This is an extremely fair and sophisticated reading of the lay of the land in economics, with a deeper grasp of strengths and limitations than most insiders! I would assign this to starting graduate students without reservation or qualification.
What I observed during covid (in Canada), was that politicians openly demurred to health authorities to determine government response, rather than the health authority providing options and trade-offs. Now, those experts certainly know health science better than me, the public, or the politicians, but they don’t necessarily hold the values that would result in the trade-offs appropriate to the public. I think something similar has occurred with economists dictating interests rates based on theories.
In biology/ecology, experts are never given authority to dictate large scale projects, their job is to define risks and impacts of different options. The advice tends to be very specific to the project. But what is often wanted is simple, broad scale maxims. However, as the issue broadens, generalisability decreases. I think that’s where economists are similar. They will be very good at predicting the effects of things like tariffs, but at broad macroeconomics their values are part of their models, and they have very few tools for large complex systems.
I think the distrust is driven because expert advice has been followed and many people feel worse off than before. I think in part that is because there are very broad frameworks that economists work within, and they don’t all work within the same frameworks (ex neoliberal, Keynesian, socialist, etc.). Therefore there are major differences in assumptions (ie how efficient are markets for public good). Hence, scaling between macro and micro economics has major implications in accuracy! But the public mostly just sees micro and macro as the same thing.
Possibly this explains a growing distrust of economists as a consequence of a more general distrust of expertise, caused by (but maybe not justified by) cases such as these.
This is where it matters that Dan mentions the IGM surveys are usually on positive (descriptive) questions rather than normative ones! It seems to me that what we really wanted for covid policy was for the epidemiologists to tell us how much case counts would be affected if we did things like close non-essential businesses, or impose curfews, or mandate masks on the subway, while economists told us how much each of these rules would affect other things we care about, and then elected officials use these, together with their sense of public opinion, to decide the best trade-offs. The normative views should come from the democratic public, and the experts should only provide positive descriptions.
Now as a matter of fact, the very first survey I glanced at on the IGM page Dan linked was in fact explicitly normative, asking whether the risk of harm is such that most people would benefit from regulation of data privacy.
Yeah it's tricky. It's pretty clear they're always trying to get at the normative question of whether some policy is a good idea, but they *usually* ask a nearby descriptive question. E.g., "would increasing the federal minimum wage to $15 substantially increase unemployment?" But even then, there's wiggle room around "substantially", which people might be importing normative assumptions to interpret. In their most recent minimum wage survey they took it out, and the results were substantially (haha) different from than the last time when they'd asked a similar question that included "substantially".
So basically, while they *usually* try to ask descriptive questions, they also often make them vague or open to interpretation in ways where you could imagine people are drawing on normative views in deciding just how to answer. And of course sometimes they just go straight for normative questions about whether costs exceed benefits, as in the one you found.
I love this essay. It almost seems as though it was written by a good philosopher, instead of a good economist.
The autobiographical point at the end generalizes to a lot of things in life. "Yes, this thing here kinda sucks in some major ways, but Jesus Christ it's better than the alternatives". I trotted out that response many times years ago when philosophers were fighting over the Philosophical Gourmet Report.
In physics experiments effects are typically observable almost instantaneously. That makes it a lot easier to deal with unexpected external influences by modifying the experimental set up. In economics, effects can take months or even years before they are visible. The life of economists is too short.
Economics is pretty good at figuring out why things happened in prior times. And we know those who ignore history are doomed to repeat it. In this way, economics provides useful tools for policy-making. But one of the lessons of economics is that all sorts of significant events are functions of policy that failed to account for some factor or another given limits to what is knowable about the future. While evolutionary biology can explain much about evolution to date, it provides little insight into the next evolutionary adaptation. Similarly, economics is limited in its ability to design policy or predict the effects of a given policy in the future.
This is a great piece! Speaking personally, I think an additional reason I give more credence to the consensus in economics is that the discipline seems much less ideologically captured and much more open to following the evidence, changing their mind, looking for evidence against their views, etc. Economists seem to generally be the best social scientists, way better than the rest!
This is a good point. Indeed, I think the mathematization of the discipline is a reason why the discipline has avoided ideology capture.
If an ideologically-inclined economist is like, "capitalism is inequality" (or on the flip side "markets are intrinsically fair"), then the rest of the community is going to be like, that's a nice slogan, but what's the model? What do you mean by "capital"? What kind of relation are you positing between that and inequality? What metrics are you using? What are the timescales? What are the mechanisms? How do you model those?
The disciplinary effect of mathematics forces economists to get a lot more precise in their claims and precise claims are anathema to ideologues.
I think you're understating the case for engineering successes by economists. It's been pretty obvious if you look at the history of the 20th and 21st century that governments that largely adopt the policies that are recommended by mainstream economists (liberalization of markets with clear property rights, central bank independence, inflation management, low tariffs, manageable taxation, investing in innovation, etc.) fare much better at reducing poverty and growing wealth. Examples abound.
The Chinese growth following the Deng reforms. The post-1991 economic liberalization of India. The post-Cold War growth of the Baltics and Poland. Singapore.
Consider also the Chicago Boys reforms of Chile. Since those reforms, poverty dramatically fell in Chile and Chile has consistently outperformed the rest of Latin America. However, this episode is often brought up as a black mark against economists because they collaborated with a military dictatorship that overthrew a democratic government. Yes, that's certainly ethically dubious. But if economics is an effective science, then it's deliverances and effects should be independent of the ethical status of its users, much like in the case of physics or chemistry. And I think that's what's the case here.
And on the other side you have countries that didn't or don't follow mainstream econ and they, well, fare much worse. Take for example Venezuela, Zimbabwe, or Peronist Argentina. (And in the latter case, we are seeing that the econ-educated Milei is working to bring down inflation.)
I think this indicates significant success in social engineering, and has helped comparably many people as the engineering effects of physics.
While I tend to agree, I also think these cases tend to be too debatable to cut all that much ice when it comes to present debates. E.g., I've seen people appeal to the success of the Asian tigers in just the way you do here, but I've also seen fans of industrial policy say: "sure, they liberalized in various ways, but they were far from laissez faire, they protected infant industries, etc." So in the context of a present-day policy debate about some matter where economists largely agree, there's room for disputing the extent to which the earlier analogous policies that were followed by success amounted to following standard economic advice.
In your article you juxtapose economic analysis with "alternatives—relying on my own intuitions, partisan punditry, or critics armed with cherry-picked case studies". I know that that was the part of the article in which you'd given up straw-manning, but boy, did you give it up!
Part of the problem is paradigm wars. It's true that a lot of the critics of the current state of economics have some new paradigm they want to foist on us. I find that pretty laughable - like jumping out of the frying pan into the fire.
For me anyway, the problem is always our own ignorance and the quality of our critical awareness. The current practice in economics is to ignore anything that can't be translated into the state of the art of formal modelling. In fact the discipline did deal with this question - of where its methods added value and where they didn't. Modern economists just ignore it.
Thus as the new mathematised discipline was being forged - one of its architects - John Hicks warned that there were some things it might be good for and other things it wouldn't be good for. But the method became the "proof of work" within the economists' guild - and "proof of stake" come to think of it!
And so the method of formal modelling spread into areas where it had little but mystification to offer.
So bits of the simplest commonsense got consigned to invisibility at the commanding heights of the discipline. Commonsense like that scale economies are an important determinant of trade flows.
As Hicks had warned as early as 1939, as scale economies grow in importance, firms gain pricing power and the methods of economic theory encounter what Newton encountered with the three body problem. The analysis gets engulfed in complexity almost immediately.
Did economists face up to any of this? Well, they did for a while, but the disciplinary incentives eventually overwhelmed the thoughtfulness. Economists went on with their careers.
And when scale economies in trade had been ignored from the 1950s to the early 1980s suddenly we reintroduced them to the analysis. Paul Krugman got a Nobel for it. But the way we did it actually vindicated Hicks's warnings. They could only be reintroduced in an ad hoc way that was short on clear policy implications.
I detailed this a little more in my debate with Krugman here.
I admire Krugman, but, like so many at the heights of the academy, he's got a real talent for making stuff that seems obvious disappear from his analysis. Then he brushes it off because no one else in the commanding heights of the academy saw it either.
This is a bit like an engineer designing a really nice bridge, without ensuring that they, or someone else makes sure it's paired with reliable geotechnical work to make sure it was appropriate for the site.
Probably the full argument here goes "Economists are the only ones who consistently promote free markets despite strong social incentives not to. They haven't built an *airplane*, really, but more of a sewer system—if the experts stopped maintaining it, you would only notice once everything went to shit."
Yeah my imagined skeptics here are the sorts of people who complain about exactly the sort of "market fundamentalism" Caplan is a fan of. E.g. "Cyrus Haghani" (who's almost certainly supposed to be Sohrami Ahmari) in that dialogue. And while I think Caplan takes seriously and makes a strong point against the Krugman/Stiglitz character, the Ahmari character is kinda just set aside. So it's question begging against my intended audience, even though I admit I'm sympathetic to the Caplan character in that dialogue.
The "applied auction theory" case studies are all pretty solid I think.
On the other hand, economists also designed the "deregulated" electricity market system (which really does centralized price setting to simulate market equilibrium). The record there is a little more mixed, ERCOT disaster especially bad. There are occasional screwups and people are starting to grumble.
I think trouble may be brewing for the "credibility revolution". For example a ton of studies use rainfall as an instrumental variable. This means they assume that rainfall directly influences the independent variable, and has no other path of influence on the dependent variable. But it's not possible for this to be correct for all these studies simultaneously, e.g. the independent variable in one may have a known effect on the dependent variable in another, etc.
I think the success of mechansim design is evidence that the mathematical modelling of economics can make important progress, and gives us evidence to think that modern economics can be applied in other fields where the effects are less clear are at least somewhat likely to be effective, and that, in the long run, we should think that mathematical modelling is fruitful path for understanding how the economy works.
I agree that mechanism design is much simpler than other problems that economists try to tackle, especially in macroeconomics, and on this basis, we should have, relative to micro theory, more confidence in heuristics and the intuitions of practitioners relative to academic economists, even if we think that academic economics has the correct methodology for developing scientific understanding in the long run.
I think the analogy with the physical sciences might be that one might trust the mechanic without formal training to fix your car, more than the academic physicists, even though academic physics is the way to understand what's going on in the car at a fundamental level (modulo debates about scientific realism).
On another note, my sense is that finance is an example of the engineering success of academic economics, and 2008 isn't very strong evidence against this. I've never worked in finance, but my understanding is that continuous time utility maximisation methods are the central method used in many cases, e.g with option pricing and portfolio management. I know that there are some practitioners who don't use these methods, but I think they're called "non-standard" for a reason.
I've read other comments talking about macro hedge funds using primarily verbal arguments based on causal models. It seems important to know whether these verbal arguments are using causal models developed by academic economists (or maybe more specifically, the mathematical tradition of economics) or are using models developed by some other method. In the former case, it seems similar to the engineer using ideas about heat developed by physicists without sitting down and solving the heat equation, which I think would amount to an argument in favour of the mathematical economics tradition. On the other hand, if they're using causal models derived from other methods, and particularly if those models contradict the qualitative implications of economists' models, then this would be an argument against modern economics.
Great story—perfect illustration of game theory in real life. Love how it went from coordination failure to enforced cooperation. Also really thoughtful take on the limits of economic modeling outside controlled settings like auctions. Makes you appreciate both the power and the boundaries of the economist’s toolkit.
This is extremely interesting. For the past year I was working at a hedge fund for where they were extremely skeptical of the economic consensus and the deliverances of academic economics. One major reason is that economics is much closer to politics than other theoretical disciplines, and economic views are nearly indistinguishable from political views. This makes ideology a much more serious concern than in physics or chemistry. Paul Krugman’s views on economics are inseparable from his political views—just look at his countless NYT editorials. So the finance community is skeptical.
But on the other hand, this points to another way of testing various economic theories. People working in finance can serve as a gauge of whether economists are in touch with reality, and perhaps a better one than the natural experiments you mention. Analysts and investment bankers who use particular economic models provide straightforward quantifiable empirical tests of economic models in the form of returns on assets.
Even so, though, I think what you’ll find is that most finance people tend to take the claims of academic economists with a fairly large grain of salt, because they just don’t think that academics have the right incentives to make sure their views are getting at reality. It’s the finance people who have a lot at stake, and half of the time they’re using their gut and heuristics and various other non-academic methods to make predictions.
Maybe it would be good to read Goldman’s client surveys instead of the IGMs!
I definitely agree traders have much stronger incentives to get things right, and so I'm happier deferring to market prices than I am deferring to economists. But I don't think it's easy to find cases where they'll conflict. E.g., I'm very happy looking at the differences between regular tnotes and tips and treating that as the best possible estimate of inflation over the next however many years. But I don't think you're ever going to have an IGM survey asking economists what they think inflation will be over the next 5 years, and *if* you did, my guess is that the economists wouldn't second guess the prices. I think part of that is by design; my sense is that the reason you're giving for why market prices should be trusted more than economists is something the people at IGM probably accept, so they're not going to ask questions where answers could be straightforwardly read off prices.
For the questions IGM asks about of the sort I discussed in the piece--stuff like rent control, or tariffs--I mostly don't know how to read answers off market prices. (Though there are maybe some exceptions. I do think that if high tariffs could be reasonably predicted to boost domestic manufacturing, then American manufacturing stocks should've gained in response to "liberation day", and that's certainly not what we saw. While they've gained back the ground they lost in the wake of the announcement, I mostly interpret that as reflecting the fact that it doesn't look like we'll ultimately get the high "reciprocal" tariffs that were announced.) But if I *could* read answers off market prices, I'd trust them more than the IGM surveys.
I'm actually thinking of doing a future post on how often it is that our opinions about the world (including about stuff like inflation) imply that various assets are mispriced, and the extent to which we should, if we accept the sensible, correct, normie advice that amateurs shouldn't try to beat the market, therefore bring our opinions in line with prices.
Yes, both points are well-taken, and I completely agree about Krugman. But since, as you say, this was my full-time job, I didn’t have to draw many inferences about the views of mainstream economists. I just build a repository of papers and models and collated their predictions. Of course, there’s no way to say whether this was truly representative, because many economists aren’t publishing on such things, but it gave me fairly decent evidence about those who were.
I think if I had no exposure to Econ or finance, and had to pick someone to defer to, I would more than likely defer to the street than to economists, if given the choice, just because of the alignment of incentives. But that’s not always easy, because as you say, they may address different questions.
Anyway, great topic, great post. Would love to talk to you more about it sometime (and I’m now back to being a philosopher, for better or worse)
I guess I don’t agree that it’s difficult to find places where they conflict. I worked a lot on tariffs this year, and many academic economists predicted huge increases in inflation from even the 10% across-the-board 60% China tariffs that were one of the original proposals floated. But looking at the EOY inflation predictions from the street, we saw no such predictions. A straightforward instance where traders’ projections for 2025 conflicted with the academic consensus (although I don’t know what the IGMs looked like). And so far the traders have been right—we had a very soft inflation print last month, although how things go remains to be seen.
Also, I am highly skeptical that the stock market is a reliable indicator of anything—including whether domestic manufacturing will infrease—since it is driven largely by retail traders who have no idea what tariffs will actually do, and are just responding to headline news about how bad tariffs will be. It’s just animal spirits, as they say. The bond market is a better indicator.
Anyway, I agree that people responding to the IGM would defer to market prices in general. But my main contention is that there are lots of places in which we don’t have prices to read off, and in which the street and various traders will diverge from the consensus of mainstream economists.
I guess absent stuff like IGM surveys, I'd be wary of drawing inferences about the consensus of mainstream economists, because I suspect the ones you hear from the most--like Krugman--are a highly unrepresentative bunch. Just for the general reason that media will select for simple, extreme, easy-to-categorize views. I imagine if Krugman had stuck to his persona from the 90s and before, he'd never have gotten to the level of public notoriety he has today, even if he'd have been just as (more?) respected within the discipline.
But I can easily imagine that if *I* were working at a hedge fund, I'd be less inclined to defer to economists, because I'd have both more time and resources to spend gathering and evaluating evidence that bears on the questions that they have views on (so my own opinion would be better informed) and I'd also have more direct access to the opinions of lots of other people like that (ie, my co-workers). It's one thing to say hedge fund traders have better info than economists, but another thing to say that people outside hedge funds, who don't really have any good way of finding out what hedge fund traders think, have a better option than deferring to economists in those cases where they know what economists tend to think (because they've been surveyed) but not what the hedge fund traders think.
To your point about the success of physical sciences in creating working technologies (like planes) and the lack of similar predictive power in economics- I think it’s interesting that macro hedge funds, whose existence depends on predicting, in a very precise way, the behavior of markets, don’t really use formal models! I interned at a macro fund, and their method of prediction seemed to be a combination of verbal argument about causal factors combined with data. There were formal elements to their trading strategy, but these did not at all resemble idealized economic models.
I'm a PhD statistician who worked with economists early in my career and have some familiarity with the field. What effect economists have overall is surely a complicated question. But at the macroeconomic level of deciding policy in a large country, the economists contribution is simply laughable scientifically, and disastrous socially. Inventing a neat mathematical system out of thin air is simply never a meaningful exercise when trying to examine a large economy. You're just a mathematical ideologist at that point, hiding your normative values behind inscrutable symbols.
And the history of economics has meant that until quite recently economists were mostly priests to the capital class, generating propaganda to embed intuitions about policy favorable only to oligarchs. And the credibility revolution has been great and all, but in terms of propagandizing the capital priests are mostly still winning. Mankiw's text still dominates 101 courses and that guy might as well be the definition of mathematical ideologist/capital priest.
The credibility revolution is much more a micro phenomenon than a macro one. It's not *impossible* to find natural experiments in macro, but it's much much harder, and requires what look to me like stronger, harder-to-verify assumptions in order to identify causation.
Yeah that's fair. I read Noah and don't dispute his point that the profession as practiced has changed a lot in the last 20 years.
My point is more about the effect of economists on public intuition and economic policy. It's been utterly dominated by fraud and misdirection. Fraud in the sense that any research that claims to scientifically demonstrate which macroeconomic policies are right in a general sense is usually mathy hocus pocus. And to the extent that economists have used their prestige as scientists to influence public policy there's basically Krugman and then everyone else arguing from the pro-rentier side. Misdirection in the sense that the simple concept of supply and demand has been twisted into "common sense" public intuition that just happens to benefit the rentier elite and harm everyone else.
When you use phrases like "priests to the capital class" youre kind of giving away the game that youre the ideologues, not necessarily "the economists"
No I'm giving away my years of frustration growing into the kind of admittedly immature, reactive dismissiveness shown in that phrase. But the frustration grows from the non-ideological fact that I have expertise in modeling complex random systems that macroeconomists don't display in their work, while at the same time they make broad, arrogant pronouncements about scientifically optimal economic policy which just happens to match their prior ideological commitments, plus mathy dressing.
I'm looking forward to the follow-up essay: "Should I trust philosophers?"
You should question them… yours truly Socrates 😎
This is an extremely fair and sophisticated reading of the lay of the land in economics, with a deeper grasp of strengths and limitations than most insiders! I would assign this to starting graduate students without reservation or qualification.
Thanks!
What I observed during covid (in Canada), was that politicians openly demurred to health authorities to determine government response, rather than the health authority providing options and trade-offs. Now, those experts certainly know health science better than me, the public, or the politicians, but they don’t necessarily hold the values that would result in the trade-offs appropriate to the public. I think something similar has occurred with economists dictating interests rates based on theories.
In biology/ecology, experts are never given authority to dictate large scale projects, their job is to define risks and impacts of different options. The advice tends to be very specific to the project. But what is often wanted is simple, broad scale maxims. However, as the issue broadens, generalisability decreases. I think that’s where economists are similar. They will be very good at predicting the effects of things like tariffs, but at broad macroeconomics their values are part of their models, and they have very few tools for large complex systems.
I think the distrust is driven because expert advice has been followed and many people feel worse off than before. I think in part that is because there are very broad frameworks that economists work within, and they don’t all work within the same frameworks (ex neoliberal, Keynesian, socialist, etc.). Therefore there are major differences in assumptions (ie how efficient are markets for public good). Hence, scaling between macro and micro economics has major implications in accuracy! But the public mostly just sees micro and macro as the same thing.
My observations and reflection.
Possibly this explains a growing distrust of economists as a consequence of a more general distrust of expertise, caused by (but maybe not justified by) cases such as these.
This is where it matters that Dan mentions the IGM surveys are usually on positive (descriptive) questions rather than normative ones! It seems to me that what we really wanted for covid policy was for the epidemiologists to tell us how much case counts would be affected if we did things like close non-essential businesses, or impose curfews, or mandate masks on the subway, while economists told us how much each of these rules would affect other things we care about, and then elected officials use these, together with their sense of public opinion, to decide the best trade-offs. The normative views should come from the democratic public, and the experts should only provide positive descriptions.
Now as a matter of fact, the very first survey I glanced at on the IGM page Dan linked was in fact explicitly normative, asking whether the risk of harm is such that most people would benefit from regulation of data privacy.
Yeah it's tricky. It's pretty clear they're always trying to get at the normative question of whether some policy is a good idea, but they *usually* ask a nearby descriptive question. E.g., "would increasing the federal minimum wage to $15 substantially increase unemployment?" But even then, there's wiggle room around "substantially", which people might be importing normative assumptions to interpret. In their most recent minimum wage survey they took it out, and the results were substantially (haha) different from than the last time when they'd asked a similar question that included "substantially".
So basically, while they *usually* try to ask descriptive questions, they also often make them vague or open to interpretation in ways where you could imagine people are drawing on normative views in deciding just how to answer. And of course sometimes they just go straight for normative questions about whether costs exceed benefits, as in the one you found.
I love this essay. It almost seems as though it was written by a good philosopher, instead of a good economist.
The autobiographical point at the end generalizes to a lot of things in life. "Yes, this thing here kinda sucks in some major ways, but Jesus Christ it's better than the alternatives". I trotted out that response many times years ago when philosophers were fighting over the Philosophical Gourmet Report.
In physics experiments effects are typically observable almost instantaneously. That makes it a lot easier to deal with unexpected external influences by modifying the experimental set up. In economics, effects can take months or even years before they are visible. The life of economists is too short.
Economics is pretty good at figuring out why things happened in prior times. And we know those who ignore history are doomed to repeat it. In this way, economics provides useful tools for policy-making. But one of the lessons of economics is that all sorts of significant events are functions of policy that failed to account for some factor or another given limits to what is knowable about the future. While evolutionary biology can explain much about evolution to date, it provides little insight into the next evolutionary adaptation. Similarly, economics is limited in its ability to design policy or predict the effects of a given policy in the future.
This is a great piece! Speaking personally, I think an additional reason I give more credence to the consensus in economics is that the discipline seems much less ideologically captured and much more open to following the evidence, changing their mind, looking for evidence against their views, etc. Economists seem to generally be the best social scientists, way better than the rest!
This is a good point. Indeed, I think the mathematization of the discipline is a reason why the discipline has avoided ideology capture.
If an ideologically-inclined economist is like, "capitalism is inequality" (or on the flip side "markets are intrinsically fair"), then the rest of the community is going to be like, that's a nice slogan, but what's the model? What do you mean by "capital"? What kind of relation are you positing between that and inequality? What metrics are you using? What are the timescales? What are the mechanisms? How do you model those?
The disciplinary effect of mathematics forces economists to get a lot more precise in their claims and precise claims are anathema to ideologues.
I think you're understating the case for engineering successes by economists. It's been pretty obvious if you look at the history of the 20th and 21st century that governments that largely adopt the policies that are recommended by mainstream economists (liberalization of markets with clear property rights, central bank independence, inflation management, low tariffs, manageable taxation, investing in innovation, etc.) fare much better at reducing poverty and growing wealth. Examples abound.
The Chinese growth following the Deng reforms. The post-1991 economic liberalization of India. The post-Cold War growth of the Baltics and Poland. Singapore.
Consider also the Chicago Boys reforms of Chile. Since those reforms, poverty dramatically fell in Chile and Chile has consistently outperformed the rest of Latin America. However, this episode is often brought up as a black mark against economists because they collaborated with a military dictatorship that overthrew a democratic government. Yes, that's certainly ethically dubious. But if economics is an effective science, then it's deliverances and effects should be independent of the ethical status of its users, much like in the case of physics or chemistry. And I think that's what's the case here.
And on the other side you have countries that didn't or don't follow mainstream econ and they, well, fare much worse. Take for example Venezuela, Zimbabwe, or Peronist Argentina. (And in the latter case, we are seeing that the econ-educated Milei is working to bring down inflation.)
I think this indicates significant success in social engineering, and has helped comparably many people as the engineering effects of physics.
P.S.--I found a paper (https://www.sciencedirect.com/science/article/abs/pii/S0304393219302028?via%3Dihub) that does a similar but more well-defined analysis, though I think my main point is plain enough to see.
While I tend to agree, I also think these cases tend to be too debatable to cut all that much ice when it comes to present debates. E.g., I've seen people appeal to the success of the Asian tigers in just the way you do here, but I've also seen fans of industrial policy say: "sure, they liberalized in various ways, but they were far from laissez faire, they protected infant industries, etc." So in the context of a present-day policy debate about some matter where economists largely agree, there's room for disputing the extent to which the earlier analogous policies that were followed by success amounted to following standard economic advice.
In your article you juxtapose economic analysis with "alternatives—relying on my own intuitions, partisan punditry, or critics armed with cherry-picked case studies". I know that that was the part of the article in which you'd given up straw-manning, but boy, did you give it up!
Part of the problem is paradigm wars. It's true that a lot of the critics of the current state of economics have some new paradigm they want to foist on us. I find that pretty laughable - like jumping out of the frying pan into the fire.
For me anyway, the problem is always our own ignorance and the quality of our critical awareness. The current practice in economics is to ignore anything that can't be translated into the state of the art of formal modelling. In fact the discipline did deal with this question - of where its methods added value and where they didn't. Modern economists just ignore it.
Thus as the new mathematised discipline was being forged - one of its architects - John Hicks warned that there were some things it might be good for and other things it wouldn't be good for. But the method became the "proof of work" within the economists' guild - and "proof of stake" come to think of it!
And so the method of formal modelling spread into areas where it had little but mystification to offer.
So bits of the simplest commonsense got consigned to invisibility at the commanding heights of the discipline. Commonsense like that scale economies are an important determinant of trade flows.
As Hicks had warned as early as 1939, as scale economies grow in importance, firms gain pricing power and the methods of economic theory encounter what Newton encountered with the three body problem. The analysis gets engulfed in complexity almost immediately.
Did economists face up to any of this? Well, they did for a while, but the disciplinary incentives eventually overwhelmed the thoughtfulness. Economists went on with their careers.
And when scale economies in trade had been ignored from the 1950s to the early 1980s suddenly we reintroduced them to the analysis. Paul Krugman got a Nobel for it. But the way we did it actually vindicated Hicks's warnings. They could only be reintroduced in an ad hoc way that was short on clear policy implications.
I detailed this a little more in my debate with Krugman here.
https://evonomics.com/paul-krugman-trade-theory-nobel-gruen/
I admire Krugman, but, like so many at the heights of the academy, he's got a real talent for making stuff that seems obvious disappear from his analysis. Then he brushes it off because no one else in the commanding heights of the academy saw it either.
As here https://clubtroppo.com.au/2019/04/02/paul-krugmans-incredible-invisibility-trick-2/
This is a bit like an engineer designing a really nice bridge, without ensuring that they, or someone else makes sure it's paired with reliable geotechnical work to make sure it was appropriate for the site.
I'd imagine you find it a bit too question-beggy, but thoughts on Bryan Caplan's "Markets Do the Good Things That Sound Bad" (https://www.betonit.ai/p/markets-do-the-good-things-that-sound)?
Probably the full argument here goes "Economists are the only ones who consistently promote free markets despite strong social incentives not to. They haven't built an *airplane*, really, but more of a sewer system—if the experts stopped maintaining it, you would only notice once everything went to shit."
Yeah my imagined skeptics here are the sorts of people who complain about exactly the sort of "market fundamentalism" Caplan is a fan of. E.g. "Cyrus Haghani" (who's almost certainly supposed to be Sohrami Ahmari) in that dialogue. And while I think Caplan takes seriously and makes a strong point against the Krugman/Stiglitz character, the Ahmari character is kinda just set aside. So it's question begging against my intended audience, even though I admit I'm sympathetic to the Caplan character in that dialogue.
Yes neoliberal economics is basically quite shitty 😆
The "applied auction theory" case studies are all pretty solid I think.
On the other hand, economists also designed the "deregulated" electricity market system (which really does centralized price setting to simulate market equilibrium). The record there is a little more mixed, ERCOT disaster especially bad. There are occasional screwups and people are starting to grumble.
I think trouble may be brewing for the "credibility revolution". For example a ton of studies use rainfall as an instrumental variable. This means they assume that rainfall directly influences the independent variable, and has no other path of influence on the dependent variable. But it's not possible for this to be correct for all these studies simultaneously, e.g. the independent variable in one may have a known effect on the dependent variable in another, etc.
I think the success of mechansim design is evidence that the mathematical modelling of economics can make important progress, and gives us evidence to think that modern economics can be applied in other fields where the effects are less clear are at least somewhat likely to be effective, and that, in the long run, we should think that mathematical modelling is fruitful path for understanding how the economy works.
I agree that mechanism design is much simpler than other problems that economists try to tackle, especially in macroeconomics, and on this basis, we should have, relative to micro theory, more confidence in heuristics and the intuitions of practitioners relative to academic economists, even if we think that academic economics has the correct methodology for developing scientific understanding in the long run.
I think the analogy with the physical sciences might be that one might trust the mechanic without formal training to fix your car, more than the academic physicists, even though academic physics is the way to understand what's going on in the car at a fundamental level (modulo debates about scientific realism).
On another note, my sense is that finance is an example of the engineering success of academic economics, and 2008 isn't very strong evidence against this. I've never worked in finance, but my understanding is that continuous time utility maximisation methods are the central method used in many cases, e.g with option pricing and portfolio management. I know that there are some practitioners who don't use these methods, but I think they're called "non-standard" for a reason.
I've read other comments talking about macro hedge funds using primarily verbal arguments based on causal models. It seems important to know whether these verbal arguments are using causal models developed by academic economists (or maybe more specifically, the mathematical tradition of economics) or are using models developed by some other method. In the former case, it seems similar to the engineer using ideas about heat developed by physicists without sitting down and solving the heat equation, which I think would amount to an argument in favour of the mathematical economics tradition. On the other hand, if they're using causal models derived from other methods, and particularly if those models contradict the qualitative implications of economists' models, then this would be an argument against modern economics.
Great story—perfect illustration of game theory in real life. Love how it went from coordination failure to enforced cooperation. Also really thoughtful take on the limits of economic modeling outside controlled settings like auctions. Makes you appreciate both the power and the boundaries of the economist’s toolkit.
This is extremely interesting. For the past year I was working at a hedge fund for where they were extremely skeptical of the economic consensus and the deliverances of academic economics. One major reason is that economics is much closer to politics than other theoretical disciplines, and economic views are nearly indistinguishable from political views. This makes ideology a much more serious concern than in physics or chemistry. Paul Krugman’s views on economics are inseparable from his political views—just look at his countless NYT editorials. So the finance community is skeptical.
But on the other hand, this points to another way of testing various economic theories. People working in finance can serve as a gauge of whether economists are in touch with reality, and perhaps a better one than the natural experiments you mention. Analysts and investment bankers who use particular economic models provide straightforward quantifiable empirical tests of economic models in the form of returns on assets.
Even so, though, I think what you’ll find is that most finance people tend to take the claims of academic economists with a fairly large grain of salt, because they just don’t think that academics have the right incentives to make sure their views are getting at reality. It’s the finance people who have a lot at stake, and half of the time they’re using their gut and heuristics and various other non-academic methods to make predictions.
Maybe it would be good to read Goldman’s client surveys instead of the IGMs!
I definitely agree traders have much stronger incentives to get things right, and so I'm happier deferring to market prices than I am deferring to economists. But I don't think it's easy to find cases where they'll conflict. E.g., I'm very happy looking at the differences between regular tnotes and tips and treating that as the best possible estimate of inflation over the next however many years. But I don't think you're ever going to have an IGM survey asking economists what they think inflation will be over the next 5 years, and *if* you did, my guess is that the economists wouldn't second guess the prices. I think part of that is by design; my sense is that the reason you're giving for why market prices should be trusted more than economists is something the people at IGM probably accept, so they're not going to ask questions where answers could be straightforwardly read off prices.
For the questions IGM asks about of the sort I discussed in the piece--stuff like rent control, or tariffs--I mostly don't know how to read answers off market prices. (Though there are maybe some exceptions. I do think that if high tariffs could be reasonably predicted to boost domestic manufacturing, then American manufacturing stocks should've gained in response to "liberation day", and that's certainly not what we saw. While they've gained back the ground they lost in the wake of the announcement, I mostly interpret that as reflecting the fact that it doesn't look like we'll ultimately get the high "reciprocal" tariffs that were announced.) But if I *could* read answers off market prices, I'd trust them more than the IGM surveys.
I'm actually thinking of doing a future post on how often it is that our opinions about the world (including about stuff like inflation) imply that various assets are mispriced, and the extent to which we should, if we accept the sensible, correct, normie advice that amateurs shouldn't try to beat the market, therefore bring our opinions in line with prices.
Yes, both points are well-taken, and I completely agree about Krugman. But since, as you say, this was my full-time job, I didn’t have to draw many inferences about the views of mainstream economists. I just build a repository of papers and models and collated their predictions. Of course, there’s no way to say whether this was truly representative, because many economists aren’t publishing on such things, but it gave me fairly decent evidence about those who were.
I think if I had no exposure to Econ or finance, and had to pick someone to defer to, I would more than likely defer to the street than to economists, if given the choice, just because of the alignment of incentives. But that’s not always easy, because as you say, they may address different questions.
Anyway, great topic, great post. Would love to talk to you more about it sometime (and I’m now back to being a philosopher, for better or worse)
I guess I don’t agree that it’s difficult to find places where they conflict. I worked a lot on tariffs this year, and many academic economists predicted huge increases in inflation from even the 10% across-the-board 60% China tariffs that were one of the original proposals floated. But looking at the EOY inflation predictions from the street, we saw no such predictions. A straightforward instance where traders’ projections for 2025 conflicted with the academic consensus (although I don’t know what the IGMs looked like). And so far the traders have been right—we had a very soft inflation print last month, although how things go remains to be seen.
Also, I am highly skeptical that the stock market is a reliable indicator of anything—including whether domestic manufacturing will infrease—since it is driven largely by retail traders who have no idea what tariffs will actually do, and are just responding to headline news about how bad tariffs will be. It’s just animal spirits, as they say. The bond market is a better indicator.
Anyway, I agree that people responding to the IGM would defer to market prices in general. But my main contention is that there are lots of places in which we don’t have prices to read off, and in which the street and various traders will diverge from the consensus of mainstream economists.
I guess absent stuff like IGM surveys, I'd be wary of drawing inferences about the consensus of mainstream economists, because I suspect the ones you hear from the most--like Krugman--are a highly unrepresentative bunch. Just for the general reason that media will select for simple, extreme, easy-to-categorize views. I imagine if Krugman had stuck to his persona from the 90s and before, he'd never have gotten to the level of public notoriety he has today, even if he'd have been just as (more?) respected within the discipline.
But I can easily imagine that if *I* were working at a hedge fund, I'd be less inclined to defer to economists, because I'd have both more time and resources to spend gathering and evaluating evidence that bears on the questions that they have views on (so my own opinion would be better informed) and I'd also have more direct access to the opinions of lots of other people like that (ie, my co-workers). It's one thing to say hedge fund traders have better info than economists, but another thing to say that people outside hedge funds, who don't really have any good way of finding out what hedge fund traders think, have a better option than deferring to economists in those cases where they know what economists tend to think (because they've been surveyed) but not what the hedge fund traders think.
To your point about the success of physical sciences in creating working technologies (like planes) and the lack of similar predictive power in economics- I think it’s interesting that macro hedge funds, whose existence depends on predicting, in a very precise way, the behavior of markets, don’t really use formal models! I interned at a macro fund, and their method of prediction seemed to be a combination of verbal argument about causal factors combined with data. There were formal elements to their trading strategy, but these did not at all resemble idealized economic models.
I'm a PhD statistician who worked with economists early in my career and have some familiarity with the field. What effect economists have overall is surely a complicated question. But at the macroeconomic level of deciding policy in a large country, the economists contribution is simply laughable scientifically, and disastrous socially. Inventing a neat mathematical system out of thin air is simply never a meaningful exercise when trying to examine a large economy. You're just a mathematical ideologist at that point, hiding your normative values behind inscrutable symbols.
And the history of economics has meant that until quite recently economists were mostly priests to the capital class, generating propaganda to embed intuitions about policy favorable only to oligarchs. And the credibility revolution has been great and all, but in terms of propagandizing the capital priests are mostly still winning. Mankiw's text still dominates 101 courses and that guy might as well be the definition of mathematical ideologist/capital priest.
Fwiw, my sense is that plenty of economists themselves are pretty quick to admit that macro is just much worse understood than micro. E.g.: https://www.noahpinion.blog/p/macroeconomics-is-still-in-its-infancy
The credibility revolution is much more a micro phenomenon than a macro one. It's not *impossible* to find natural experiments in macro, but it's much much harder, and requires what look to me like stronger, harder-to-verify assumptions in order to identify causation.
Yeah that's fair. I read Noah and don't dispute his point that the profession as practiced has changed a lot in the last 20 years.
My point is more about the effect of economists on public intuition and economic policy. It's been utterly dominated by fraud and misdirection. Fraud in the sense that any research that claims to scientifically demonstrate which macroeconomic policies are right in a general sense is usually mathy hocus pocus. And to the extent that economists have used their prestige as scientists to influence public policy there's basically Krugman and then everyone else arguing from the pro-rentier side. Misdirection in the sense that the simple concept of supply and demand has been twisted into "common sense" public intuition that just happens to benefit the rentier elite and harm everyone else.
When you use phrases like "priests to the capital class" youre kind of giving away the game that youre the ideologues, not necessarily "the economists"
No I'm giving away my years of frustration growing into the kind of admittedly immature, reactive dismissiveness shown in that phrase. But the frustration grows from the non-ideological fact that I have expertise in modeling complex random systems that macroeconomists don't display in their work, while at the same time they make broad, arrogant pronouncements about scientifically optimal economic policy which just happens to match their prior ideological commitments, plus mathy dressing.
I guess that you aren't a fan of using the Laffer Curve to determine macro policy for taxation?