In 1849, the historian and philosopher Thomas Carlyle referred to economics as the “dismal science.” The pejorative stuck, and is still slung by critics of the field today. 

But what if economics is worse than “dismal”? What it’s…harmful? 

George DeMartino’s recent book, “The Tragic Science: How Economists Cause Harm (Even as They Aspire to Do Good)”, makes exactly that claim: that economists aren’t just ineffective at solving social problems; they often end up creating new ones. Worse still – since economics lacks a meaningful criteria for defining what harm is, economists often don’t know how to measure (and fix) the problems they create. 

George is an economist himself, and his work isn’t just a pile-on against the field. Rather, his critique points a way towards a more socially engaged version of economics – one that takes the notion of harm seriously. 

Learn more about and purchase “The Tragic Science How Economists Cause Harm (Even as They Aspire to Do Good)”

Learn more about the Watson Institute’s other podcasts

MARK BLYTH: From the Rhodes Center for International Finance and Economics at Brown University, this is the Rhodes Center Podcast. I'm the Director of the Rhodes Center and your host Mark Blyth. In Eighteen Forty-Nine, the historian and philosopher Thomas Carlyle referred to economics as the dismal science. The pejorative stuck and is still slung around by critics of the field today, but dismal is at least better than harmful. So what if it's really the harmful science? George DeMartino's recent new book, The Tragic Science, makes exactly that claim that economists aren't just ineffective at solving social problems. They often cause new problems when they try to solve old ones. Moreover, the field is hamstrung by its ability to not do harm since it lacks any criteria of what harm is and how to avoid it.

George is an economist himself, and his work isn't a simple pile-on against the field. Rather, his critique points a way towards a more socially engaged version of economics that takes the notion of harm seriously. We started with why economists seem to be so very sensitive to criticism. Hope you enjoy it.

[MUSIC PLAYING]

Hello, George. Good to see you.

GEORGE DEMARTINO: Great to see you, Mark.

MARK BLYTH: It's great to see you because, once again, you're doing what you do best, which is trouble the conscience of your profession. You've done so with your new book, The Tragic Science, how economists cause harm even as they aspire to do good. Now, I want to get into the main themes of the book shortly, but I do want to just suggest something to you and get your reaction. Earlier on this semester, I had a conversation with Jacob Saul. He's an intellectual historian. And he wrote a book called The Free Market; The History of an Idea. And being an intellectual historian, he didn't start with Smith and then basically do the pantheon of 19th-century greats which builds up through Hayek de Friedman and then freedom for the rest of the world.

What he did was went back to the Romans, went back to Cicero, and walked us through a much longer story that essentially says, no, all of these stories about self-regulation and self-regulating things, whether the markets or not are really power plays by elites to basically preserve their powershell, we say. And when his book was surveyed by illustrious members of your profession, they didn't like it. They really didn't like it at all.

GEORGE DEMARTINO: I can imagine.

MARK BLYTH: The effect on the fact, they had a misplaced footnote, but really, the critique was deeper than that. You've gone one further. You're not just saying that they don't really understand their own history. You're saying that they don't understand that they cause harm. So let's just-- before we really get into harm, why so sensitive? What's going on? What is it that when you push these boundaries, you create such a reaction?

GEORGE DEMARTINO: I think that the economics profession, more so than the other social science professions, is quite defensive about critiques coming from within the profession and coming without-- coming from without. And I think a lot of it has to do with the fact that, well, since the founding of the modern profession, let's say, toward the end of the 19th century, this profession has sought influence, influence, influence over public policy affairs. There's a deep-seated belief in economics that a world in which economists have more influence is to be preferred as a better world than a world in which we have less influence.

And so there's been an extraordinary amount of energy put into expanding economist's influence. Critiques of the sort you just described and the critique of the sort I have just written are problematic from this perspective because they start to call into question the authority of economists. The idea that economists should have that much authority and influence in the world. I think we've overstepped ourselves. We've gotten ahead of ourselves. We have more influence than we are prepared to bear, and that's an argument that the profession really does not want to hear.

MARK BLYTH: So you maintain that this is the tragic science rather than any of the usual epithets which was put on there, the dismal science or whatever that the profession itself causes harm. Let's get into this. What kinds of harm? You open the book with the discussion from Steckler and Basu's book Twenty Thirteen about Russia. And then another example of what happened to Midwestern communities in the United States because of trade with China.

GEORGE DEMARTINO: Sure.

MARK BLYTH: Tell us why these things aren't just economic events. Why do you think they're classes of harm?

GEORGE DEMARTINO: I think every economist has been trained for the last 100 years to know that when you embrace a policy and when an economic policy is implemented, even if it's one, you have reason to support. There are going to be those who benefit from the policy, but there are very likely to be those who suffer harm from the policy. That's not the problem. The profession does recognize that its interventions, even its preferred interventions like free trade, cause harm to some people. The problem is in how we understand the nature of the harm.

Economists theorize harm and simply in economic terms. So, for example, if the United States embraces a free trade agreement, and you lose your job in a North Carolina textile mill, well, we theorize you as having suffered a loss of income because you've lost your job. Full stop. But, of course, that's not the end of the story for a community that's been ravaged by unemployment because of, let's say, trade liberalization.

Those economic harms cascade out to reach into other kinds of harm. I mean, the evidence by now is clear communities hit with unemployment, high levels of unemployment face physical harms, psychological harms, political harms, losing political efficacy, social exclusion, increases in mortality rates, increases in homicide, increases in homelessness. The initial economic harm radiates out across the full range of human harms and, therefore, can be far more devastating than we teach our students when we teach them international trade.

MARK BLYTH: So that's a compensation problem, and I want to get into that the fact that you can theorize compensation, but you don't actually do it. I'll play defense on this one. But that's just bad governance. I mean, ultimately, if the politicians were up to their job, they would design compensation mechanisms, and these board harms would be alleviated. Would only get the good stuff.

GEORGE DEMARTINO: No. I would put it this way. In the first instance, it's a failure of professional responsibility in economics. Economists have treated harm as if it's something that's always limited that it dissipates very quickly. So, for example, a community harmed in the way I just described a North Carolina textile community that's now losing jobs because of free trade. What you were taught, what I was taught, is a story like this. There's an initial shock to that community, but it's like a pebble going into a pond.

Initially, the ripples are very big, but they very quickly dissipate. That's how we're taught to think about harm. Yeah, there might be some short-term harms, but very quickly, those harms dissipate, and so we can think about the economy moving from one equilibrium to another. But what if some of those harms aren't like a pebble going into a pond? What if, instead, they're like an avalanche? Now, an avalanche is snow moving down the mountain. It starts very slow and small, but as it goes through time, it gains force and power and becomes more destructive.

What if some of the harms to that community are of that sort such that they don't just pass away or dissipate like ripples on a pond, they actually radiate out and deepen over time? Well, here's the bad news, Mark. More and more research. Empirical research is starting to show that that's exactly what happens. That the harms of the community suffers by virtue of, say, something like trade liberalization actually deepen over time. They don't self-correct over time.

The obligation of the profession is to be talking about that. To be warning policymakers. Those policymakers have been trained by people like me to think of these harms as ripples on a pond; hence no need for big government interventions. The point being this we have misled our public about the nature of the harms, the depth of the harms, and how long these harms endure.

MARK BLYTH: So, as you know, little while ago, 10 years ago, I wrote a book about austerity, which struck me as I used to call it self harm. That was just my only way of understanding why any society would want to do this when you know it's not going to work. And then you had the really extreme version of this, the expansionary austerity. Not only will it dissipate, it's actually good for you. That was actually the claim that was being made.

GEORGE DEMARTINO: Yeah.

MARK BLYTH: And to me, that was always astonishingly irresponsible and reprehensible and not way because there was no evidence for it, and eventually the IMF and practically everyone except the European Commission finally came and went, yeah, that was a bad idea. Let's not do that. Does learning happen, or do we just get so stuck in our particular ways of thinking that there no way to jolt us out of, if you will, the do-harm equilibrium?

GEORGE DEMARTINO: Learning is a very difficult problem in my profession. Let's compare with medicine. Let's say I'm a clinical doctor in a clinical setting, and you're my patient, and you come in, and I misdiagnose you, and I mistreat you, and you suffer harm. The connection between my intervention and your outcome is rather direct. Typically, it happens almost immediately. And so there's a feedback loop that I get information right away that my intervention was wrong, and I, therefore, have a good chance of learning.

In economics, the impact of policies that we're talking about, let's say, trade liberalization. The impacts are indirect. They're never direct. They are filtered over many years. And at the same time that policy is generating its effects, there's a million other things going on in the world, such that, at the end of the day, an economist who advocated a policy in the year Twenty Twenty and now sees that bad things have happened by Twenty Twenty-Five, it's very difficult for that economist to accept the policy that he or she advocated was causal of the harm.

It's much easier to say, well, the policy wasn't implemented the way I indicated it should be implemented. This was Jeffrey Sachs and Russia. That they didn't do everything I said they should have done, and of course, therefore, the policy of course failed, but it wasn't what I was advocating. We have hindsight bias. We have all these mechanisms by which, in economics, we can insulate ourselves from the harm associated with the policy. We advocated, now, this is irresponsible, but it's even worse. It prevents collective professional learning.

MARK BLYTH: There's a fabulous line in Karl Polanyi's, The Great Transformation, where something along the lines of for every single failure of liberalism, its advocates cling to the belief that if only we'd applied it more regularly and more thoroughly, the outcome would have been different.

GEORGE DEMARTINO: That's exactly right. I mean, you can always point out that no policy is ever implemented exactly the way the architect of the policy had intended. All that tells me is that the policies aren't robust. If you've got policies that require that everything be done perfectly in order for the policy to work, you've got your hands around a very, very dangerous bad policy.

MARK BLYTH: Let's get into some of the weeds on this. A critical factor in what you're exploring. You have to basically unpack a set of moral and causal architectures in order to make the book work. I actually found it to be more in-depth than I thought and also more far-reaching. A couple of things that stood out. The first one is a very [INAUDIBLE] term which you've invented, which is irreparable ignorance, and the other one is econogenic thinking. These play very important anchor roles in the book. Can you unpack both of those for us?

GEORGE DEMARTINO: Sure. Let's start with the last. In some professions, not all, there is tremendous attention to the fact that well-meaning, well-trained practitioners are going to nevertheless cause harm as a part of their professional practice, that this problem, to some degree, is irresolvable. Medicine comes to mind in which the term iatrogenic harm, physician-induced harm from the Greek, has been a central concept for, well, half a century, at least. Much of medical ethics and much of medical practice focuses on this problem of iatrogenic physician-induced harm and looks for ways to minimize it. Ameliorate it, and so forth.

In economics, we have never had this conversation about economist-induced harm. We don't even have a term to describe it, and so I've borrowed from iatrogenic harm, and I've introduced the concept econogenic harm. Economist-induced harm. The harms that come about by virtue of what economists write, what they teach, and what they do outside of the academy. With respect to irreparable ignorance part of the argument of the book and is the central argument to the book, and I should tell you won't be surprised. It's the argument in the book for which I get the most pushback from economic audiences.

The argument is that economists do not know and cannot ever know much of what they need to know to do much of what it is we're doing. So, for example, long-term forecasting, cost-benefit analysis. All the kinds of policy analysis tools. We can't ever know enough to actually do the kinds of work that we are doing. I call it irreparable ignorance because it's never going to be overcome. It's not a problem that can be overcome with more knowledge.

One of the arguments in the book is when we generate more knowledge. We also generate new domains of salient ignorance. When we learn something, there are new things we need to know that we cannot know at the time that we've created the new knowledge. Current example artificial intelligence AI. Two weeks ago, Elon Musk and another thousand, I think, researchers and executives in the field of AI was called for a moratorium on continued development of these tools.

Their argument was, in my view, perfectly correct. They argued when we generate this new knowledge, we're generating whole new realms of ignorance about things we would now need to know about how these technologies are going to be used, what will be their impacts, and so forth. We need to know all these things because to pursue the technology at this rate is going to be far too dangerous. Going to create far too much harm. I argue that that's a universal condition. New knowledge creates new domains of ignorance.

And so, we can't just think that science will progress and eventually will banish ignorance. Ignorance is with us forever. The question is, how do we manage it, not how do we banish it.

MARK BLYTH: So let me take this from my bits of econ that I know. I can absolutely see that if you're basically stuck in late 80s, early 90s macro, and it's a toy model, but you confuse the toy model for the real world, and therefore, you make very buzzy policy recommendations based upon a toy model which has no grounding in reality and that causes a great deal of harm and disruption.

But these days, when I look at economists who are younger than me, they're much more micro-focused. They're much more careful about what they do. They're very much into trying to get causality done with RCTs and all that stuff. Is that better, or does that suffer from the same types of ignorance, shall we say?

GEORGE DEMARTINO: Yeah, the problem can't be resolved. The problem is the nature of causality and causal claims. I mean, causal claims in the social sciences almost always involve counterfactual claims. We're comparing what did happen against another world in which the causal event did not happen to claim that, in fact, the event was causal.

MARK BLYTH: But I'll put on my micro hat and go, yeah, and we have a control group. We have a treatment group, and everything else is constant. So what's the problem?

GEORGE DEMARTINO: Yeah. So let's talk about economic methods more generally going back, oh, I don't know 50 years. For 50 years or more, the profession has been looking for a method that will deliver unimpeachable claims about causality. That's what econometrics is-- in fact, that's what every method that you and I could do if we were to make a list of the predominant methods that have achieved standing in economics over the last 50 years. That's what they're all trying to do.

What they're all trying to do, I would argue, is not to describe what they see out there in the world. The big task of these methods is to describe or to identify the uniquely correct counterfactual because if you can't identify the uniquely correct counterfactual, you cannot sustain an unimpeachable causal claim about the world that we do inhabit. Now, the problem, of course, is that there's never going to be an unimpeachable test about what would have happened in this alternative world that only exists in my imagination had some causal event x not occurred.

RCTs, randomized controlled trials, are now the new shiny method that's supposed to deliver to us unimpeachable causal claims, but it's grounded in counterfactual thinking all the way down. For example, had the control group not been the control group but instead been given the treatment, its results would have been identical to the group that was given the treatment, and vise versa. Had the treatment group not been given the treatment, its effect would have been the same as what happened in the control group.

Of course, can we ever know this? We can't ever know this. And the reason we can't ever know this is because of what's called the fundamental problem of causal explanation. In order to really know if some x caused some other event y, we'd have to be able to run history twice. Once where x happened and one identical world in which x does not happen. And then, to use the evidence from your RCT, you've got to be prepared to hold in your head that the world out there isn't changing in some fundamental way such that you can then export the finding to some other community on some other continent, and you'll get the same results.

Often not always, I find these claims undependable and unpersuasive.

MARK BLYTH: What I've always found puzzling about this the reason we do comparisons is because things are different over there, and, in a way, the logic of the RCT is to say no, everything is the same everywhere. And given that we know that things really aren't the same over there and lots of other over there is a bit of a stretch to just make that claim to universalism.

GEORGE DEMARTINO: Yes.

MARK BLYTH: So that's one of the things that you pick up on. So there's ignorance. There's universalism, et cetera. You also talk about responsibility without control and paternalism. How do they play into your story?

GEORGE DEMARTINO: Brief correction. Influence without control.

MARK BLYTH: Oh, sorry. Influence without control, yes.

GEORGE DEMARTINO: Yeah, that's such a fundamental point that was actually driven home to me by one of my colleagues, a political scientist. I was talking about the power that economists have. And she said to me economist don't have power, maybe they have influence, and I thought, of course, that's right. Economists don't have control. They cannot control anything in the world, but what they do have is an immense amount of influence in the world.

They can nudge decision-makers to make fundamentally different decisions than they might have otherwise. But beyond that, they cannot control what happens next. The image I give in the book is of little Johnny, 10 years old driving down Main Street in a humvee. Does little Johnny have influence over what's about to happen next? Oh, yeah. But does little Johnny control what happens next? The answer is no. Do not bet your life that little Johnny can control what happens next. That's the position of the influential economist. That's the Ben Bernanke problem.

The central bankers have all kinds of influence, but they cannot control what happens next. When you have influence but you don't have control, you are in position to do tremendous damage. There are going to be all kinds of unintended consequences because you lack control. And as a consequence of that, you have a responsibility to those purport to serve to share with them the limits to your knowledge and the limits to your control so that they themselves can be more centrally involved in what risks to take in pursuit of what social objectives.

We've taken it upon ourselves to make these decisions about what harms we should feel OK about imposing for what benefits rather than engaging those who stand to be harmed to engage them in a more meaningful way in decisions about what risks they're prepared to take in order to achieve what social goods. After all, we purport to serve them. Well, if we're serving them, maybe they should be more centrally involved in the decisions that are going to bear directly on their lives.

MARK BLYTH: You could think that way, or you could think that basically, it's probably bad if we just nudge them along.

GEORGE DEMARTINO: Yep. That's where my profession has headed over the last 15 years or so, coming out of behavioral economics. The idea is that people aren't rational in the way that neoclassical theory has presumed for 100 years. OK. Fine. But the project has been to discover if people are-- and here's the key point-- predictably irrational because if their behaviors though irrational, are predictable, then we can use that knowledge of their irrational patterns to get them to do what we know is in their best interest.

And this is the case of nudge. This is the case of people like Cass Sunstein. The idea that we know-- as utilitarians, we know what's in people's best interest. Now we have the knowledge from experiments to nudge them to make decisions that we know are best. That is not engaging people as autonomous agents with integrity. That's treating them as animals to be herded around to get them to do what we already decided is in their best interest.

MARK BLYTH: And, yes, it has that deep problematic core of essentially erasing any notion of common futures or citizenship and replacing us with a class of people who are very clever who know what we should all be doing, and we get to nudge everybody else into those behaviors. I just can't help but think that a little bit of the current upset and upswing in populism and the rejection of elites is somehow tied up with that.

GEORGE DEMARTINO: I couldn't agree more. And in Twenty Sixteen, I was in charge of the economics meetings setting the agenda for one of the economics associations, and the theme that I chose was the illiberal turn. The illiberal turn and the responsibility of economics. And I got people like Daniel Rodger to participate and Jamie Galbraith, a number of wonderful economists to present on this topic. And in fact, there was a widespread feeling among those who chose to participate in these sessions widespread feeling that we had violated our obligations to society.

We had been dishonest to them. We oversold them on what it is we know. We oversold them on the dream of free trade and globalization. We sold them on this myth that everyone would get rich from the kinds of economic liberalism that my profession was selling in the nineteen-eighties, nineteen-nineties, and since. And instead, a large swath of people the world over got badly hurt. Well, they are not interested anymore in what experts like economists have to say, and moreover, not only are they tired of economists, they've turned against professions more generally.

Like we saw this with public health during COVID. There's a large swath of people around the world who are really angry at experts and who are then vulnerable or available to the Donald Trumps, who rail against the idea of expertise altogether. It is just a frightening moment, and I am with you in believing that my profession contributed to this skepticism and even cynicism around expertise.

MARK BLYTH: Well, essentially, by departing from society or at large of becoming, if you will, a priesthood, economists are not alone in this. There are many other strata of society. One can think of finance. One can think of other professions that have similarly pseudostratified themselves away. But it does seem to be part of the same pathology.

Let's plow further into the book. You have this very interesting term, moral geometry. What is it, and why does it matter?

GEORGE DEMARTINO: Moral geometry is a strategy that my profession uses to adjudicate policy. The idea is this that policy often harms some groups of people while benefiting others. We talked about that earlier, and economists know this. Now, this raises really difficult moral problems. When is it OK to harm some for the benefit of others? In what ways is it OK to harm people? For what kinds of benefits, and on and on and on?

These are daunting moral problems. But economists don't want to get bogged down in these daunting moral problems. What they want to do is to be able to quickly, effectively, efficiently pass judgment on policy. And so what we are trained to do in economics is to transform these really daunting moral problems into simple little math problems like cost-benefit analysis. And then we are trained to solve the math problem and think when we've solved the math problem that we've actually had something useful to say about the deeper moral problem.

I call this moral geometry. Looking for simple solutions to really daunting moral questions. It is professionally irresponsible. I think it's professional malpractice to use this moral geometry when people's livelihoods and even people's lives are at stake. By moral geometry, the austerity that you write about so effectively in your own work. It's easy to do a cost-benefit analysis to show that Greek austerity, actually the benefits are greater than the costs. Does that mean it's OK to see children's more-- child mortality rates going up? Public health crises of all, but it's easy to do moral geometry to show that yes, that's A-OK because, in the long run, the gains to the winners are going to be greater than the losses to the losers.

Who else gets away with pronouncing like this other than my profession? I find it abhorrent.

MARK BLYTH: Let's go back to the austerity example and work this one through. If it's Twenty Twelve, or something like that. Greece is really kicking off this contagion risk in the markets. They're adopting the pigs and all that stuff. They're basically doing a rerun of the East Asian financial crisis. You liquidate one pot to cover your losses and another part, and that creates the contagion, and that's what everyone got freaked out about, and that's why we quote-unquote had to do austerity.

Is there any room for a type of logic that says, OK, we did horrible interventions? We know we're playing moral geometry, but if we don't given what we don't know, it's just as likely that things will get even worse, so we had to intervene. Is there any room for that type of logic?

GEORGE DEMARTINO: Yeah, I mean, the idea that we face tragic choices in which all the decisions are bad. I get really anxious when people embrace the tragic choice argument without doing a really thorough demonstration drawing on lots of different theoretical traditions to show us that there really is no alternative to the one that's being advocated. Too often, these judgments about China there is no alternative. Comes from one theoretical paradigm that sees the world in one particular way in which there is no escape but to pursue a policy that we know is going to be damaging to many, many people.

Now, one of the problems in economics is that economics is allergic to pluralism. The idea that there can be-- Mark might have something to say. Adam Smith might have something to say. [INAUDIBLE] And so forth, and that we might actually confront a problem like the fiscal crisis with the possibility of contagion. We might confront these from lots of different theoretical lenses to see whether any of them shine light on avenues that we might pursue that don't involve such damage to so many people.

That is never done in economics. What's done is the high priest of economics draws on whatever model typically neoclassical model she or he takes to be true and dictates this is a tragic choice too bad some people are going to suffer but we have no alternative. The fact of the matter is there are innumerable policy strategies always available. And how it can be said that in any particular crisis, there's only one available and moreover, Mark, that we know what it is, I find this to be the height of arrogance and professional irresponsibility.

MARK BLYTH: One of the other things-- we'll have to wind it down a little, but I do want to get into this. This is something that many trade economists such as David [INAUDIBLE] and everything will freely admit now, and particularly the younger generation are following those folks, though we just lied about compensation. Compensation was like a theoretical possibility that we said oh, that'll happen. And then, of course, it didn't happen. And that's when you get really, really antigenic harm coming that way.

Are we smarter about compensation now when we know that we have to do things that involve trade-offs when there are difficult choices? Are we more sincere about compensation now, or are we just as blase as we always were?

GEORGE DEMARTINO: I guess it depends on who you mean by the we. The economists you mentioned are certainly-- there are some economists I agree with. There are other economists I trust. So the two you just mentioned economists I don't always agree with, but I find that I trust. And because I do think that they're raising some of the critical questions about the uneven impact of what we do and the ways in which we need to be much more responsible and sensitive to the harms that our activity generates.

I agree with something you said earlier, which is that we're going through a generational shift in economics, which is extraordinary. We're moving away from the old theoretical battles of the 20th century toward empiricism. I mean, in some ways, young political scientists and young economists strike me as the same.

MARK BLYTH: Very similar, absolutely.

GEORGE DEMARTINO: Theories being left beside-- to the side. And it's just doing high-quality empirical work on pressing problems. That's a whole different world than the one that I was trained in and that I've populated over the last 30 years as an economist. And so it's a real open question for me how that generation is going to come down on these issues, but they are not ideologues. That we're-- to say it another way, if they're ideologues, their ideology is empiricism. It's not neoliberalism.

And so I think there's going to be-- I think there probably will be much greater openness to the idea that the state has more very important roles to play when we also are pursuing things like trade agreements then that then has been the case with my generation of economists, but we'll have to wait and see.

MARK BLYTH: Yeah, the trade agreements. One is always a puzzle for me because even the most mainstream people would say oh, yeah, you're going to get half a percentage point of GDP over 10 years. It's not really worth it. Well, why are you putting so much effort into it then? Well, 'cause of the consultancy gigs that we get with all the firms that are going to get the investor protection courts.

So yes, there's always a duplicitous game going on there. But I'm going to bring it to a bit of a close there. I want to turn to two things. You mentioned cost-benefit analysis that you alluded your critique, but you have a much more fundamental critique of why this seemingly neutral standard really is not neutral, and I want you to flesh that out and then to give us the alternative, which is decision-making under false uncertainty.

So let's talk all those one at a time. Cost-benefit analysis seems reasonable. Why do you think it's so not reasonable?

GEORGE DEMARTINO: Cost-benefit analysis is the way in which the profession enacts what's called the Kaldor-Hicks compensation test. Kaldor-Hicks is an approach to adjudicating policy where there are winners and losers. It's a very simple decision rule. It's a perfect example of the kind of moral geometry that I criticize in the book.

The Kaldor-Hicks says that if that a policy is good if the gains to the winners are greater than the losses to the losers. Full stop. The idea is that if the gains to the winners are greater than the losses to the losers, then in principle, the winners could fully compensate the losers for their losses and still have something left over. We call this a potential pareto test because if the winners really were to compensate the losers, then we'd meet the pareto criterion, which is to say there'd be only winners and no losers' potential, but no one requires that the actual compensation take place.

Cost-benefit analysis is the job, of course. Benefit analysis is then to do the math to find out if the gains to the winners are greater than the losses to the losers is how we enact a Kaldor-Hicks. The long-standing critique of this is that the winners are not expected to compensate the losers. But I think that-- I share that criticism. But there's a deeper criticism, and it's this. A lot of the harms that people suffer in the economy by virtue of what economists do are not even in principle compensable. Some of the harms are irreparable.

If you are a great father in Two-Thousand Nine, and you can no longer afford a pharmaceutical to keep your daughter alive because of austerity policy, you've lost your daughter. Is there really some amount of compensation you're prepared to accept from the winners such that you're just as well off with the money and without the daughter, as you were with the daughter and without the money? If you were to tell me that there is such amount of money, I'd say you are unfit for parenthood.

The point being there are some things that are not irreparable, and because they're not irreparable, they're not compensable, and therefore, economists have no business using these tests when people are suffering harms that, in fact, are irreparable and non compensable.

MARK BLYTH: Now, tell us what else to do then. There's another approach that gets away from all the stuff we've been using. You spend quite a bit of time at the end of the book talking about it. It's about the fact that the world is not probabilistically certain is deeply unknowable and uncertain, and not Knight Keynes formulation I've always been fond of. So this takes advantage of that rather than just seeing it as something to chuck into the error bucket and pretend isn't there.

GEORGE DEMARTINO: Exactly.

MARK BLYTH: How does it work? The last chapter of the book is on this relatively new approach to policy-making by practitioners who know we cannot know the causal connections out there in the world, and therefore, we can't make predictions. As you say, the world is uncertain, not just risky. It goes by the name decision-making under deep uncertainty, and I don't offer this as a panacea. I offered as one approach that's got its finger on how we might support good decision-making in the world, even once we understand that we can't the future or control the future.

The idea is to bring together stakeholders, those who are going to be most directly affected by a policy, and rather than dictating to them what they should want, getting them together to search for policy possibilities, then to stress test the policies that come from the stakeholders against 10, 20, 30,000 possible futures without assigning probability distribution to those futures to see which of these policies that are being suggested by the stakeholders seem to have a chance of doing well enough, not optimally well enough against a very large number of futures.

The process is iterative. You engage stakeholders in the problem and look for policy solutions. You stress test the policies that they come forward with. You then bring back the stakeholders to review the results. Look to see if some of the policies can be amended in some way to make them more robust. Run the new policy through the process again, and you continue with this process until you find a policy or set of policies more likely that the stakeholders think are good enough in the sense of being robust enough and, therefore, worth risking taking a chance on.

It fundamentally changes the relationship between the economist and those we purport to serve. It brings the stakeholders who are going to be the ones to suffer the consequences or the benefits of what we do directly into the policy deliberation process, and it doesn't pretend that we have a time travel machine such that we can see tomorrow today. Critically important.

MARK BLYTH: So, in closing, I'm going to suggest that the world may be a little bit closer to what you imagine than you think, and I'm going to give you two examples. The first one is in the academy. People like Jonathan Zeitlin and Charles Sabel and others have been talking about what they called experimental governance for a long time. And there's more and more examples of that which seems very, very similar to what you've just been talking about.

And that-- and that proves its worth. People do this 'cause it works. And the second one is an interesting one. I don't know if you've got the same interpretation as me, but people are looking at the central banks just now and inflation and so on and so forth. And there's a lot of they don't know what they're doing. And I think that there is a conversation inside, which is really goes like this, yeah, we really don't know what we're doing. And what we're doing is what we figured out is that the big models and the big theories that we have really don't work, and what we're doing is becoming much, much more data dependent.

We're looking at the data. We're trying to figure out what it means, and we're taking baby steps. That strikes me as being conscious of the harm you can do if you don't get it right. What do you think?

GEORGE DEMARTINO: I absolutely agree. The question is, how do we hang on to these new insights? I mean, these are hard one. I mean, we've been through crisis after crisis after crisis. And I think over time, the effect has been for, at least on the more sensitive economist's recognition that there's so much we don't know and we'd better find a way to change how we promote good decision-making.

The question is, how do we hang on to this? How do we prevent the next period of great moderation when economists really thought they had it all figured out? My profession goes through moments where we really think we know it all. Then we get turned upside down by events. We become a little bit more humble.

And we're in one of those moments now to which, I say, thank God, we are. How do we hang on to it? How do we institutionalize it? And I think a lot of it has to do with what we train our students to believe about our profession. I think we've got to bring these insights into PhD level training, Ma level training, and even undergraduate training to train future citizens who are going to be listening to economists about the limits of our expertise so that they don't ask of us things that we cannot deliver like long-term cost-benefit analysis and forecasts and the like.

MARK BLYTH: In closing, if you have an NSF grant and you're a PhD Student, you have to attend a responsible conduct of research seminar or series of seminars to discuss these issues. There are no such requirements for recipients of grants that are in the social sciences.

GEORGE DEMARTINO: Yeah.

MARK BLYTH: Isn't that interesting?

GEORGE DEMARTINO: It is interesting, and it's really, really dangerous.

MARK BLYTH: George, it's been a pleasure. It's a great piece of work. It really makes people think. I hope it's widely read, and it's been great to talk to you.

GEORGE DEMARTINO: Thank you, Mark. I've so appreciated the opportunity.

MARK BLYTH: This episode was produced by Dan Richards and Lella Wirth I'm Mark Blyth. You can listen to more conversations like this by subscribing to the Rhodes Central Podcast, wherever you listen to podcasts. We'll be back soon with another episode of the Rhodes Center Podcast. Thanks.