrators and open source software community managers experience something close to tenured positions in their respective communities after a certain point of engagement even though the groups have no formalized version of tenure.

Mar 17 2009 at 3:47am

One problem I have with some of these arguments is an artificial dichotomy that is created. One argument Klein keeps emphasizing is FDA vs. private regulation. This, I feel, is a false dichotomy. Without getting convoluted, it seems to state that any formulation that a state institution could create would be inferior to any system a free-market institution could create regardless of the market incentives. I’d be interested if anyone has a view on this.

Mar 17 2009 at 7:19am

I was struck by Klein’s dipleasure with questions directed to him that start with, “You know that Friedman economic theory has been discredited..” because in 90% of these programs Russ Roberts states that “Keynesian economic theory has been discredited.” These podcasts are generally fascinating, but when two GMU professors get together to smugly puff about their great wisdom, it is a waste of time.

Mar 17 2009 at 10:09am

AHBritton, check out http://www.freedomainradio.com
The creator and forum followers believe any state involvement is inferior based on the argument from morality, althouh the argument from effect is often used also.
The common theme of the website is Libertarianism/anarchism and there is some very interesting podcast, e-books and forum conversations on the site. Some times it’s a bit full on but still rather interesting

jim mcclure
Mar 17 2009 at 10:13am

I enjoyed the podcast. I have to say that I am much more optimistic than Dan is about the future of the economics profession especially because of the positives that Dan and Russ noted ARE going on at GMU: Econ talk, EJW, the blogs thanks Alex, Tyler, Don and Russ), and the fact that a decision has been made there NOT to chase rankings (this is absolutely a brilliant decision by GMU). Dan’s somewhat pessimistic tone is a probably a form of modesty about the impact of EJW. OK, fine, but I wouldn’t bet against EJW’s impact eventually proving Dan’s modesty (expressed as pessimism) being almost as misplaced as Ehrlich’s pessimism about future availabilities of “natural resources”.

Thanks for a great podcast!

Daniel Klein
Mar 17 2009 at 4:56pm

Thanks Russ for having me, and everyone for listening and commenting. Much appreciated.

Some replies:

To ScreenNom: Do check GMU groupthink, but with specifics. I didn’t speak to Keynesian theory.

To AHBritton: On the FDA, I’d say the key distinction is freedom vs. coercion. When Congress banned all new drugs (till permitted by the FDA), it stuck a gun in the face of would-be makers and sellers of meds, people who had coerced no one. Congress initiated coercion. I’m not saying that initiating coercion is always wrong , but I definitely think that, even the repugnance aside, it is not redeemed in this case – very far from it.

To Alan Forrester: Funny, I was just reading Criticism and the Growth of Knowledge. But I take Kuhn over Lakatos. For me, Kuhn’s response is the highlight, full of gems. His remarks about Sir Karl not being really any less vague than he nicely pertains to talk of “economic efficiency” – it’s actually much vaguer than economists like to think or admit. I like pragmatism — Kuhn, James, Rorty, Quine, and McCloskey. Others I think of this way include Adam Smith, M. Polanyi, F. Hayek, and R. Coase.

To Greg Ransom: Yes, knowledge’s richness, particularly asymmetric interpretation, can’t really be captured in an equilibrium model, which generally assumes common knowledge (not perfect information!), which is symmetric interpretation. Yes, the math stuff is way over-valued, and, as Pete Boettke and many other Hayekians say, the math modes relate to the prevalence of bad judgment on policy (not that you need a math fetish to have bad judgment). As for the distinction between conversing and arguing: Besides seeking shared understanding with one’s interlocutor, best achieved by conversing, there is seeking influence and understanding with the audience, the spectators, as they hear you reduce the interlocutor to rubble. Arguing isn’t as confined by congeniality as conversing is. Some people overrate congeniality, I think. In fact, I think Adam Smith does. But I don’t mean to favor arguing/debating over conversing, both are good and too scarce. I’m just saying that I don’t think it’s obvious that we should favor conversing over arguing/debating.

To Lee Kelly: What you write sounds fine to me, but it doesn’t seem to speak to major disagreement – Smith vs. Marx, Hayek vs. Keynes, Friedman vs. Galbraith. – Though it’s not clear what to say about major disagreement.

To JP: What you say about how programmers deal with open software bifurcation is interesting – I need to listen to the Eric Raymond podcast. I do think there are parallels with academic culture, especially lock-in. And your point about the elite’s costs of engaging “crackpots” is well taken. On the other hand, if they really are the champ, and people suspect the game is rigged, shouldn’t they take on a Rocky Balboa now and then?

To Jim McClure: You are kind. But how exactly are the innovative media to change what the economics pyramid does and is? Perhaps there is influence on the new generation, but what evidence of that?

Thanks again for reading and writing.

jim mcclure
Mar 17 2009 at 9:26pm

The pyramids are tombs made of stone that have not moved for thousands of years. Your analogy, “the economics pyramid”, I submit, is hyperbole, part of your modesty/pessimism. Paradigm shifts and dominance hierarchies can and do occur suddenly. As you are aware, in the 1970’s the percentatge of critical commentary articles routinely was above 25% (in the AER, EJ, JPE, and QJE) in the past decade it is routinely below 10% for these journals taken in the aggregate. As chaos theory illustrates about the butterfly in BEIJING,so it may be in our discipline (the tipping point is as real as it is difficult as it is to anticipate chronologically).

My view is that something as trivial as the foundation of the dominance hierarchy you analogize as a pyramid, is likely to crumble as rapidly as the Maginot Line before the Blitz.

The evidence of history illustrates again and again that the “might” have fallen. If I were you, I wouldn’t bet on the perveyors of ultra-mathematically complex, game theoretic, multiple equilibrium models to be dominating economics 15 years from now.

Lee Kelly
Mar 17 2009 at 11:18pm

Daniel Klein,

My comment was primarily concerned with addressing the role of evidence in science, and by extension, economics.

Evidence, whether simple observation or complex statistical data, does not suggest, entail, support, verify, make probable, or justify anything. Before evidence can function as a premise in a logical argument, it must be interpreted. But there are infinitely many possible interpretations. Only habit, custom, bais, ignorance, or whatever, can lead one to think that the evidence speaks for itself.

The language of argument is infused with language of war and violence. Implicit in many arguments is the belief that a evidence should be compelling, that good arguments are forceful, that any evidence or argument that can be doubted, dismissed, or deflected in some ad hoc manner is faulty. Opponents in a debate defend their castle from all attacks, and will not surrender unless coerced by the force of reason.

It is a totalitarian vision of rationality, where no inidividual is responsible for deciding what kind of evidence or argument will be accepted in debate. It is even seen as a triumph to reinterpret evidence in a way that is compatible or favourable to your preferred ideology. A good argument, in this view, should leave no room for anything as irrational as a decision. If an argument does not compel, then the argument has failed, rather than ourselves.

I see this attitude so often. Those who practice philosophy seem to consider it a conclusive refutation of almost any conjecture or criticism that they are capable of doubting it. But to insulate theories from criticism and refutation is the easiest thing, done simply by interpreting away any inconvenient evidence or denying that any standard if criticism is acceptable. But theories are only refutable if you choose to make them refutable i.e. clarify the problem which your theory is an attempt to solve, and then specify what kind of argument or experiment could be deployed as a test of that theory.

It seems to me that, when this attitude is eschewed, a more rational debate can take place, even between those with differing “ideological commitments”, because I do not think that any such commitments are necessary. We can argue all the way down (our ability to entertain hypotheticals is much undervalued). But it is important for those who wish to take part in rational debate that attachments to ideologies are avoided.

Mar 18 2009 at 3:03pm


Economists have very different policy opinions than most of their fellow citizens, and most are very interested in changing the opinions of the citizenry to agree with their own.

Economists are also very very eager to quantify myriad types of human interactions, to the greatest extent possible.

Why don’t economists have a quantified approach testing the efficacy of various strategies in aligning the opinions of the citizenry with their own?

(seems like great work for someone like Brian Kaplan)

George Peacock
Mar 18 2009 at 3:48pm

Another helpful podcast. I have not read all of the previous comments but would be interested in any information about a potential “book club” as mentioned by Russ in the podcast.

Vjim Filomena
Mar 18 2009 at 5:17pm

Wonderful work, Russ. Been listening for about 6 months; good timing; great info.
Would be interested in book club !

Love, Vjim

Mar 19 2009 at 4:22am

I would like to repudiate the fatalism of Klein and other commenters and echo mcclure in pointing to what Russ said was a sign of hope: new technology. Klein is already on the leading edge of this with EJW, publishing online and avoiding the traditional gatekeepers. The pessimism shown is particularly amusing given that we are on the cusp of an information revolution, that will destroy existing socialized institutions like academia/education and medicine, which have accreted stupidity and laziness over decades of being isolated from the market, just as newspapers are being destroyed today for the same reasons. In academia, online learning will become dominant and universities will be decimated. Research will be funded by private research institutes and researchers won’t have to waste their time teaching, something they’re usually horrible at. It’s very likely that the market for intellectuals is currently artificially inflated by govt subsidy and will shrink greatly once subject to market forces and the weak property rights inherent to intellectual work. Many people who are currently employed in academia will find other, presumably more useful work in an information economy so all but the completely useless ideologues will be fine. One thing’s for sure, a punctuated equilibrium in education and other information markets is coming. My analysis is that this future will be decidedly more libertarian than the past, both because of some fundamental properties of the new technology and because some old arguments in favor of free markets will finally be fully realized.

Mar 19 2009 at 5:16pm

Reading, again, The Theory of Moral Sentiments strikes me as a good idea. So good that I’ll be reflecting on my re-read at my own website. If and when you start the book club up, I will undoubtedly comment on your debate at my own site.

The institutional nature of change is absolutely important. It is worth researching how liberalization ever works. How did it happen in the late 18th century and early 19th century? How has it happened, in places, here and there, since those “early days”?

I wonder how much influence the background of the religious wars had to the Enlightenment. By my reading, it was inter-church/inter-state warfare that provided the Big Fact against which most ideas percolated. The critique of mercantilism was successful (to the extent it was . . . after all, in America the mercantalists took control after the new Constitution was put in place) in part, perhaps, because it conformed to the general idea of liberty which seemed the most natural way to curb the excesses of religious strife.

One might think that the fall of the Soviet Union would have provided something similar to the West, but it didn’t turn out that way.

Mar 20 2009 at 3:54pm

Mr Klein spoke repeatedly against the FDA creating barriers to new drugs. He also spoke on the importance of disclosure. Did he happen to disclose if he had a relationship with any drug development companies?

It seems to me that the FDA is the best friend of the drug companies given its role to legitimize drugs in the marketplace. Otherwise nobody would put themselves for example through the torture of chemotherapy.

Mar 20 2009 at 5:42pm

Great podcast; I was fascinated by Klein’s point that economists who have published a policy preference on a certain issue (e.g. the FDA) tend to be more liberal in their recommendations than the average of randomly polled economists. This seems like a corollary of Brian Caplan’s research showing that the more people become economically literate in general, the more libertarian they become and is something that I think deserves a lot more attention.

Is there anywhere I can find more about this phenomenon? Are there any counterexamples (i.e. issues on which economists who have studied it favor a more statist solution than those who have not). Prof. Roberts, would you consider elaborating upon this idea in the future?

Mar 20 2009 at 6:48pm

This interview is fantastic. One can throw out any point that is politically distasteful to their particular liking and still gain much insight from a much broader notion that the inclusion or exclusion of political or social leanings are intrinsic when considering opinion and commentary. I am going to pass this particular piece along to some friends of mine who seem to believe that Keynes was the Jesus Christ of modern economics rather than simply a good cowboy in a certain rodeo that was void of talent at the time.

Outstanding job.

Don Crawford
Mar 21 2009 at 9:04am

The same university department biases are true in education where progressive educational philosophies prevail and critics are seldom hired, published or tenured. The conservative think tanks are providing the only home for those who don’t worship Dewey and Piaget, but think tanks don’t produce new Ph.Ds, so the balance is not changed.
I agree with the point that argumentation is not always especially enlightening–especially when each side simply puts forward its most “compelling” arguments and fails to engage the claims of the other side.
As I understand the relationship between empirical evidence and theory, a theory should produce testable hypotheses if one is looking for them. Sometimes it takes a lot of creativity to find such. (In education the government monopoly schools do not wish to engage in public research because their mission is to appear effective and ideal in the eyes of the public and experiments run the risk of showing that their policies are wrong–before the administration has had a chance to change.) It seems to me that a valuable exercise, when economists of differing theoretical “biases” get together is to search for some testable hypotheses–if their biases are not faith-based then they should be falsifiable. What empirical test of the theory’s predictions would cause its proponents to give up the theory?
One other observation. Statistical gatherings common in economics (like John Lott’s work) are often subject to so many conflicting interpretations that the data set often are quite unconvincing. But isn’t there a body of research in economics where simulations (classrooms of students) can test economic behavioral predictions? Didn’t someone get a Nobel prize for that kind of research? And wouldn’t that sort of experiment be a more fruitful venue for investigating some of these theoretical predictions? Especially if economists with differing theoretical biases joined together to design the simulation (I know that would be unlikely) so that each side felt it was a fair test of their model’s predictions?
Finally, listening to EconTalk has been a real joy since I discovered it a few months back. It seems odd that I like it–can’t explain to my friends–but I really do. And I’d like to recommend Peter Wallison on the subprime crisis as his presentation in December at reason.tv was exceptionally clear and lucid. I’d love to hear Russ delve into it with him further.

Mar 21 2009 at 3:02pm

Thank you for another interesting podcast. I am a non-economist (a lawyer by trade) but over the past year I have delved into the subject, reading extensively, working through a couple of textbooks, keeping current with a number of blogs, and listening to this podcast. I find the study of economics fascinating and — at the risk of hyperbole — intellectually exhilerating. It teaches a whole new way of looking at the world. Thinking overtly about incentives and their role in both our problems and their potential solutions is, I think, the great contribution of the economics profession.

Economics has its limitations, however. In law school, they taught us to “think as lawyers”. As a way of thinking, legal analysis is also very valuable. A problem arises, however, when a profession (in its collective mind) elevates its particular viewpoints and approaches to a level of preeminence – something that I think happens all to often in the legal profession and in economics.

Things are compounded when proponents, of say an economic or legal solution, began to assume that they have a “correct” solution to problems that are human (and thus unquantifiable) in nature. I doubt that we will ever truly understand human behavior or the problems that go along with it. Will we ever truly understand the cause of crime waves or the business cycle? I am equally doubtful that we will ever truly solve crime or create a perfect economy. Yet, people want solutions and there has never been a shortage of those who are willing to serve as “priestly caste” to provide the answers and solutions asked for.

A problem arises when the priests (todays lawyers, economists) present their views as infallible or seek to make a science out of subjects that are at bottom philosophical. Economics, it seems to me, took a wrong turn when it began relying so heavily on math and tried to turn itself into a science (for the “free marketers” out there, maybe you can blame Samuelson). Mathematics, although valuable as a tool, is based on assumptions that are in the end highly subjective. Can economists, for example, predict or measure with accuracy the multiplier effect of a dollar spent on road construction? Can they accurately measure the marginal utility of an additional movie ticket to a particular consumer? Personally, I doubt it. As a layperson, I find dubious the proposition that math can provide true answers to these types of questions. Mathmatical models may be interesting and useful, but I think it is a mistake to confuse them with scientific precision.

Economic methodology is valuable, I just think we should take its results with a healthy does of skepticism –particularly since the biases and philosophies of individual economists seem to play such an important role. This podcast, for example, rarely fails to provide valuable and intriguing conversation. It would be disingenuous, however, to suggest that it is completely unbiased or neutral. EconTalk is not “unbiased”. For the most part, it reflects (what I presume to be) Russ’ predisposition towards an individualistic, free market oriented view. But that is not a criticism. I listen to Russ and his guests and agree with much of what is said even though I am philosophically drawn towards the view points of more “liberal” economists such as Krugman and Stiglitz. The opinions expressed on EconTalk are not wrong — by any means. Rather, in my view any way, they emphasize one set of truths. Individuals are, for the most part, best at determining their own self-interest. To the greatest extent possible they should be left alone to act in their own interest and solve their own problems. Those points, I think most would agree, are beyond debate. The deeper question centers around the factors that can disrupt the individual’s quest for personal and economic freedom. For some people, the biggest danger comes from centralized authority that, according to their viewpoint, is more often than not incapable of providing for the individual good. The truths expressed in this view are emphasized on EconTalk. In the year that I have been a listener, however, I have not heard much discussion of other factors that can limit individual choice, such as asymmetric information, principle-agent or adverse selection problems. I don’t know for sure, but my guess is that the more free-market oriented economists do not place as much emphasis on these problems as their more “liberal” colleagues. That is not to say that one school of thought is right and one is wrong, but rather that ultimate beliefs as to human nature and psychology play a large role in economics. It seems to me that we will all be better off when this fact is acknowledged.

Mar 22 2009 at 8:11am

Except for a small percentage of lawyers, economic thinking usually runs counter to “thinking like a lawyer.” Lawyers are adept at splitting hairs and finding the smallest loopholes.

Stiglitz and Krugman are excellent examples of great economists turned lawyer.

Mar 24 2009 at 6:03am

I enjoyed the discussion and was intrigued at the theory that people don’t change their ideology/opinions much after age 25-30. Has there actually been a study done on this? I’ve always heard the conventional wisdom line about how people supposedly move from liberal leanings to conservative leanings as we age. Wasn’t it Humphrey who said something like ‘If you’re not more liberal when you’re young, you don’t have a heart; if you’re not more conservative when you’re older, you don’t have a brain.’ (hopefully I didn’t butcher that quote too much).

And yes, I think some kind of book club/discussion format for “The Theory of Moral Sentiment” would be good. I’d certainly try to participate/follow along….especially when you’ve made the book free via download.

Simon C
Mar 30 2009 at 4:34am

Another interesting example of how scientific results can be influenced by the scientist’s preconceptions can be found in Feyman’s Surely You’re Joking Mr Feynman.

“One example: Millikan measured the charge on an electron by an experiment with falling oil drops, and got an answer that we now know not to be quite right… It’s interesting to look at the history of measurements of the charge of the electron, after Millikan. If you plot them as a function of time, you find that one is a little bigger than Millikan’s, and then the next one’s a bit bigger than that… until finally they settle down to a number which is higher.
Why didn’t they discover that the new number was higher right away? It’s a think that scientists are ashamed of – this history – because it’s apparent that people did things like this: When they got a number that was too high above Millikan’s, they thought something must be wrong – and they would look for a reason why something might be wrong. When they got a number closer to Millikan’s value they didn’t look so hard. And so they eliminated the numbers that were too far off, and did other things like that. We’ve learned those tricks nowadays, and we don’t have that kind of disease.”

It may be that the hard sciences have learnt to avoid this manipulation of results. But when you consider that just the potential embarrassment of giving an usual value was enough to cause the manipulation of results in the hardest of sciences, what can it say about the value of results in macro economics or climate science where the data is so much more ambiguous and the political pressures so much stronger?

David S
Apr 8 2009 at 3:17pm

I find it really interesting that even in the midst of your wonderful discussion about how biases and groupthink work, your own root biases were so clearly on display — thus proving your point that these are inescapable!

In particular, your use of the word “coercion” with respect to action by the state is itself a blinking LED sign screaming “I’m a libertarian!” It frames the discussion right from the start — I mean, who could possibly be *for* coercion, right?? Action by the state must therefore be an evil only to be allowed as a last resort.

This is a moral assumption though. It lies outside the realm of logic and provability. That doesn’t mean it isn’t important and I mean no criticism by pointing out your moral beliefs. We all have them, indeed we *must* have them. All our logic and reason rests on a foundation of moral value assumptions. That is, those things that we hold to be self-evident.

The problem is, of course, that we all subscribe to somewhat different basic moral tenets. Most of us don’t seem to understand this so we shout over eachother in the incompatible languages of our own belief systems. This is the heart of what’s always frustrated me about the abortion debate, for example.

As you say, there’s no way out of this. But we can at least try to figure out where everybody is coming from and speak on those terms.

Anyway… I love the podcast. Keep ’em coming!

Dave S
Apr 8 2009 at 4:04pm

To emerich,

In the spirit of this show, I think your confirmation biases are showing.

With respect to the french revolution, ok, so it was pretty tyrannical, but what does that prove? It seems like a big stretch to draw any conclusions either way about the relative merits of the state vs. individuals from that. Would you say that power more or less concentrated under the directorate that it was under the monarchy that came before it? I dunno. For that matter, it was the ultimate embodiment of concentrated power called Napoleon who finally enacted something like what we might call a liberal economic system which France then flourished under. So what does *that* say?

Besides, what about the hybrids? A somewhat more state-oriented economic solution doesn’t have to mean a more tyrannical one. That seems to me to depend more on the decision making apparatus. Dictatorships and one-party systems like Nazi-germany and the USSR are absolutely not the same type of creatures as the social democracies of Europe today.

I for one grew up in the US but live in Sweden now, which I can say anecdotally is a wonderful place to live.

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Books:


Web Pages:

Podcasts and Blogs:



Podcast Episode Highlights
0:36Intro. Hanson podcast continuation. Disagreement, sciences and social sciences, [taping Feb. 5, 2009]. Is there progress in the social sciences? Pessimistic. Public policy, ideological in nature. Can't separate one's own sensibilities about those matters. Culture's too socially democratic. Sort of group-think problem writ large. History, psychology, sociology, group think dynamic going on. Broader society-wide group think, descended since decline of classical liberalism. Unravel threads, revealing one's own ideology. Resented being labeled as a conservative economist. Have an ideology, but that's different from my economics, which is science. More realistic and think about my biases over the years, but most economists would disagree. Journalists similarly think of themselves as reporting objectively. Cheering in the press box in sports, but most media like to think they are independent seekers of truth. Naive, illusion, goes against human nature. Then it's only a question of what label to use. Smith-Hayek label, which doesn't convey anything. Gunnar Myrdal, Nobel Prize winner, social democratic economist, wrote Objectivity and Social Science, useful to reveal one's biases, disclosure. Frank Graham, Peter Lewin also have endorsed self-disclosure. Top 1000 economists, though, would say "I got a Ph.D. from x" where x is a top-five Ph.D. program. Wouldn't reveal bias. Just need somebody well-trained. Self-deception, disingenuous; know there are certain people you hang around with at a cocktail party, those with similar ideology. Naive, not thinking hard about the norms in social science, evolved to exclude certain kind of discourse, within the 40-yard lines, focus, anything outside that talking about more radical reforms is being an advocate. Naive about how normative they are.
10:00Group-think issue: Now that more open and honest, call self a free-market economist. Better understood. Problem with this honesty strategy is that it is prone to ad honinem attack: You're like Milton Friedman, or Free markets have been discredited. Cocktail party phenomenon: does that mitigate the tendency to be honest. Can rub people the wrong way or be self-important. Can draw you into larger issues, changing the topic. Distracting. Also issues of group think within the group. "Group think" is pejorative by construction, like the term rent-seeking. Defectiveness of the ideas is being presupposed by the term and the analysis. Free-market libertarian circles, "our circles"--don't fall easily into a box. Not so much that the policy judgment is defective but the argumentation for it is defective. Think tanks can be kind of hackish but may agree with the punchline. Plagues all of our groups. Defend group-think: Virtue is that it embeds a conservatism in the evolution of what is thought to be right. Can argue that that's a good thing, better to move in small steps. Counter: Hazards of not enough group think is big Thomas Kuhnian theme; Deirdre McCloskey plays up; the essential tension between tradition, keeping to established focal points, and innovation. If you allow too much to the latter, nothing is focal, enterprise no longer coherent. Convention vs. innovation. Faith in the leaders of the group, the apex of the pyramid. Worry because economic science coming to judgment on those the right answers are not being arrived at, well-formulated, and conveyed. Example, Food and Drug Administration (FDA): apparatus, structure of banned until tested: Friedman, Becker, other economists, Klein, who have studied the FDA argue for liberalization. Count noses, most economists are for liberalization of the FDA. If you poll economists at random, though, they are supportive of the FDA. Specialists versus people who are just interested. Issue-expressive economists reach a conclusion that's different from the mainstream. What is going wrong in the culture?
20:03Put you on the couch: Maybe you're wrong, FDA is a great thing, maybe experts are wrong. What's the source of your confidence? Entry-level argumentation: rationales for the FDA, no good market-failure rationale, economies of scale, can't do your own testing, need independent body, role for government. Last part doesn't follow: nothing special about government. There are market-failure arguments for government action: government in its specialness brings something to the problem; generally speaking its coercive power. Externalities, free ride on them--government can bring its power to tax. But in the FDA case the government doesn't bring anything to the table. Market might be imperfect but rationale has to go further than that. Ground-level details; but higher level. Can turn to all these figures, important guys who do this, make their case. Where is the other side? Recent book in defense of the FDA; but confining discussion to economists. Start picking and choosing; no way out of this.
24:29Sociology of academic life, why more skeptical of mathematical work than mainstream. Could be wrong; or other side could have set up rewards that make it harder to get a hearing. Academic group thing in terms of ideology: In minority in economics, Smith-Hayek contingent is maybe 10%; in other social sciences practically no free-market people. To the extent that they deal with policy and politics, big problem there. Defectiveness, starting point for suspecting group-think; group-think presupposes it; feeling that there is defectiveness: FDR saved America from laissez faire is common view in history. How could it be that defective ideas can persist? Two mechanisms, one more micro and one more macro. More micro: What goes on within a department. Departments are very autonomous, majoritarian politics. If a majority has a certain point of view and if they believe sharing it is part of the scholarly judgment, if 55% in hiring, become 60% and then 65%. GMU econ department is pretty uniform, similar view, lots of diversity but similar. Scientific judgment ought to reflect that. Tendency toward uniformity. Where does the department look to for its norms of fairness, excellence? Looks to wider discipline. Rankings of departments, better schools looked up to and followed, also journals. Pyramid that leads all the departments. History Department is a creature of history, not so much in being a George Mason employee. Making and placing of new Ph.D.s: large departments at the apex of the period.
32:35Strange viewpoint for a market-oriented individual to have. GMU wisely has not tried to be a top department like others; only 20 schools can be in the top 20; most in next rung are vying for excellence. Fool's game. Story: Department that wouldn't hire someone who couldn't get a job at MIT. GMU eschewed that strategy; do something different and do it as well as we can. Moneyball strategy, looking for a niche. Why don't departments do that? First, talking about culture, network externalities kind of thing; furthermore talking about political culture. Power of politics plays a role, creating prestige and status. Something like 70% of college teachers in the U.S. are government employees. Research money comes from the government. A lot goes wrong in analogizing the market for historians with the market for waiters. Waiter department, waiters minted like new Ph.D.s, etc. Some parallels, but cultures go wrong institutionally and individually. Beliefs about religion and politics, people rarely change their mind after age 25 or 30. Taboo, carry on as if there is hope for changing the minds. Who are the cultural leaders? Not going to be the neophytes, but those who are locked into their beliefs.
38:12Story: Example of Semmelweis, doctor who was alarmed that so many women were dying in childbirth from puerperal fever. Theory: doctors would go to the morgue and examine women who had died and then go deliver a baby, passing on the infection. Suggested washing hands. Profession didn't agree, thought cause was vapors and instead suggested closing windows in that section of the hospital. Book. Semmelweis designed experiment. Not the most pleasant fellow--often the case with renegade thinkers. Did experiment half-heartedly, which gave an opportunity for older doctors to dismiss it. Eventually world came around to his opinion. Raises troubling story about the entire scientific enterprise. Even in the physical sciences, it is hard for truth to out. Data matters. Kuhn's book: eventually generation dies. A lot of truth to that. Keynesian revolution was viewed as coming in when the old guard died out. Iatrogenic malady: malady from the treatment. Other mechanisms persist because the older generation in academia trains the younger generation. What role does empirical evidence play? No Hayekian forces that would make cultures get better. Pessimistic story. Some hope. Not that there aren't Hayekian forces, certain power to wisdom which gives it a chance in the fight, but it's not a free market. Government is a huge cultural player. World isn't really an infinite space. If things aren't getting better, could just be that Hayekian forces are overmatched by other forces. Some hopeful things--like EconTalk! Better quality discourse. Talking plainly, back and forth, publically so that accountable.
46:34Prediction markets--Robin most passionate about them. Great example of possible meta-evidence. Simon-Ehrlich bet on resources running out: Paul Ehrlich on Johnny Carson, vs. Julian Simon, no attention, and said there should be a profit opportunity in the form of futures; should be willing to bet me that prices will be higher in the future. Ehrlich agreed; picked five metals to bet on inflation adjusted prices. John Tierney, NYTimes; Simon won on all five. Simon told Klein he never cashed the check, put it on his wall. Putting money behind opinion conduces to better opinion-making. Limitations: problems in what you can actually bet on, what kinds of metrics you can set up to look at down the line. Always will be some guy who got the last 20 coin flips right--he's the guy who knows. Some things don't pay interest, so no one wants to take those bets. Next big spending package: if you are voting for it, perhaps your compensation should be based on the unemployment rate that results over the next six months. Would play to the test. Teasing out problems, so many factors. Interest: regulations might stand in the way of paying interest.
51:34Econ Journal Watch: innovative kinds of discourse. Very little real criticism in economics, especially of the most influential people. Get more back and forth, debate, which has disappeared from the top journals. Australian econ journals similar. Criticism of work and criticism of institutions and practices. Sociology of academic economics. One section: Do economists reach a conclusion? Look at whether issue-expressive economists reach a conclusion relative to random economists. A lot of people say economists don't agree; but that's only one consultation. A more meaningful consultation is with those who publish judgments. Makes yourself more accountable to publish as opposed to filling out a questionnaire. In general, these do reach conclusions more than random economists. Judgments toward liberalization. FDA, licensing, road pricing, rail transit, drug prohibition, postal services: in all six, issue-expressive reach conclusion more preponderantly than a random economist. If there is an impasse with random economists, there's a problem. Why don't other economists know the findings of their own profession? Why aren't the findings more publicized? Power in society is part of the problem. Crucial journals aren't willing. Journal of Economic Perspectives somewhat better. How successful is EJW? Seven Nobel Laureates on the advisory committee, indexed. More variety in the media, widened voices available to people on economic policy, world is a little more competitive.
1:00:12Philosophical question: Once people turn 25 or 30 they get locked in. Large disconnect: three broad categories for why people believe what they believe. Evidence: empirical work, facts, experience, studies. Logic: common sense, consistency. Narrow self-interest, better friends at parties if you hold this or that view. Example, Minimum wage: evidence, morally wrong, injustice of the market, my friends all believe it. Categories are not all that distinct. Evidence entails logic. What we can articulate as our reasons emerges from deeper judgments and sensibilities that are not yet articulated. Dialectic. Important to listen to the people you disagree with are saying. Engagement. What parts of their beliefs are they putting forward that you could get them to reconsider? Web, 90%-true statements; counter arguments don't disprove the claim. Smithian way to look at it. Confirmation bias. Sowell, framework used to process. Would like to hear from listeners: Smith's Theory of Moral Sentiments. Any interest in book club reading and talking about it? If interested in pursuing this, email mail@econtalk.org.