Transcript provided by Kensho.



Nathan Hunt: This is the Essential Podcast from S&P Global. My name is Nathan Hunt.

We are rational beings. We make rational decisions in our own self-interest, buying only the things that we should buy at a price that is both prudent and within our means. As a result of all of these individual rational actions, our markets are flawlessly efficient, free of ranked speculation and animal spirits.

I'm being sarcastic. Homo economicus, that supposed rational actor, beloved of classical economic theory, has been steadily discredited by experience. Well, a few economists working today would defend the notion of homo economicus in its pure form.

My guest today is perhaps unique in considering the implications of the fall of this idea for economics and for economists themselves. Diane Coyle is the Bennett Professor of Public Policy at the University of Cambridge. She is the author of numerous books, including her latest, Cogs and Monsters: What Economics Is, and What It Should Be.

Diane, welcome to the podcast.

Diane Coyle: It's a pleasure to be talking to you, Nathan.

Nathan Hunt: Diane, there is no shortage of people lining up to tell us what is wrong with economics. But one of the things I found so fascinating about your book is that you insist that, well, there are many things to criticize about economics. The typical criticisms are mostly invalid. How do people typically criticize economists? And what are they missing?

Diane Coyle: Well, I should start by saying I am an economist. And so maybe people want to aim off for some of the answers that I give. There are a suite of typical criticisms that go something like, well, of course, we are not calculating rational beings in the way that you described.

We're impulsive. We are subject to emotions and so on. And so how unrealistic is economics? Criticized the use of mathematical models as being too abstract and nothing to do with real life and a very familiar range of criticisms, which, I think, largely misfire.

Sometimes people act like rational homo economicus and sometimes not, but it doesn't matter to economists because many, many models that we use now take account of different ways of forming judgments and making decisions. And the behavioral economics revolution is incredibly widespread.

And actually, in some situations, people do behave rationally. And in fact, pigeons and monkeys and fungi behave rationally. So let's talk seriously about the decision-making context that shapes that. So that one, I think, is no longer valid or that it might once have been.

The idea that it's too mathematical is, again, or models are too abstract, is something that I think misfires. Historians use abstract models. If you're talking about the courses of the second world war, you're doing the same thing.

You're taking a really complex, messy environment and trying to pull out the key features of it so that we can make some sense of it and think about perhaps what we ought to do differently. Those, I think, are overstated criticisms. And the trouble with these is that it lets economists off the hook for things that actually do need to improve about the profession. I count myself as being quite mainstream. But there are things that I would definitely criticize, and that's what this book is about.

One of them is the individualism that you flagged out. Sometimes you make decisions as individuals. But often, and particularly in the digital economy, we influence each other in quite profound ways. So we ought to think about collective decisions.

We are not a very diverse subject. It's mainly male, mainly white, mainly posh. And that's a problem in a social science because you don't know that you're asking the right questions. You don't understand the lives of the people that you're trying to analyze.

We need to think more deeply about what our benchmark ways of thinking about the economy in a world of digital platforms, data, where there are all kinds of features that make up what economists call public goods, where the economic characteristics are so different from the characteristics that underpin the free market intuitions, the idea that the market largely sorts things out, and the government just needs to fix some of the failures.

So that's a long introduction to the themes of the book and what it's really -- I want people to criticize economics for the right things and not the wrong things.

Nathan Hunt: One of the other points you make in the book is that we assume every economist is a macro economist, but that is not true for the profession today.

Diane Coyle: No, indeed, and probably always was a minority of the profession. Since the revolution in applied microeconomics, for which the Nobel Prize last time was awarded, bulk of what economists do in academia and in government economics is what many people might think of as applied social policy.

It's looking at areas like health economics or pensions or obesity and how to tax food effectively. So many people have this impression of people in city firms, perhaps in gray suits, who stand up and talk about what markets are doing or what's going to happen to inflation and unemployment.

And necessary as that is, that's not what most economists do and misleads people. Even way back into school, where you find that the reason girls often say they don't want to study economics is that they're not interested in going into the city and making money. They want to make the world a better place. Actually, economics is a great tool for making the world a better place. But you need to understand how wide its scope is.

Nathan Hunt: Cogs and monsters is a great turn of phrase. I would argue, it's one of the best turns of phrase in economics that I've heard anyway since the invisible hand. So let's start with cogs. What are economists getting wrong when they think about people as cogs?

Diane Coyle: It's the idea that you referred to in your introduction, Nathan, it's this individual, atomized way of thinking about how does the economy operate, whether it be a single market or the macro economy as a whole. And it isn't that individualism is useless. I mean, we do often take decisions. If I go and think about what kind of sandwich I want to buy for lunch, that's, to some extent, influenced by social norms, culture, what's typically available in my cafe, but that's my decision.

So in many context, the methodological individualism, as we social scientists would call it, is fine, but it's not the whole story. And I think economics has, for too long, ignored those more collective social parts of the story. For example, that social norms do change the way that we make decisions, that the range of decisions open to us or that we bring in the social moral values. For example, if there are shortages, normal people think that rationing is the right way to approach that, whereas economists think that raising the price is a more efficient way to do it.

In the economist's sense of efficient, that's absolutely correct, but that's not actually how people think about these kinds of context. And as I mentioned a moment ago, in the digital economy, the social is actually expanding its territory. If you think about data, even if I make an individual decision to sign over my personal data to a tech company that's collecting, it. It's an individual transaction, a property transaction. But me giving information about me actually gives the company information about people like me as well. There's an inherent spillover in that.

Or similarly, the value of a social media company or a platform company like Uber is having more people on each side of the platform. There are network effects. The more drivers, the better for each passenger. The more passengers, the better for each driver. That's an inherently social context.

Advertising, innovation, new goods -- we don't know that we want new inventions until we hear about them and other people start using them or they get marketed. That's the bit that I think the discipline has ignored. And it's becoming more important when you think about policies in the context of the digital economy, which has these increasing returns to scale everywhere, public good features and nonrivalry. The fact that it doesn't get used up or depleted when one person uses it, many people can use it at the same time. And these make for just a different kind of environment if we're thinking about policy in particular.

Nathan Hunt: So let's move to the other side of this, who are the monsters? And how does traditional economic thinking help to create more of them?

Diane Coyle: Well, I suppose, the monster then, it's a reference to those antique maps, where you see the outline of known coastlines and known cities. But there's so much that's not known, and it gets labeled here to be monsters. And it's a reference to the fact that in the economy that's taking shape now with digital transformation, there are these unknown territories, which traditional economics doesn't yet give us very good tools for thinking about. So if I go back to the point that there are massively increasing returns to scale and new technologies and feedback loops, network effects.

This means that there are many, many potential outcomes from where we start today. If you think about something like the green transition, the actions of the government and thinking about policy can determine whether or not we can achieve that by using government procurement tools to favor certain technology by setting or guaranteeing a market, derisking it for investors, setting technical standards.

So it's not subsidies, but it's a way of shaping what kind of outcome does the economy go to out of many possible options, what we'd call in the job and what we call equilibria. How do we think about policy in those areas? Well, you need to be able to put some estimates on what the potential nonlinearities are, the potential externalities, the dose response function in -- another kind of jargon. And we've not got very practical tools for doing that. And actually, the standard approach is still to start with a more linear and predictable world without these tipping points in what the potential outcomes might be.

Nathan Hunt: Some economists seem to wish to be pure technocrats, objective clinical observers of the economic life of a society. Is this a realistic aspiration, kind of social scientists stand outside of this society in which they live?

Diane Coyle: Well, you picked up a very important point I want to make there, and the short answer is no. The idea that economics can be objective has quite a history. And in -- all the way back to Adam Smith, he advocated for acting like the impartial observer. But that got taken to an extreme in the early 20th century under the influence of logical positivism, which says that you can only know 2 kinds of things, either logical truth or empirical facts. And anything else is speculation or metaphysics, as they would have referred to it.

It's got picked up in economics by Lionel Robbins and by Keynes, who said economists ought to be like dentists, the practical people, fixing known problems. It was affirmed again by Milton Friedman in the 1960s, who talked about separating what is and what ought to be by Esther Duflo, the more recent Nobel laureate, where she talked about economists being plumbers.

So the idea that economists tackle objective problems in an objective way. Data, as applied to the situation, you bring the best empirical message you have to bear on it and you come up with the best answer. And this is a delusion, actually, because the idea that we have a best in economics is something called Pareto efficiency.

And it says that an efficient outcome is one where you can't make anybody else better off without making somebody else worse off. And that sounds pretty reasonable as a kind of criterion for taking a policy decision. But actually, it rules out any redistribution because that implies making somebody worse off. And actually, there's no policy context that doesn't have winners and losers. You're always talking about taking something away from somebody.

And we've misled ourselves, I think. We're using this word efficiency. It's not an efficiency criterion like engineering or plumbing. It's a value judgment about efficiency. So economists talk about things being the best outcome by using efficiency, in this sense, and there's a lot of implicit judgment in that. And I think this hasn't done the profession any favors because when people say what economics is, etiological will go no, no, no.

We're just looking at the data from big objectives and looking at the evidence -- evidence-based policy. Actually, there are value judgments implicit in that. And I think we should acknowledge it and at, the same time, do our best to be objective about it, to be impartial about it, even while admitting our own value judgments and preferences. So it's complicated. But I think the backlash against expertise is partly about the fact that economists have claimed now for almost a century to be objective, when actually, we haven't been.

Nathan Hunt: This false objectivity, the idea of a pure intellect engaging with pure economic ideas, free of bias or favor, is a very male way of thinking. I say this as a man, we find the idea of our own objectivity quite seductive. Do you think this problem in economics might be related to the well-established gender imbalance in the profession?

Diane Coyle: That's a very interesting question, and I'm not sure I've got a neat answer off the top of my head to that. There is something about the admiration for technical brilliance and fancy mathematical or econometric techniques, a certain kind of showiness us that I think is very characteristic of men in the male-dominated disciplines. And you probably find the same in some seminars in philosophy, for example, which is another very male subject. So I would agree with you to that extent. But I think it's actually a really interesting research question.

We spoke a bit earlier about does the narrowness of the social makeup of the economics discipline, restrict the range of questions that people ask, restrict what people know about because you don't know what you don't know. And I think you've just posted a really interesting question that I don't know that we got the answer to, but I'd be quite interested to find out if somebody wants to research it.

Nathan Hunt: In the book, there is a chapter on the use of AI and algorithms in economics. Does the use of these technologies lead to more of the same, more fall subjectivity?

Diane Coyle: So yes, it does. And that's one of my concerns. I distinguish using machine learning tools and algorithms as a kind of empirical technique, which is obviously absolutely fine. But it's when you translate that into recommendations for action in the real world, but I think it becomes a little bit worrying. Because if you think about it, an AI algorithm is homo economicus, but much better than any actual human doing it.

They are set objective functions, reward or loss functions, and they maximize subject constraints. That's exactly what we do in conventional economic models. And so applying that in context where reality is messy, there are conflicts of interest, so incentives are not aligned. You might have missing data, you might not have the full picture.

I find that incredibly troubling. Issues about data bias, I think, are well known because we collect data from a society that has an inherent structural biases. And there's work underway in the machine learning community, the AI community, to try to address that.

But I think there are other challenges as well that are not very much discussed. One, for example, would be, what about the data we don't have? You might be thinking about planning a public transport system, and you could do that much better if you can take advantage of all the data that's out there, about the journeys that are made and congestion times and using road sensors.

But that doesn't tell you about the journeys that are not made because that data is missing. What about people who might have better access to labor markets and well-paid jobs if there were a bus service from where they live into town? We don't have that kind of information. And the other thing that worries me is where incentives are not aligned. If your bank is using machine learning algorithms to check fraud, that's fantastic. Because they don't want the money to be stolen, and you don't want the money to be stolen.

But if you're somebody whose amount of benefits is in question, or you're being sentenced by court, which is using an algorithm to determine the length of sentence, that's a much more troubling area. And then lots of policy domains, actually, such is useful. Not being clear is the way that we get compromised and manage to move things forward. So to give an example from one paper I did, you might agree that this right sentence for burglary is 5 years. But you might disagree profoundly about whether you should be maximizing rehabilitation on minimizing reoffending.

And those could give you different kinds of objective functions if you're trying to code them into a machine learning system. So I think we need much, much more thought, particularly about the data, but also about conflicts of interest and value judgments before we start giving advice as economists about these policy areas where there are winners and losers.

Nathan Hunt: How do you solve for these issues? They seem so massive? How can an economist become a tool for better policy within the limitations of their own bias and their place in the society?

Diane Coyle: The key thing for me is actually just acknowledging that there are these limitations. And for many people, the impression of an economist is, it's the guy who -- and it's usually this guy, of course, who stands up on the TV news or appears on the radio, making some statement about what is definitely going to happen to inflation or to unemployment in the economy.

And then there's another one who says exactly the opposite. So it's a kind of Punch and Judy show, which is partly about the way the media treats it because they want to get stories, but it's also partly about overconfidence. And I would hope that we learn to better acknowledge our limitations.

After the financial crisis, of course, there was a lot of calls for economists to be more humble. And I don't think the humility lasted more than a year or so. And in macro in particular, which we talked about earlier in the interview, it is back to that Punch and Judy show, "Oh, yes, you did. Oh, no, you didn't." And what is the public supposed to make of that?

Nathan Hunt: Diane, I had an interesting conversation recently with one of your Cambridge colleagues, Gary Gerstle, who's a historian. And he brought up the idea that we are at a bit of an historical inflection point. Do you think economic thinking is at an inflection point, and should it be?

Diane Coyle: Gary is a very dear colleague. I was talking to him about this just the other day. And the short answer is, yes, I do. There's a pendulum in the way people think about how the world is ordered, their sort of public philosophy, if you like. And for the period of the 1960s and 1970s, largely, that philosophy brought into [ kinds of ] ideas and demand management. And we had a period of prosperity, which then started to crumble as the economy changed and various aspects of that reached their limits. So we had the experience.

The 1970s, the oil price crisis happened at the same time, the stagflation, and [ dependent ] and then swung. And since then, we've had what some people call the neoliberal view, the Ronald Reagan-Margaret Thatcher revolution in politics that incorporated a certain set of economic ideas about markets first, being most efficient about the importance of private sector wealth creation.

And it's not that, that was an entirely wrong, of course, but it was an extreme swing in the pendulum that got to extreme lengths. And I think it's going into reverse now. These things are stimulated partly by events. And we've had the financial crisis. We've had the coronavirus pandemic and [ net shock ]. We've had the subsequent supply chain disruptions and the war in Ukraine. So the events are coming together in that kind of crisis, about the way we think about how the world is ordered.

But at the same time, the structural change from digital technologies, and we spoke about that a little while ago, the way that, that's leading to fast returns to scale, non-linearities and so on. And so that hasn't yet crystallized in an alternative world view. But I think, and I probably also hope that, that is emerging. And if we look back in 10 years' time, we'll say, yes, this was the hinge where we got into the next phase of how we think about organizing the world.

Nathan Hunt: The book, once again, Cogs and Monsters: What Economics Is, and What It Should Be. Diane Coyle, thank you so much for joining me on the podcast today.

Diane Coyle: A pleasure talking to you, Nathan. Thank you.

Nathan Hunt: The Essential Podcast is produced by Kurt Burger, with assistance from Kyle May and Camille McManus. At S&P Global, we accelerate progress in the world by providing intelligence that is essential for companies, governments and individuals to make decisions with conviction. From the majestic heights of 55 Water Street in Manhattan, I am Nathan Hunt. Thank you for listening.