Visit techpolicyinstitute.org for more!
Jan. 4, 2022

Diane Coyle on How Economics Can Evolve with a Changing World

Diane Coyle on How Economics Can Evolve with a Changing World

Professor Diane Coyle is the Bennett Professor of Public Policy at the University of Cambridge. Professor Coyle co-directs the Bennett Institute where she heads research under the themes of progress and productivity. She is also a Director of the Productivity Institute, a Fellow of the Office for National Statistics, an expert adviser to the National Infrastructure Commission, and Senior Independent Member of the ESRC Council. She has served in public service roles including as Vice Chair of the BBC Trust, member of the Competition Commission, of the Migration Advisory Committee and of the Natural Capital Committee. Professor Coyle was awarded a CBE for her contribution to the public understanding of economics in the 2018 New Year Honours. Her new book, "Cogs and Monsters: What Economics Is, and What It Should Be," is available now.

Liked the episode? Well, there's plenty more where that came from! Visit techpolicyinstitute.org to explore all of our latest research! 

Transcript

Bob Hahn:

Hello, and welcome to the Technology Policy Institute’s podcast, Two Think Minimum. I am your host, Bob Hahn. Today is December 16th, 2021, and I’m delighted to be speaking with Diane Coyle, who is the Bennett Professor of Public Policy at the University of Cambridge in the United Kingdom. We will be talking about her new book, Cogs and Monsters: What Economics Is and What It Should Be, which was recently published by Princeton University Press. Diane, welcome to Two Think Minimum

Diane Coyle:

Hello, and thank you for inviting me on to talk to you. 

Bob Hahn:

So, this is a very ambitious book. You cast many aspersions on the economics profession, and you compliment it at various times. There are lots of provocative ideas in this book for fixing some of the problems that we face. But my first question for you is where did you come up with the title, and what does it mean?

Diane Coyle:

The title is one that I used many years ago for a program on BBC Radio Four. And I adopted it here because of the contrast between the stylized and actually inaccurate portrayal of economics as a very mechanistic discipline, which treats everybody as cogs in a machine. It used to be like that at one time, not so much anymore, but the contrast between that and the monsters who used to get marked on medieval maps, where people didn’t really know where they were going, there were exciting and dangerous things out there in the unknown oceans, and they used to mark the maps, “Here be monsters.” And so, it’s about the journey in economics, from the world of cogs in a machine to the world of monsters that we haven’t fully explored yet.

Bob Hahn:

So what was the motivation? You’ve written several great books. One of my favorites is your short book on GDP. What was your primary motivation for writing this book?

Diane Coyle:

There’s a lot of criticism of economics, and I suppose it’s gained intensity again since the financial crisis. And we seem to be back in a world where macroeconomists, in particular, shout past each other. They’ve got very different, incompatible approaches to understanding what’s going on in the economy. And yet I think a lot of the criticisms misfire and they cover, for example, the claims that economics is unrealistic because it uses models, which is just absurd. We might use algebra to notate our models sometimes, but history uses models. “The causes of the Second World War” is a model that’s used in history. Models are a way that you reduce the dimensionality of a very complex world to make sense of it. The idea that economics is only about money, which is simply untrue. Equally the idea that all economics is macroeconomics, and it’s a minority sport among economists, although it’s obviously important and high profile, and the trouble with all of these kinds of everyday critiques is that they’re lazy and inaccurate, and they kind of let economists off the hook because there are things that I think the subject can and should be criticized for.

And so, really that’s what I want to focus on here. It’s a book that’s strongly in support of economics and its power and the contribution it can make as a discipline, yet at the same time wanting it to improve. And there are some serious things that I think need improving.

Bob Hahn:

All right. So, let’s talk about some of those serious things or the problems in the way economics has been done, say over the last few decades.

Diane Coyle:

The one I start with in the book is the social makeup of the discipline. It’s very male-dominated. This is very well documented now. As male-dominated as a few other disciplines, such as computer science or philosophy, funnily enough, or parts of engineering, that people of color are underrepresented, although the data are much less good, I think people from low-income, working-class backgrounds are underrepresented, and you might ask, why does this matter, because there are other disciplines that have a gender skew, for example. And I think it matters because of the claim to be a science of society. If you’re trying to understand and improve the society in which you’re operating, and economists are very influential in government and in business, then to be so unrepresentative is an issue because you don’t know what you don’t know as an individual. You won’t know what questions to ask if there are no people in your cohort who’ve got very different kinds of life experiences than you do. I mean, it’s also for that matter, quite an English language, Anglo American, particularly US-dominated profession. And, you know, in danger of just emitting different perspectives on the course of economic history that come from other parts of the world, different cultural perspectives. So, for a social science, I think that lack of diversity really matters.

Bob Hahn:

Okay. I certainly accept that economics isn’t the most diverse of disciplines, and I’ll take your word on the stats or the statistics. What do you think we should be doing about it? Should we have quotas?  How should we encourage people who are underrepresented, or should we be encouraging or incentivizing people who are underrepresented to get in the game?

Diane Coyle:

There’s no quick fix to this, it’s a system problem. But I think there are things that economists can and should do. One is to talk much more, including talking in schools, about the whole range of things that we do. Because the very strong perception among the public is that economists are about making money and they’re about macroeconomic issues and they shout at each other a lot. And yet, as you and I know, most economics now is applied micro and it’s making a fantastic contribution using new datasets and techniques to understand causal relationships and problems and try to offer ways to fix them. And sometimes that’s about money and valuing things in monetary terms, but it’s not about financial markets and making money in the markets, which is the impression a lot of people have. So we can talk much more about what we do, and that includes going and talking in schools.

It includes curriculum reform, and I’ve been very heavily involved in the Core Economy Curriculum free online course. So, anybody in the United States who looks at the price of economics textbooks, there’s this much better, free alternative online that you can use. And that brings some of the big questions into the first year of undergraduate study. What causes inequality? How can economics give us tools to address issues like climate change? How do innovations come about? What’s the right kind of structure to incentivize innovations? Why does competition matter in markets? 

So, all of those important societal questions that young people are really interested in, we do those. We should put those in the first-year textbooks and not have this bizarre apparatus of socializing people into thinking about indifference curves and isoquants which are not real things at all. So, let’s talk about the real things that we do. 

So, we can do all of that. The other thing I think of is the culture of academic departments. My institute is now interdisciplinary, but it now sits in a political science department. And the culture in other disciplines is really different. There’s a very aggressive culture in economic seminars. Senior men in the discipline heads of department, people who lead seminars, should be doing something about that. So, that whole business of immediate interruptions, quite aggressive tones… I remember I once went to a conference in Toulouse. It must have been in the early 2000s, and it brought together cognitive scientists, psychologists, and economists to understand, to start thinking about behavioral aspects of financial markets, and to my absolute astonishment, the psychologists in the audience applauded at the end of every presentation. You know, the culture was just miles away from the one that we’re used to. The American Economic Association, the Royal Economic Society here, are starting to take action on these things and not be [inaudible]. So, I really applaud that, but we’ve got a way to go. 

Bob Hahn:

So, you started your remarks by talking about macroeconomics. So, could you just give us a primer on what macroeconomics is about, and then maybe we can dive into what macroeconomics can do well and not so well.

Diane Coyle:

Macroeconomics is about the economy in the aggregate. So, looking at measures for the whole economy at the national level, or perhaps regional level, what’s happening to inflation, unemployment, interest rates. And the purpose really is to think about what should the settings of monetary policy and fiscal policy be? And it’s really hard. Macroeconomics is really hard because you are trying to understand and predict the behavior of millions or hundreds of millions of people and millions of businesses. And these are likely overdetermined. It’s very hard to establish causality. The amount of data that we have is not large. They look like big datasets, but they’re highly correlated with each other and highly correlated over time. So, the information you can get out of them isn’t fantastic. And then whenever there’s a structural change, new technology transforms the way business is done, a pandemic hits the world, then those relationships are quite likely to break down.

So, it’s very hard, and I’ve got every sympathy with the people who are trying to do it. I think the problem is that a lot of people who talk about macroeconomics, and certainly in public, talk about it in ways that are much more confident than is plausible. And they substitute for the ability to identify causal links in the way that you can in much richer micro datasets. They substitute for that with a framework of ideas or an ideology, if you like. And when I was a young student, it was all about Keynesianism versus Monetarism. And it reminds me very much of what we have now in terms of Modern Monetary Theory and whatever other alternatives there are shouting at each other about how should monetary policy be operated. It’s just, you know, after the financial crisis, there was a call for more humility among economists. And we certainly haven’t got that.

Bob Hahn:

So, it strikes me when you talk about the problem of macroeconomics, you know, given how complex our economy is, or the world economy is, that it’s virtually an impossible task to be predicting things, you know, say six months, a year, two years down the road. I mean, is this a worthwhile endeavor and what should we be doing there?

Diane Coyle:

I think there have to be some efforts to forecast over the short to medium term because it’s needed for monitoring fiscal policy. So, somebody’s got to have a go, being more aware of the margins of error is really important. We all applauded when the Bank of England introduced what are called its fan charts, and those are the range of outcomes with attached probabilities, for inflation and GDP growth. What I like about their GDP chart, in particular, is that the range of uncertainty actually goes back from the present as well, because the figures will get revised. And there’s so much uncertainty in the way we even measure any of these macroeconomic aggregate aggregates that just doesn’t get acknowledged in debate. There’s a sort of, I don’t know if it’s a media problem or a press release problem or a political problem, but this focus on percentage point changes in GDP growth from quarter to quarter is just absurd. And, you know, I can’t quite understand why we’re not paying less attention to those things. 

You know, alternatives would be to sort of look at a range of different indicators in the way that Alan Greenspan famously did when he was Chairman of the Federal Reserve Board. So, you get a richer picture that way of what’s going on. Anyway, to get back to your question, it’s got to be attempted short term, but one of the things I’m interested in thinking about is the self-fulfilling character of economic policy. In some cases, clearly, policy statements and policy actions affect expectations. Recessions can be self-fulfilling if there’s too much gloom around and that affects people’s expectations. So, this is Keynes, this is old stuff. And I wonder if we need to revisit that kind of approach, which Bob Shiller, in his book, a couple of years ago called Narrative Economics, it was about thinking about the way that the things economists say and the policy actions taken as a result actually change the outcomes of the system that you are analyzing and thinking about that, seems to me, a fruitful way for microeconomics to be going.

Bob Hahn:

I want to pursue that idea a little bit because I had a mixed reaction to what I thought you were trying to say in your book. So, I think your idea was that economists affect the system or the policies because they’re working in government or they’re influential people like Milton Friedman was, for example. And we don’t often take that into account somehow when we’re doing our models. Was that your idea or your suggestion?

Diane Coyle:

Yes, exactly. So, the idea that the actions taken will affect outcomes. They’re not operating on a set of givens. It’s not… we’re not social engineers looking down at a system and tweaking a lever here or a setting there…

Bob Hahn:

So, here’s my take on that, and tell me if you agree or disagree. I’d like to get your reactions. Certainly, central bankers like in the UK and the US are extremely sensitive to what they say in public. And I’m reminded of Greenspan’s statement about irrational exuberance or whatever when the markets tanked and went down 500 points or something like that. But as a microeconomist who studies different industries and studies the impact of “X on Y,” you know, using field experiments, things like that. I don’t see that many of us actually have much influence on public policy that’s easy to identify. So, I guess a couple of questions. One is, okay, let’s take it per your argument that economists do influence the process. How should that affect the way we think about policies, and two, do we have that big an impact on policies? And what’s the evidence for that?

Diane Coyle:

I fear it’s the case that what I say doesn’t apply in all contexts. And there are well-defined situations where you can do an economic analysis and what you say about it won’t affect anything. But I can think of many examples that are not just macro examples, as well. One would be certain behavioral nudges or interventions, whatever you want to call them. And maybe it’s just a question of giving them enough time to play out, but I was very struck by the idea that you could reduce automobile accidents by changing the layout of roads, and in particular, ending the distinction between the sidewalk and the road where the cars drive. So, that drivers become much more uncertain. Therefore, they slow down. And when they’re driving slowly, they hit fewer pedestrians. And there was an example in the Netherlands where that seemed to be working. It was tried on a street here in London, and there was a lot of hype about that, but then it turned out that actually drivers got used to the new pattern, and then they revert to hitting as many pedestrians as ever they did.

So, it’s kind of a risk compensation mechanism, and we know about that in theory. I just think it’s not always brought to bear on the real situations that people are analyzing. Another example though, would be cost-benefit analysis and deciding when you’re going to make an investment in a project like a piece of infrastructure. And this is treated as, you know, cost-benefit analysis applies in contexts where you’re talking about marginal changes, but it’s applied to big projects where there’s a possibility of big changes rather than marginal ones. And actually, you can affect the prospective benefits you get from an investment by the way you do it. So, for example, by aligning other policies, if you’re going to build a new rail link and rail station, are you also going to make sure that the bus routes connect? Are also going to make sure that there are houses nearby or there’s adequate car parking?

So, all of the other things that are likely to deliver an uplift in productivity locally, which might be non-marginal, and yet that’s not done when cost-benefit analysis is used in practice. So, I think there are also situations where we should be thinking about actually the positive scope to change outcomes in a big way. And if you think back to the Victorians and why did they build infrastructure that we are still using 150 years later? Why did they build a century and a half capacity? I think it’s because they had that kind of alternative world in mind that they were trying to build. They knew what they were trying to build. So, those would be two examples of needing to think much more about how both the economic analysis, and the advice and actions that result, could actually change outcomes.

Bob Hahn:

Mm-hmm. So, you mentioned cost-benefit analysis, and I’m a big proponent of cost-benefit analysis, but just listening to your examples, I think what you’re saying is there are some types of cost-benefit analysis where you can drill down and focus on a small problem and some types of problems that are going to affect the economy in a very broad sense, and we may need to approach those problems differently. But are you saying we shouldn’t do cost-benefit analysis, or we should just do it sensibly depending on the problem we’re trying to address?

Diane Coyle:

We should certainly do it sensibly rather than mechanically, and I think where it’s done, it’s often done mechanically and in quite reductionist ways. So, the treasury here has the “Green Book,” which has been updated several times in response to critiques that it’s too narrow. So, it now takes into account and environmental externalities to the best extent possible and recently has started to include what are called well-bs. So, single metrics in monetary terms of changes to people’s wellbeing as a result of a project. That seems to me, quite a reductionist way of thinking about broader impacts of cost-benefit analysis. I’d rather keep them in separate buckets and have cost-benefit analysis stick to its knitting.

Bob Hahn:

So, you touched on the issue of climate change, which is something I recently taught a course on. And you talk a little bit about economic approaches, and maybe you want to talk about that now, and then you talk about your concerns about putting a price on everything. And I guess I wanted to get your take on problems like climate change, what think we should be doing, and what economics is doing well here and not so well, and what leaders might be thinking about doing.

Diane Coyle:

Gosh, big question, big problem. Hmm.. where to start on this? I do think it’s important to have metrics, including monetary metrics, trying to think about the scale of the challenge and put the costs of responding to it in appropriate perspective. Having said that there are ecologists, environmentalists, who are strongly opposed to putting a price and monetary value on aspects of nature. And I can understand why they do that. My counterargument is that we know that zero is the wrong number to put on the value of nature. So, let’s have a number and start to have a debate about it. 

So, I’m not at all against putting measures on either the cost of addressing climate change or the cost of the damages it might cause in future. So, I do think of this, you know, if this is a collective action problem, and I do think that governments haven’t done enough, not so much in terms of the usual debate about carbon taxes, but in terms of things like de-risking private sector investment that’s going to be necessary, investment and innovation in particular, by setting standards, setting switch-over dates, and guaranteeing that there will be markets that have certain characteristics, because for any, almost any private company at this stage, these are highly risky investments to undertake. [They’re] high-risk, high cost.

There has to be some public sector coordination here, and the most market-friendly, pro-competitive, pro-innovation way to do that is to find ways to guarantee what the size of the market will be, what technical standards might be, when it’s going to come into being. And so, I’d favor quite strong action by governments along those lines. This is pretty standard economics. There’s nothing heterodox or outrageous about saying this, but it hasn’t been part of the policy discourse very much, which is largely focused on some quite narrow channels, such as should we have a carbon tax or not?

Bob Hahn:

So, one of the issues you raised in your book is something you call “political economy.” And I think you suggest that economists should be more sensitive to political issues and issues of what’s feasible in the real world. And when I hear you talk about climate change, I would be an advocate, in an ideal world, where let’s say you were the Prime Minister or President of the United States, of you setting a carbon tax for the US or the globe or something like that. The problem I see with that in places like the United States is it’s very difficult for politicians to credibly commit to a tax over time, or a tax even today. So, it seems to me that your suggestion of creating markets that provide more certainty, say to the Elon Musks of the world or budding entrepreneurs, is in some ways a response to the fact that we can’t, in some ways, have a first best where we might have a price on an externality where, like Pigou suggested, to reflect its damages. Can you react to that?

Diane Coyle:

It’s only partly that, but it’s also partly incorporating what we have already learned from some markets like [inaudible] about the pace that change happens because there’s learning by doing and very large economies of scale. And across the board, quite a lot of policy analysis comes from what we learn in school about economics, which is a constant returns competitive world. And doesn’t so often incorporate these kinds of pervasive increasing returns characteristics, particularly with new technologies. So, you know, I guess I would like a carbon tax, but I think that’s not the only tool in our armory. I think we should be thinking about these learning by doing scenarios for new technologies as well.

Bob Hahn:

But also, we have research and development, which addresses the knowledge externality that an individual entrepreneur or whatever can’t appropriate all the benefits of his or her invention. So, that provides another motivation, I think.

Diane Coyle:

Sure. And there’s a very strong, classical economic case for public support for basic research. So again, nothing at all strange calling for that, but the political economy point responds to what you often hear economists say, which is, “We’ve figured out the perfect analysis, the optimal solution in this situation. If only the politicians would implement it.” 

And my point is that if it’s not implementable, it’s not the perfect solution. You’ve done the analysis wrong. You’ve got to take the politics into account in the analysis.

Bob Hahn:

Okay. So, the late Marty Weitzman, I had him down for a seminar in Washington, DC, and Marty was very much oriented towards defining what economists might call the optimal solution, in letting what politicians engage in the fine art of sausage making and do what they do. It seems to me, you’re suggesting, perhaps taking a step beyond that, and forcing the economics or asking the economics profession to think more about feasible alternatives in the real world, is that where you’re going? 

Diane Coyle:

Yes, and actually even further, because I don’t really believe that you can separate the analysis from the political decisions in the way that Marty suggested, much as I hesitate to disagree with somebody so brilliant. And the reason I say that is because if you think about how we would define analyzing a situation, “What’s the optimal thing to do?” We’ve got this concept of efficiency behind it, and that’s a word that to many people sounds like engineering efficiency. You’ve got a leaky tap. You figure out how to fix it, but it’s not. It’s a value-laden concept because it requires this underlying conceptual machinery of utility curves and preferences and Pareto optimality, which are about values. They are about ethical choices or political choices. So, although we ought to be trying to be the plumbers and/or the engineers, and sticking to the analytical part, I think we’ve also got to recognize that we can’t completely divorce that from political or value-laden choices, and it’s not particularly comfortable, but I really think it’s hard facing some of the challenges that we face now to say, “Economics does not involve ethics. We don’t need to engage with all of that.” And actually, part of the reason for the low standing of economists in some circles is that we, you know, there are people who just downright refuse to accept that there are ethical considerations that need to be taken into account in economic analysis.

Bob Hahn:

Okay. So, let me try to parse this or what we’ve been talking about over the last few minutes. So, Ronald Coase, many years ago, presented a version of what I think you’re arguing now, which we should compare… He argued we should compare reals with reals, so realistic alternatives amongst themselves, as opposed to something that may not be politically feasible. My question to you is how do we figure out what’s realistic? That’s a really tough problem. 

Diane Coyle:

Gosh, I don’t think I’ve got an easy answer to that one. And part of it is experience and engagement with these kinds of issues.

Bob Hahn:

So, let me try this out on you, based on my very narrow, limited experience in government. So, when we were designing a cap-and-trade system for limiting SO₂ in the United States, one of the reasons we elected to go down the road of cap and trade was because it gave politicians a lot of flexibility in how they handed out what were then called allowances or property rights. So, in that sense, we were trying to address the feasibility question really in two ways. Giving them the flexibility to hand out the property rights, and also the flexibility in determining how much pollution they thought society should be willing to accept.

Diane Coyle:

Which sounds a very pragmatic way to go about things. So, I’m not sure I know what the question there is.

Bob Hahn:

I’m not sure what the question is there either, but I guess, the question is that what you mean when you urge economists to think about solutions that might be feasible, to take into account the political motivations of legislators?

Diane Coyle:

Yes, although I think about it more in a negative way, I suppose, to think about the constraints on politicians, what can they not do? And that defines the territory over which you can search for optimal policies, because what is the point of recommending something that can’t possibly happen? And I think perhaps, well, maybe in a more positive way also, you know, if you think about substitutes for a carbon tax, it might be increasing duties on gasoline, petrol, and that’s very regressive. They tried it in France. It brought protests onto the roads every weekend, the Gilet Jaune protests. So, maybe there that directs you towards thinking about policies to price carbon in some way to address climate change issues, but to do it while also thinking about the constraint of fairness. What about the increment inequality implications? Do you need to pair it with some kind of compensation for the losers, such as rural drivers? So, thinking about it in that more holistic way.

Bob Hahn:

So, I think you’re saying be a little bit more creative about the solutions we come up with, and don’t just sort of narrowly focus on what might be sometimes termed “first-best efficiency.”

Diane Coyle:

Indeed. I’m very fond of the “Second-Best” paper, I have to say.

Bob Hahn:

Ok. So, on that, and, and I know you’ve written about this in the context of GDP and thinking about welfare. How should we be evaluating policies? What are the dimensions that you think are really important when we put our pen to paper?

Diane Coyle:

Some of the old ones are still important. I call my book about GDP, a brief, but affectionate history because there are some good things about GDP. It’s correlated with things that matter to us, like jobs and incomes. The thing I am pondering a lot at the moment is about what other dimensions should we be incorporating? Environmental issues, for example, environmental externalities, depletion of natural resources. How do we incorporate that in our metric of success? And I think there’s quite an obvious, very economic-based answer there, which is to think about the assets and the depreciation of those and assets over time. If you’re thinking about natural capital and look at net GDP, taking that into account, so we can use the conventional metrics as a base and start to think about how to improve them. 

We can collect figures about what’s happening in the household economy, particularly over the pandemic, and understand better how members of the household have to use their time to deliver care. What are the transitions between the market economy and the unpaid economy in the household, and collect those kinds of metrics as well. There’s a lot of interest in understanding well-being more broadly, and that takes you to psychological factors, such as autonomy or a sense of purpose, which contributes significantly to an individual’s well-being. That has an economic dimension to it because many people are in jobs which remove all of their autonomy, that are surveilling them, or telling them exactly when they must work, and under what conditions, for low job quality. 

So, that gives you an array of things to think about. And I think the challenge in policy is how many of those can you pay attention to, and how do you think about the trade-offs among them? It’s a very complex world. There are many dimensions of things that we care about. There’s a whole philosophy, Amartya Sen’s capabilities approach, which is multidimensional and argues that we should think about the whole array. And that led him to advocate for dashboards of indicators, but they’re not very politically salient and they’re complicated to interpret. So, I think this is a big question for me. And, you know, if I’m a public policy professor, I want to make things better for people. How do I define and measure what I mean by better? And how do I translate that multidimensional, complex world, into the kind of, “We will do this, or we won’t do this, decisions” that are practical politics?

Bob Hahn:

So, you are Distinguished Public Policy Professor, and one of the things that you talk about in your book as a way of improving public policy is by rooting public policy in solid empirical evidence. So, if you were the Prime Minister or the President of the United States, and you’d have my vote, what would you do to help ensure that outcome?

Diane Coyle:

Evidence-based policy has been in fashion and out of fashion. And there’s a wonderful quote from Keynes that I will mangle, but it’s along the lines of “No politician ever wants any evidence to overturn his dearly held theories.” So, I would increase funding for national statistical offices to produce more open data. Data’s a public good in the classic sense that it’s non-rival. There’s a lot of data around that’s privately held. There are property rights over it, and people are excluded from it, but we should be investing in data as a national asset to help politicians make better policy, also to help businesses make better decisions, and to inform any interested individual. So, I would invest in that. If we’re talking so much about a data-driven economy and how important it is for business, it’s important across the board. So, I would definitely do that.

I would make economics students learn more about data in a practical way than has been the habit over recent decades. And students are taught quite a lot of theoretical statistics in their econometrics courses. I think it’s improving a lot now, but actually hands-on, not only how do you do the econometrics and how do you write the code in R or Python, but also, how do you think about how the data was constructed? Where has it come from? Who put that data together? It’s the only lens we have to understand society, but it’s a lens that reflects existing social relations, past structures, classifications, and so I would like people to be much more questioning about that evidence. And I say this partly because machine learning is being deployed so much in certain sectors of public policy without enough thought being given to what that data embeds. Data bias has become quite a prominent issue, and I think people are aware of that now, but there are some subtler issues about how data get constructed and shape what we think is going on in the world.

Bob Hahn:

So, let’s talk about a related issue, which you do touch on in your book. And Ted Miguel talked about this in a previous episode of Two Think Minimum with me, which is how data gets used by economists and whether we can trust the results that they actually get and publish in papers. So, you’re well aware that editors have an incentive, or academics have an incentive, to get results that are different from others, and they sometimes torture the data to get those kinds of results. How do we deal with that side of the question that social scientists or scientists may not be… maybe publishing things that aren’t actually always true? 

Diane Coyle:

There’s a lot of dysfunction in the academic publishing system, and particularly economics, where the gateway for the general journals is so narrow. You know, we’ve got far fewer of those journals than many other disciplines. So, if five Nobel Prize winners can’t change that easily, that tells you it’s a hard problem to fix. The incentives are quite embedded in the academic world. So, I don’t have a quick solution for it. I think a slow solution is actually to teach the next generations of social scientists to be much more cautious about data, but you know, even grand as I am these days, not needing to worry too much about publishing billions of journal articles, I get editors saying, “This is very interesting. Can you run this regression?” 

And my answer is, “Well, no. The data quality won’t support that the power will be very low. I’d have to beat it up to get any statistical significance, really. What’s the point of doing that?” And you know, that demand for a certain mode of discussing evidence is so strong. It’s very hard to escape it.

Bob Hahn:

So, one of the suggestions that Ted Miguel had, he’s a professor at Berkeley who was on Two Think Minimum, was to register what your hypotheses were before you torture the data, with the idea that you’re less likely to be going fishing for spurious correlations and things like that. What’s your take on that as at least a first step to address this issue?

Diane Coyle:

I like the idea. I think it would involve a different way of training graduate students how to go about it, and indeed, existing scholars, how to go about it. I suspect that the number of positive results, if you like, that emerge from such a system would be much, much lower than appears to be the case now. 

Bob Hahn:

So, now there is a lot of evidence that people do, you know, go fishing for particular findings that may be attractive to editors. So, I’m inclined to agree with you. One of the points you make in your book is how economics has changed a lot in the last few decades, sometimes for the better. Do you want to give a couple of examples of that? I was reminded of some of the use of randomized control trials, but you may have other examples.

Diane Coyle:

Well, I think the whole kind of explosion of applied microeconomics is a great example and it’s the techniques, it’s the databases, and the computational power that’s available to do that. And, you know, it’s the kind of thing that I think many non-economists don’t really appreciate as part of economics, it gets reported in the media sometimes, but they probably think it’s to do with education or to do with health. It’s not really economics, but of course, in areas like that in all of these field disciplines, there’s been an enormous contribution in recent decades. I suppose another, thinking about the past couple of years and the financial crisis, is that actually we’ve learned about the macroeconomic responses to crises and comparing outcomes in 2008, 2009 or outcomes in the past couple of years to the Great Depression, we really have improved through all the fact that there’s no, you know, kind of intellectual consensus about why it works. It does seem to have worked incredibly well.

Bob Hahn:

In your book, you single out, or maybe not single out, you highlight the importance of the digital economy and how, in some ways, the characteristics of economics have changed. Why did you do that? And what is significant about the digital technology or the digital economy in your mind?

Diane Coyle:

Well, this is my own area of real interest. And my first book about it was published 25 years ago. And there are two aspects, really. One is that we’ve had this amazing transformation in the way we lead our lives, particularly since 2007, with the advent of smartphones and 3G and the design that made it possible to have all of the free apps on our phone. And so businesses, business models are changed, and the way we lead our daily lives and go shopping has changed. We spend 24 hours out of every week online, and this is pretty much invisible in economic statistics. So, it’s quite hard to tell, you know, you wouldn’t notice it if we were just looking at statistics, and it raises lots of questions about how do you really measure price indices or productivity in the face of this kind of big structural change that economic statisticians are starting to grapple with, and there’s a lot of interesting work there.

But the other thing that strikes me about it is that although we’ve always had increasing returns in important parts of the economy, important spillovers in the economy, they’re now everywhere. The dynamics of digital markets, the fact that network externalities are so pervasive, means for me flipping our benchmark where the starting point for thinking about how our market operates or what kind of policies to devise. So, rather than starting from the textbook conversion that we all got taught in school of increasing of constant returns, perfect competition, and all of that framework of the welfare theorems, we should be starting from a place where those assumptions just don’t hold. 

Bob Hahn:

So, can you explain that a little bit more? You have a wonderful table that I was swimming through. I found it on page 180 of your book, where you do talk about sort of 20th-century economics and 21st-century economics. You know, 20th-century economics was linear. 21st is nonlinear. 20th was static and 21st was dynamic and whatever. So, one, could you elaborate on that a little bit? And then once you explain it, can you talk about what all this means for a research agenda for the profession and ultimately what it might mean for policy?

Diane Coyle:

Well, I think, you know, if you could read out the table, that might be the best elaboration of it. It is about what you take to be the best description, the best thumbnail description of how the economy operates. And it is that it’s non-rival, it has increasing returns. It has non-linear dynamics, such that things happen slowly and then they happen very quickly, and markets tip. There are a lot of spillovers between people, a lot of externalities. And this adds up to, I suppose, it’s partly theoretical, but it’s also just partly empirical questions. If you are sitting in a competition authority and you want to think about merger control or market dominance, do you have the tools to understand what it is that makes a market tip to one dominant player? When will that happen? What are the metrics? How do you think about the scale of that?

So, I think there’s both a sort of conceptual issue to get the framework in a way that this would be the framework that you would teach in graduate school at the starting point, and your perfect competition, constant returns, linear world, would be a special case of that. And then really quite a wide empirical applied and policy agenda, thinking about if that’s your world, how do you translate that into understanding setting competition policy or thinking about cost-benefit analysis? The natural world is full of tipping points. Do we have any idea on the empirics of when biodiversity in a certain area will collapse in agricultural productivity will collapse? And I think the answer is, no, we don’t.

Bob Hahn:

So, one of the issues that you raised earlier in the book was government failure. When the government might intervene in ways that are not so beneficial, I presume for consumers. How should we weigh government failure in sort of thinking about public policy, as opposed to just sort of allowing market participants to do things and perhaps address problems?

Diane Coyle:

Well, clearly there’s a lot of government failure. Governments do stupid things, and sometimes that’s the fault of their economists, and sometimes it’s because analysts haven’t persuaded politicians about the right thing to do. There are some things that you could fix about training the kind of people who go into bureaucratic roles so that they have a more statistical or scientific background. I don’t know about the figures on that in the US, but in the UK, Whitehall civil servants are largely from arts and humanities backgrounds, social sciences, but not the sciences. So, you can do some things to fix it. But I think the truth is that we will always have government failure, as we will always have market failure, and actually the reasons they fail and the context in which they fail are pretty much the same. And they’re just underlying problems of information asymmetries or non-rival public goods problems that it’s going to be hard to do. And we are imperfect humans. We ought to try to learn from experience. We ought to have more evaluation of public policy, so we understand what went wrong. There’s a very dysfunctional political media-analyst relationship where changing course is seen as U-turns and called out in the press as if it’s a terrible thing, when it actually might be a really good thing to do. So, we could limit the scope of government failure. We could improve the quality of government, but I don’t think we’re ever going to eliminate it.

Bob Hahn:

So, I want to conclude by asking you one question related to a point that you brought up in your book and you brought up in today’s podcast as well, and that has to do with economists or other experts getting on TV and having an air of certainty that might not be warranted if you actually look closely at the data. Is there anything we can do about that?

Diane Coyle:

Oh, goodness me. I’m not sure there is. 

Bob Hahn:

So, let me tell you my take on that. And this is from my being in the UK and being a fan of the BBC, which you obviously support and were very helpful for. Journalists who are interviewing people can take what they say is given, or they can press them on it. And so, my take is we’re not going to necessarily get rid of what I might call “the talking head problem” completely, but somehow if we could get journalists to be a little bit more insightful and aggressive in the kinds of questions they ask, I think it would force people to be a little bit more honest in their responses. And my factoid for proving that is that in the UK, it seems that they do that more in the US, and I think we observe more honest discourse as a result.

Diane Coyle:

Well, the BBC is a fine institution that is the kind of stake in the ground for journalism here in the UK. The ability of newspapers to sustain specialist correspondence has obviously been undermined by their decline in advertising revenue. But specialisms are also really helpful because you get journalists who understand better what they’re talking about, but there are other means of communication. And I have always put a lot of time and effort into both curriculum, work, teaching, but also public communication of economics. Partly because having an audience that can question is part of the solution as well, I think. And the whole economics blogosphere and Twitter-sphere also has its downside, but there are at the same time, many good people are trying to do that kind of communication. So again, a lot of difficult problems, no quick fixes. I’ve been a very unsatisfying podcast guest in that respect.

Bob Hahn:

Actually, I beg to differ. We’ve been talking with Diane Coyle about her new book, Cogs and Monsters. Diane, thanks for joining Two Think Minimum.

Diane Coyle:

I’ve really enjoyed talking to you, Bob. Thank you.