The Joe Walker Podcast - Radical Uncertainty - Mervyn King

Episode Date: August 8, 2020

Mervyn King was Governor of the Bank of England from 2003 to 2013.Show notesSelected links •Follow Mervyn King: Website •Radical Uncertainty, by Mervyn King and John Kay •The General Theory of Employment, Interest, and Money, by John Maynard Keynes •'Debt deflation: Theory and evidence', paper by Mervyn King •The End of Alchemy, by Mervyn King •Recollections of a Bleeding Heart, by Don Watson •The Poverty of Historicism, by Karl Popper •Obliquity, by John Kay •'Truth and Probability', essay by Frank Ramsey •Frank Ramsey: A Sheer Excess of Powers, by Cheryl MisakTopics discussed •Why did Mervyn choose to study economics at Cambridge? 11:06 •Keynes' General Theory. 17:41 •Debt-deflation. 23:58 •When Mervyn met Ben. 29:20 •Was Mervyn caught off guard by the Global Financial Crisis? 31:40 •Does stability lead to instability? 38:18 •The stability heuristic. 41:43 •What is radical uncertainty? 49:54 •How technology creates radical uncertainty. 1:04:57 •Why has the economics profession overlooked radical uncertainty -- and when did this blindspot begin? 1:13:10 •Narratives. 1:25:18 •What does it mean to be rational? 1:31:21 •Speculative bubbles. 1:40:25See omnystudio.com/listener for privacy information.

Ladies and gentlemen, welcome to the show. This episode of the podcast is brought to you by none other than the Dollar Shave Club. Do you have hair on your face or elsewhere on your body that you'd like to remove on a periodic basis using the most effective and luxurious products available without the hassle of going out to the shops to buy expensive shaving gear, which you always forget to do anyway, so you end up resorting to a rusty razor that renders a regrettable red rash. Well, you need Dollar Shave Club. I love these guys and I get genuinely excited when their boxes arrive in the mail. Here's how it works. You sign up for their starter set, which includes a weighty executive handle,

six blades, and a tube of their shave butter. The blades are the best I've ever used. It's like shaving with a lightsaber. Then you tell Dollar Shave Club how often you want regular shipments, whether that's every month or three times a year. There are no long-term commitments and you can cancel your membership at any time. Now, for a limited time, listeners of this podcast who join Dollar Shave Club will get the starter set for just $15. What? To get your starter set for just $15, go to dollarshaveclub.com slash swagman. That's dollarshaveclub.com slash swagman. This episode is also brought to you by Goodwill Wine, our newest sponsor and one for the Australian listeners. CEO and founder Dave is a listener of the podcast and I love his story. Just over 10 years ago,

he lost everything he owned in the Black Saturday bushfires and thanks to donations from around the country, he was able to rebuild. With just $15,000, he built Goodwill Wine. Goodwill Wine produces awesome Australian wines and gives 50% of the profit back to a charity of your choosing. So far, they've given $350,000 and counting to charities around the country. Half of their team are long-term unemployed or living with a disability. And also, the wine's not shit. I've tried it and it's not shit. So, head to goodwillwine.com.au and buy the mixed red case.

If you enter my exclusive voucher code SWAGMAN, you'll get free shipping and an upgrade on the Pinot Noir. So if you're an Aussie and you like red wine, let me repeat, you must go to goodwillwine.com.au, select the mixed red case and enter the discount code SWAGMAN. This show is also sponsored by Blinkist. Blinkist is an app that condenses the best nonfiction books in the world into 15-minute

summaries or blinks, which you can read or listen to. Now, this is perfect for people who want to cheat on their book club or people who want to get ahead and consume as many non-fiction books as possible in a short space of time. But how I use Blinkist is in none of those ways. I use it as a way to filter out and choose which books that I'd like to read. It's kind of like the Amazon look inside feature or the Kindle sample feature, which allow you to see what a book is like, except whereas those features generally only give you the introduction, Blinkist summarizes the core thesis of the book, so you know what you're committing your money and most importantly, your time to,

because after all, the opportunity cost of reading is the time that you spend on that book, time that you could be doing something else or reading a better book. So, you want to make sure it's the right book. Blinkist is where that decision can be made. It's like a really smart friend who's read all the best non-fiction books in the world and can tell you what is worth reading. So, what you need to do is go to blinkist.com slash swagman, where you'll get 25% off an annual subscription and get to try Blinkist Premium free for seven days. That's Blinkist.com slash Swagman. You're listening to the Jolly Swagman podcast.

Here's your host, Joe Walker. Ladies and gentlemen, boys and girls, Swagman and Swagettes, welcome back to the show. It's great to be back with you. This episode was recorded recently, and it's a conversation with Lord Mervyn King. Mervyn was governor of the Bank of England, that's England's central bank, from 2003 to 2013. I'm not going to spend much of the Bank of England, that's England's central bank, from 2003 to 2013. I'm not going to spend much time by way of introduction because I think we cover a lot of Mervyn's background and also the nature of the topics that we're going to discuss at the beginning of the conversation. But I do want to apologize up front for the at times impaired audio quality. Unfortunately, that's the nature of the beast

when you're speaking with someone over Skype or over Zoom. You're constantly at the mercy of a fragile internet connection. So if at times the audio sounds kind of wobbly or wobbly or like digitally impaired, I apologize. It doesn't last for the whole conversation and I implore you just to try and push through. It's still totally audible and I think you're going to enjoy the conversation.

This is a conversation about radical uncertainty, which is the topic of a new book Mervyn has co-authored with John Kay and it's a topic that is close to my heart, a topic I'm very interested in. We also talk about financial instability as well, because obviously Mervyn was at the helm at the Bank of England during the global financial crisis. So without much further ado, please enjoy this conversation with the great Mervyn King. Mervyn King, thank you so much for joining me. It's a pleasure to join you, Joe. I'm very excited to speak with you once more.

We've spoken previously, and I thought I'd do something a bit different this time around and not come with an agenda or any list of questions prepared and just try and have a freewheeling, open and discursive conversation with you. First, I want to introduce you to everybody, Mervyn King, born in 1948. Where were you born? I was born about 30 miles northwest of London, very close to the town where my father and his parents and family all came from, Chesham in Buckinghamshire. My grandfather worked in a factory moulding wood into furniture.

My father worked on the railways before the war, well, during the war, was an engineer during the war. And then after the war, he joined the program, which had been put in place to train teachers, new teachers, servicemen coming out of the Second World War could train to become teachers. And he did that and started his career as a teacher, first in a primary school, then a secondary school, eventually becoming a headmaster. And as a result, we moved around the United Kingdom a great deal.

And what did your mom do? She stayed at home and looked after my father and my brother and I. Before the war. Would you describe? before the war she actually um was in service um for a short period when she left school at the age of 13 14 and then she became a seamstress and sewed the curtains for the queen mary the old queen Mary, which sailed across the Atlantic. Oh, wow. Would you describe the atmosphere in your house as intellectual when you were growing up?

Oh, yes and no. It was a mixture. Being a teacher, my father was certainly bringing books home and had his own little library and encouraged us to read books. He'd give us books from time to time. But the family was also very religious. My father had been brought up in a Plymouth Brethren family, which is extremely rigid. And you didn't read anything on Sundays, no sport, had to sit indoors, go to church three, four times a day.

My mother was a Methodist, and my father converted to being a Methodist on his marriage. Essentially, went to the Methodist church and eventually became a Methodist local preacher. After school, you initially wanted to study cosmology at university. What attracted you to cosmology? I think I used to lie in bed at night and worried about what could be beyond the edge of the universe. If there was an edge to the universe, what was the other side of it? And this was a time when there was a lot of interest in radio astronomy. There'd been a new radio telescope built in Jodrell Bank in England, and they were getting radio signals back from outer space.

So there was a lot of interest, and it was something to which I could apply mathematics, which was my main subject at school. And I thought it was a lot of fun. But the great drawback was that I discovered that at Cambridge, you couldn't study cosmology without really doing three years of natural science first. And in order to do that, you had to be doing experiments in the laboratory. And the one thing I was absolutely hopeless at was the practical side of doing experiments. You know, I dropped the pipette, couldn't really identify chemicals when I sort of mix them together. I've always come out top on the theory, but when it came to the practical side, I was close to the bottom. And I just decided that three years of laboratory experiments was not for me.

So I then decided to do something else to which I could apply my mathematics, and that was economics. And I'd become interested in policy and politics. My father had started off in secondary schools teaching geography and then moved on to teach economic geography. So there are a number of books around on that subject. So I became an economist. It's not a ringing endorsement for the economics profession that you chose the field because you weren't practical. Well, of course, you don't have to do physical experiments in economics. It's all in the mind.

And even if you do, there's a growth now of experimental economics, but people don't have to carry around bits of equipment to get them to work. Why did you pick Cambridge? I think because I had, well, there were two reasons. One is that I read in a newspaper about a year before I applied and had to make a decision about a computer model of the UK economy, which was being carried out by Professor Richard Stone and his colleagues in Cambridge in the Department of Applied Economics and it sounded absolutely fascinating the fact that you could capture what was going on in the economy in terms of a

mathematical model that you could simulate on a computer using real numbers. And I was intrigued by that. And I had also read about Maynard Keynes, obviously. So I don't quite know what came over me, but I wrote to King's College Cambridge, which was Maynard Keynes' old college, and said that I was studying mathematics, was interested in doing economics. Would they have any interest?

And they encouraged me to apply. And the school was a state grammar school, but it had a long tradition of sending people to Oxford and Cambridge. Remember in those days, this was the mid-60s, that only 8% of each cohort went to university at all, unlike the near 50% now. So this was a special thing to go to university, and Cambridge expressed some interest. I wanted to go to Keynes' old college which was also as it turned out the college of

Richard Stone who was running this model in the Department of Applied Economics. So I applied and took the exams and went up for interview in Cambridge. That was an interesting experience. I had to travel on a Sunday I think it was across country from Wolverhampton where we lived to Cambridge arriving on a sort of dark December evening um show my room was dinner in hall with a handful of other people who were also there for interview um and then we were interviewed the next morning and then we went back and then one day I got a an envelope it's one of the best days of my life. A letter came to home saying I'd been accepted, and I remember walking to school feeling on an absolute high that I had got into King's College Cambridge,

and that was one of the key turning points of my life without any doubt. The funny thing is that three years later on, when I graduated from Cambridge, I was living in a nice set of rooms in the courtyard by the river at King's College. And one afternoon, I came back and found an envelope had been delivered under my door and opened it. And it was something rather odd on one side of the paper, turned it over, and there was a letter from the head of the Department of Applied Economics offering me a job and to work on Richard Stone's computer model of the economy.

So having decided initially to go to Warwick University to do a master's degree, together with Oliver Hart, my close friend at King's College, Cambridge. I stayed in Cambridge and joined the Department of Applied Economics. Oliver went off to Warwick and then to Princeton for a PhD. And now he's at Harvard and he won the Nobel Prize three years ago, four years ago. Was Keynes' long shadow still looming over King's College when you joined? Yes, and over the whole of Cambridge. I think one of the rather sad things about

Cambridge economics, the way it went, was that in the 60s, there were two distinct groups, one trying to maintain the tradition of Keynes and the other wanting to move into the more modern mathematical era. And I think the sad thing was that they didn't really find a way of combining the best of both. It may have been very difficult but I think it meant that Cambridge was held back in terms of its development of economics as a subject. On the other hand many of the things that the people who were Keynes's followers believed in I subsequently came to believe in strongly myself so I think that you know there were the people who worked with Keynes when he wrote the general theory people like Joan Robinson Richard Kahn even Mead, who came to Cambridge for a year from Oxford.

James had then subsequently moved to Cambridge as professor when I was there. They were all there in Cambridge. And half of them were keeping the Keynes tradition going and the other half were trying to move into the more modern era. So it led to a very divided faculty, which was unfortunate. And people would knock on your door and instead of saying that what are you working on today they'd say which way you're going to vote in the faculty election so it became highly political and I was I started as a graduate I was working in the department of applied economics as a graduate student um paid a salary to work as a research officer and then i went to harvard as a kennedy

scholar and that was another key turning point because then i really learned what modern economics was about and it didn't have to be anti-kensian or pro-kensian it was how does the world work and that was a key turning point because the contacts and the friendships I made there, particularly with Martin Feldstein, who subsequently became the president of the National Bureau of Economic Research, which he took from New York to Cambridge, Massachusetts, and turned it into a most fantastic think tank for economics, full of faculty, graduate students mixing. You know, when the cinemas were empty at night, people would flood into the National Bureau

and start working on their computer work all night. That was a great experience. And I went back to Cambridge as a faculty member then for five years, but it was a highly political atmosphere and not a terribly pleasant one. So I then moved on to my first chair at Birmingham University at the age of 28. And then I taught at Harvard and MIT for a couple of years and then back to the London School of Economics. I just want to come back to Keynes momentarily. I don't know about you, but I found the general theory to be generally incomprehensible, except for a few parts which

I find deeply insightful. But there's one sentence at the beginning which is enigmatic, where Keynes writes that the difficulty lies not so much in developing new ideas as in escaping from old ones. What did he mean by that sentence? Well, I think it's a very important reflection on the difficulty of changing people's views about the intellectual framework that they have grown up with and have been using as the basis for their own research and teaching. And I found this to be exactly the same when John Kay and I were writing

our book on radical uncertainty, which was published recently. And I think what Keynes meant was that in the 1930s, when he was trying to talk about the Great Depression and explain why there could be mass unemployment for a long period of time. What he found was that the rest of the profession had adopted the view that really if a market doesn't clear and the labor market is a market, then there must be some impediment to the speed at which prices, in this case the wage rate, can change. And that if only we could find a way to reduce some of these impediments to markets clearing quickly then we would never find ourselves with a Great Depression.

And Keynes was saying this was completely wrong and Keynes was right because what we discovered later on in the early 1950s when people like Ken Arrow and Gerard Dubreuil were writing highly mathematical work about the conditions under which a market economy would clear and produce an efficient allocation of resources including full employment what Arrow and Dubreuil discovered was that you know how do you cope with markets for goods and services that will be produced in the future? So expectations about the future are of fundamental importance

in driving the economy. Now, what our own de Broglie did was to say, well, let's imagine that we could buy and sell today in markets for future possible goods and services, including goods and services contingent on various possible outcomes. And this was an important theoretical description, although of course it's utterly unrealistic. We couldn't possibly have markets in all goods and services under every possible outcome in the future.

We don't even have futures markets in oil, one of the big commodities, for more than a certain number of years. So the question of what happens if these markets just don't exist is of fundamental importance in driving what's happening to the economy today so the economics profession in the 30s which basically took the view look all the markets that are open today if they all clear then we won't have unemployment there must be some friction or impediment preventing us kane said no and he didn't really put it in this language because if he had been able to,

I think he would have found more ready converts. And as you said, his book would have been easier to understand. But basically, what he was arguing was, since these markets for future goods and services don't exist, expectations about the future are a critical driving force in the economy. So if you cut the wage of people who are working today, the conventional analysis at the time said, oh, that means that employers will have more incentive to hire workers, unemployment will go away. What Keynes pointed out was that if the reaction of workers is to interpret that wage cut as a signal that all their incomes in the future will be lower than they had previously expected, then actually they will cut spending not just today, but they will also plan to save a bit less for the future.

But one of the consequences is they'll spend less today so you get into this downward spiral and only changes and expectations and improvement in animal spirits can prevent that and that's why Keynes recommended government spending at a time when animal spirits was particularly were particularly low and thinking about it in that way is at one level obvious, but actually it was inconsistent with the way the economics profession thought about the world in the 30s because they didn't have a framework which enabled them naturally to think about markets in future goods and services under all possible contingencies that had to come until the 1950s and this i think was a fundamental importance and that's where i you know it's quite interesting that the economists

today who have been brought up in the tradition of the almost of the tradition of that framework in the 1930s still think that the big contribution of Keynes is to argue for government spending when there's a downturn as a means of overcoming frictions in the labor market of all kinds. Whereas I think the real contribution of Keynes was to say, even if there were no frictions in markets open today, you could still get unemployment if people were so pessimistic about the future that they are overcautious and don't spend. And prices moving aren't really going to change that. And I think this goes to the heart of the debate about negative interest rates today, when there are people who believe if only the interest rate can be more negative,

then people will be bound to spend more and save less. But not if that changes people's expectations about the future and about the policy framework, if they lose faith in what central banks are doing. I guess the other interesting piece of the puzzle goes back to Irving Fisher and his theory of debt deflation when we're thinking about financial instability. And in 1994, you gave a presidential address to the European Economic Association titled Debt Deflation Theory and Evidence. Can you just give me an outline of the argument you made in that address and why it was important?

So we had been through a number of recessions in the post-war period. And in Europe, and particularly in the UK, we had deep recessions in the early 80s and again, the early 90s. And the question was, why? What was driving them? Why were they so deep? And I think conventionally the idea was that if people had become rather pessimistic, animal spirits were low, then businesses would invest less and therefore you'd expect the business cycle to be driven by investment. When the economy turned down it's because investment had fallen off and when the economy turned down, it's because investment had fallen off. And when the economy picked up, businesses had become more optimistic and were investing again to produce for the future. But consumption would be relatively stable. And it seemed to me that actually the data suggested that in the early 90s, the causes of the depth of the recession were much more to do with weak consumption. And the question then was why?

And again I went back and looked at what Fisher had written on debt deflation and it did seem to me that when debt levels were very high then households could become super cautious if they felt that they were moving to a point when they would be unable to service their debt. And that would lead them to want to cut their consumption, to give a margin of income available to ensure they could service their debts rather than suffer the penalty of falling into the position in which their credit ratings would decline and deteriorate or their homes

might be taken away from them. And I wanted to construct models to look at this. And it became clear that to do so, you really had to move away from the idea that the economy could be explained by changes in the behavior of an average household and instead you had to look at the behavior of different types of households and that was one of the main conclusions from the paper that macroeconomics could not really sensibly be divorced from the analysis of different types of households interacting with each other. So the paper essentially was to construct arguments about debt deflation, which would explain why consumption could be much more sensitive than conventional pre-existing theories

could predict when you got close to the boundary where debt levels were very high. And of course, debt levels are even higher now. So I think these things we will see depending on which country you're looking at and which type, which group of consumers, whether these factors will have a role to play in the next three or four years, because debt levels today are even higher than we experienced in 2007 before the financial crisis, and higher than in the early 90s when I was writing that lecture. Yeah, and at the moment, globally, you're more concerned about corporate and, to an extent, sovereign debt.

What about for a country like Australia that has the second highest household debt levels in the world? So I think the debt levels we will see causing problems in the next few years will vary from country to country according to where the debt was before the COVID-19 crisis. I think COVID-19 itself has not led to a significant rise in aggregate household debt, though it certainly led to a rise in debt for the self-employed small businesses. But I think for countries which have very high levels of household debt,

this is something that will have to be looked at very carefully for two reasons. One is that people may have taken on debt in the last five to ten years thinking they could service the debt with interest rates at such low levels and if interest rates were to rise then even a very small rise in interest rates could cause problems for that group of households. And the second reason we could be concerned is that there is a risk, I'll put it no more than that, but there is a risk that over the next decade, inflation and hence nominal interest rates will be much higher than we have got used to in the last sort of 15, 20 years.

And if that were to come about, then that would cause potential debt service problems for a much broader range of households. Mervyn, I want to go back to when you were teaching at MIT and sharing an office with Ben Bernanke, who would go on to become the chair of the Federal Reserve in the United States. And what was Ben like back then? And did either of you have any inkling that you would go on to become the head of the central banks? So we had adjoining offices. We didn't share an office. We had adjoining offices. We didn't share an office. We had adjoining offices.

Ben was on leave for a year from Princeton, and I was on leave from Birmingham. And MIT had lost – well, the people teaching public finance had in large part retired. And the idea was I'd go for a year as the senior public finance person, and Jim Perturba would go as the junior public finance faculty member, and together we would try to regenerate public finance at MIT. We had a wonderful seminar series with meals at a Chinese restaurant afterwards. And so I spent one year there, and Jim Jim Paterba took over and still runs the public finance at MIT to this day so Ben was as he always has been actually quiet thoughtful rather introspective doesn't make a lot of noise until he's got something worth saying then he

speaks but if you had asked either of us whether we would be head of the central bank i think the answer would have been clearly no certainly in my case i wasn't even teaching monetary economics and um i think it's a great shame that we didn't together go to a local bookmaker and try and place a hundred thousand dollars or something at odds of a million to one on on this outcome uh so then we could have had you know pleasant retirements when when we left in spending the proceeds of this extraordinary bet that came off no it wouldn't but you see it wouldn't have occurred to us to even to make the bet that's the point it wasn't that we thought about it and said oh we can't get good enough odds. It would never have crossed our mind that we would both have

ended up as respective heads of our central banks working together to deal with the crisis. Yeah. So you began as governor of the Bank of England in 2003. When the global financial crisis eventually struck in 2008, were you surprised by it? Were you caught off guard by it? go back. I'd been at the bank from 1991 as first chief economist and then deputy governor, putting in place the new inflation targeting framework, then became governor 2003. My very first speech as governor, I talked about the nice decade, the non-inflationary, consistently expansionary decade. And the point of the speech was it couldn't go on. That something, you know, this was not a situation that could continue. So I think we all knew that the macroeconomic picture

in the world economy had to come to an end at some point. These imbalances, low saving rates, high saving rates, trade surpluses, trade deficits, simply could not continue indefinitely. But of course that doesn't enable you to know when it will come to an end or indeed how it will come to an end. We had various seminars at the IMF which suggested, there was a seminar which got lots of leading economists together and the majority view in that was that the unsustainability

would end with a sharp fall in the US dollar. What no one imagined was that it would end with a collapse of the US banking system. Then we moved ahead and in 2007 it was very clear that there were some serious problems in the financial sector. I talked to a number of people, not just in the bank, but academics and others about the complex financial instruments that were being created. I gave a speech in June of that year precisely about the dangerous nature of these complex instruments, saying

that rather like a bottle of champagne, it may say AAA on the bottle, but when you open it, it may be flat. And that all previous financial crises have been characterized by excessive leverage. And I remember saying that the last line of it was, why do we think we're so much wiser than the financiers of the past? And there was booing from some of the audience at the Mansion House at that event. Of course, only a few months later, they were all completely convinced that they knew this was going to happen.

So then we went into the crisis. For the UK, it was dominated initially by Northern Rock. the day in August 2007 when the BNP Paribas, the French bank, announced that it would end redemptions from three of its funds. In itself, that was not a major event, but it was a big signal to the world, and you could start to see jitters in financial markets spreading. But it didn't lead straight to an immediate crisis. It led to sharp rises in interbank interest rates. People didn't want to lend to banks at all because they didn't know which were long and which were short

of instruments whose value had fallen. And really for 12 months, this waxed and waned. And there were months when the LIBOR spreads rose sharply, months when they narrowed sharply. You know, there was hope that we might get through that that was a't too many losses, but nevertheless, you weren't sure which banks were solvent or not, then the rational thing to do was not to lend to any of them and just to lend to the government for six months until things had calmed down. So these liquidity problems weren't going away without some further action and we had

to recapitalize the banks. And right through 2008, I worked hard trying to ensure that we had a recapitalization of the banks in the UK. In the end, we did it. We were the first country to do it. The Americans followed shortly thereafter when they diverted the money that had been provided by Congress for the so-called TARP scheme, Troubled Asset Relief Program. And then they put in place the stress tests and the money was there to ensure that all banks, either through their own efforts or through intervention by government, would be recapitalized.

And I think, you know, I would say the banking crisis ended in May 2009, when the US Treasury announced the results of their stress test, and the resulting recapitalization of all the major US banks. And that, I think, was the turning point of the banking crisis. The underlying imbalances, which had never really gone away, came back again. And that's why we had a decade of stagnation, I think. But I think the big shock of the serious collapse of the US banks came as a terrible jolt to everyone. Day by day, you could see the problems emerging. But somehow, it was still a great shock when, as Ben Bernanke put it, all the major U.S. financial institutions would have failed without government intervention.

And that illustrated, I think, A, how fragile a banking system is normally, and B, that most times it it carries on but it's potentially fragile and then when the moment comes when people lose confidence the whole thing can collapse extremely quickly a lot of finance types like to sagely quote Hyman Minsky, that stability leads to instability. Do you agree with that insight? I think it's too simple. I think it implies a sort of rather mechanistic causal relationship that you must oscillate from stability then to instability and then back again as if it's a regular cycle. And I just don't think the world is like that. All attempts to identify cycles in economic activity,

if you look backwards, you can always apparently identify a cyclical pattern, which seems to be a regular one. Because for any given set of data from the past, there is some relationship that will fit it. Just as an aside, I've never forgotten a paper almost in jest that Bob Hall produced saying that, you know, who are all these people saying you can't predict the stock market? I can predict it very well. And what he did was to say, look, here's the stock market for the last 30 years, and here's some regressions.

I can explain 80% of the variance of the stock market with these variables. And of course, he could, because looking backwards, there was always some correlation exposed that happens to fit that particular configuration of data. But of course, when you come then to predict in real time for the future, the whole thing breaks down. This is not a stable relationship. It's something that just happened to be a pattern for that period in the past. And I think that's true of the attempt to identify cycles. So there is certainly a key element of Minsky in the sense that beliefs and expectations matter. And if those beliefs become ones where people are very optimistic and then something happens to mean that that optimism is not justified, then you can see a big change in behavior and a change in asset values. And if people have borrowed a lot against those higher asset values, then you can get, as I mentioned before,

when we were discussing debt deflation, debt itself, high levels of debt can act as a dampener on spending. So you can certainly see movements and changes. But I don't think they follow a regular pattern. I think they're completely unpredictable. And that's why financial crisis can't easily be predicted. People go backwards and just identify patterns in data. But I don't think they necessarily give us a good feel for what will happen in the future.

Yeah. You remind me of a story about Jeremy Grantham, the billionaire investor. He once joked about John Templeton's famous phrase that the four most dangerous words in the English language are, this time is never different because who are we to assume that the laws of economics are stable and that regression to the mean is like a fundamental law of the universe when it applies to some sort of economic metric. Absolutely. Yeah, nothing is preordained or set in stone. In your book, The End of Alchemy, which was first published in 2016, you argued that people were relying on something you coined as the stability heuristic. What is that?

Well, it's an attempt to get away from the Minsky view that there is this mechanistic link from stability to instability. It's a way in which people grapple with an uncertain future and they have rules of thumb which seem to work and you stick to those rules of thumb until they stop working. So I think in the run-up to the financial crisis, what was fairly clear was that, as I said at the IMF meetings, economists said that these imbalances can't go on forever. But people living their ordinary lives didn't go home and say to each other, you know, darling, I'm very worried. Britain's got a big trade deficit. This can't go on forever. some point something will change the price level will go up if we pay more for the imports or we'll have higher taxes something has got to change in the

future so we shouldn't spend quite so much today and instead what people did was to say well you know how much can we spend uh consistent with the world carrying on as it is. So that we were spending collectively more than our means, but it seemed to be something that was stable and we could carry on doing. The fact that we were building up large stocks of overseas debts was not something that affected individual behavior. The rule of thumb that seemed to work was we could carry on spending at this rate. And something would turn up in the future to vindicate that. Come the crash, it became very clear that what we've been spending was more than we could sensibly sustain. So the rule of thumb changed. And I think this low consumption, not just to fall, but to stay low and remain weak.

And so we've had, you know, what Larry Summers calls secular stagnation, what I would call is, you know, it's a low demand equilibrium. And the heuristic has altered. It's that the narrative that we tell ourselves has changed. And it's a narrative revision downturn that we experienced during and after the financial crisis. And it's now turned out into a longer period of just really low growth. And it's not going to change, I think, until the narrative about how our world economy operates changes. And that's gonna require governments to recognize that the levels of saving rates in some countries

are unsustainably high in countries like Germany and China and unsustainably low saving rates in countries like the United States and Britain. And until we take this on board and do something about it, then I don't think we're going to get the corresponding narrative revisions for households to change their levels of consumption. German and Chinese households need to consume more. British and American households need to consume less relative to their national income. I love your idea of a narrative revision downturn. I think there's a lovely

example of it in Don Watson's book, Recollections of a Bleeding Heart. Don was Paul Keating's speechwriter, Paul Keating, the Australian Prime Minister, and the book was based on diaries Don kept in Keating's office. And they talk the 1991 recession and Don quotes Don Russell who I think was Keating's senior economic advisor at that time as saying they heard the economy snap like a neck and I thought that was a really vivid description of what you would call a narrative revision downturn. There was something palpable that had shifted in the public mood. Incidentally, you know, Keating, you were going to come out to Australia before the pandemic ruined those plans, sadly. But how do you know

Paul? How did you first meet him? So, Paul became a great statesman-like figure after he left office and joined a number of international meetings where he discussed politics and economics. And he had come to meet Martin Feldstein, who was my mentor, who ran the National Bureau of Economic Research, and with whom I had carried on working and being very close friends with for a long time and one day when I was at the Bank of England Martin sent me an email and said look Paul Keating is coming to London I think you'd like to meet him you know set up the meeting and Paul got in touch with my office and we met at the bank and that started a series of meetings that paul and i both enjoyed very much i one of the most interesting people i have ever met is paul keating and uh very astute views on what is going on in the world what is happening to politics, why things have changed. And extraordinary that someone who had rather little formal education but became so learned and so artistic.

And I have enjoyed my contacts with contacts with paul ever since the house that he has created in sydney is the world's best example of what a french empire house would have looked like and he is stunningly beautiful you know furniture artifacts the floors, everything has been done to make it look perfect. And it is absolutely, it's better than anything in France, and Australia should be very proud of it. It is wonderful. But Paul's view on what's going on in the world was always a view that I wanted to hear and listen to.

And I was very struck when, you know, he would tell me, you know, he'd work on his red boxes, whatever colour they are in Australia, on Saturday afternoons, listening to classical music. And the music would create a sense of imagination and excitement about the new policy ideas that could be created, discussed, put in place. A remarkable man. Do pass on my best wishes. I think I will do if I bump into him up in Potts Point in the street there.

And I think he's a big fan of Mahler amongst others. But you're a classical music fan as well, if I'm not mistaken. Yes. And this September, I become the chair of the Philharmonia Orchestra, based in London, but plays around the world. I hope one day we can take the orchestra all the way to Australia. Find a suitable sponsor, we'll do it. Sadly, at present, I'm becoming chair of an orchestra that can't

perform at all. But we're beginning to put on some performances with a subset of the orchestra filmed in a recording studio and then streamed. So you'll be able to see in Australia the Philharmonia performing. You can get it on YouTube.

Brilliant. Well, I'll look that up. And if you do come to Sydney, I'll be there in the front row. Well, I very much hope I can. Goodness knows when it's going to be possible to travel again. Just absolutely unknown. That's a real example of radical uncertainty. There's no point speculating or putting odds on the chances

of coming on a particular date. We simply don't know. So speaking of radical uncertainty, you have a new book out with John Kay called Radical Uncertainty. And I enjoyed it immensely. It's sort of like a capstone for me for a lot of ideas that I've been thinking about for the last few years. I think ideas that I probably originally became interested in via Nassim Taleb and his inserto. And I'd like you,

Mervyn, just to take as long as you like to give us a sense of the content of what you mean by radical uncertainty. So radical uncertainty is any kind of uncertainty which can't be quantified. There can even be uncertainty about the present or in some cases the past. We don't know something, but typically in the past we can look it up. There are mysteries about the past, things we will never really know about the past. There's no point saying, you know, it's a 30% chance that the dinosaurs disappeared for one reason or 60% for another theory, if we can't ever discover the truth of it.

The real essence is when you look forward. And let me give an example, a typical example of what I mean by radical uncertainty. COVID-19 was not something that we had expected. It hit us as a surprise. But in a way, the fact that there was an epidemic shouldn't have been a surprise. We actually wrote in the book, and we wrote this last summer, that we should expect to be hit by an epidemic of an infectious disease resulting from a virus that does not yet exist because we knew that pandemics existed. But the fact that you knew that there were things called pandemics did not enable you to make a statement of the kind, well, I think the probability

that a virus will emerge from Wuhan in China in December 2019 is 32% or 3% or 80%, that statement would have been a meaningless statement. You could say the fact that it may come in Wuhan in December 2019 is not very likely because in any one city in any one month, it's not very likely. But it is likely that at some point there will be a pandemic and therefore we should prepare for it but quantifying it doesn't really add anything and it can be very misleading the best example in some ways is what happened to president obama when he had to make the decision about whether to send in the Navy SEALs to the compound

in Abbottabad in Pakistan to capture Osama bin Laden. They knew there was someone living in the compound. There was a man with family around him living there. They didn't know whether it was bin Laden or not. So they had a meeting in the White House to decide what to do, and the CIA provided briefing about whether this was bin Laden or not. Now, after the intelligence failures in Iraq in 2003, Congress told the CIA that when they briefed the President of the United States, they had to put their judgments in a probabilistic form. So the agents came in to see Obama, and he said, well, is it bin Laden?

And the first agent said, yeah, I think it's 90% that it's bin Laden. The next one said, I think it's 60%. Someone else said 30%. Someone else said 30 percent. Someone else said 70 percent. Obama said in an interview after this meeting, he said, the fact that these people were giving me probabilities was concealing the truth and not helping me think about it. What I wanted to know was some information about, you know, what did they, what was the basis of their judgment? What was the information they were using?

I wanted a narrative from them as to why they were either confident or not confident. And it didn't help to exaggerate the degree of precision of that confidence by attaching a number of probability to it. It was just confusing everything. And in the end, the only answer to the question, is Bin Laden there or not, was, we don't know. And that is the answer to many, many big questions that we confront. You know, will people buy my new product, which I love, I've invented it, but when I come to sell it to the rest of the world, will people want to buy it? We don't know. And there's no way of knowing the answer to that.

You can think about strategies for getting information. And it may be that you can think about the issue in a way that tells you what information to collect. But it doesn't help to try and pretend you can quantify this. And yet, around the policymaking world and around the business world, it is felt that there have to be consultants used or experts employed to calculate via spreadsheets or some other mechanism a quantitative estimate for whether this decision should go ahead or not. And very often, we have no basis for that quantification. That doesn't mean to say that numbers aren't important. They are. You can find numbers that help us understand what is going on. But what you don't do is to

hand it over to a consultant or an expert who's got some black box model where they are forced to make up numbers in order to get out the final number at the end which if it's positive means go ahead and if it's negative means don't do it and these numbers are meaningless typically and they lead to very bad decisions so radical uncertainty is much more than Nassim Taleb's black swans. Black swans are a limiting case of radical uncertainty, the case when you literally can't imagine what might happen in the future. So the example he gave was that it gave rise to the description black swans. The people who first went to Australia could not possibly know that there were black swans because

the only swans they had ever seen were white so a swan in their view was always white so the idea that they might sit on the deck of the ship traveling slowly to Australia betting each other bet you three to one there's a black swan inney when we get there you know it just it was to make absolutely no sense it was not something that people could ever do and the but radical uncertainty is much more than that because there are things that we do know there are pandemics there are banking crises but that doesn't mean that you can attach a probability to a crisis occurring at a particular date or in a space in the future. You can't possibly know that. And it really matters because if you can't quantify it, then you can't rely on traditional cost-benefit analysis where you have to come up with numbers for everything. What you need to do is to just ask the question always, what is going on here? And we stress that in the book a lot. What is going on here? And you have to tell a story about what's going on to reach a judgment at the end of the day as to what the right decision to make is. And if you do that, you can certainly use models of an abstract kind which may give you a feel for the nature of the problem you're facing

or numbers that tell you you know this number suggests that the cost of engaging in a project is going to be truly enormous so why do we think the benefits hard though they are to quantify do we really believe that the benefits could possibly justify a cost of this magnitude? Sometimes relatively small pieces of analysis can actually help you understand what's going on in a way that the black box model doesn't. And I think my favorite example of that in a way was came from a great Australian, Bob May, Robert May, great scientist. I mean, he did everything, physics, mathematics, biology, and was at Princeton, then Oxford, government chief scientist

in the UK. And he and Roy Anderson, well, this is Bob May, they wrote the textbook on infectious diseases. So all the models that are being used are derivative of their work. But Bob May was asked by the World Health Organization to go out to Southern Africa when AIDS took off. And he was asked, you know, we've built this very impressive model of population movements and dynamics in Southern Africa, linking together demographic models of the different countries in Southern Africa together. So we want you, we've done this because we want to know

how quickly AIDS is going to spread in Southern Africa. And we want you to look at the model and tell us whether you think it's a good model or not. So Bob May went out there, and a massive great model, lots of detail on different age groups and so on, different demographics. But he could see that one of the parameters in the model

that was absolutely crucial was the average number of sexual contacts per person per year in a country. And he looked at them and said, you know, this number is fundamental. And don't you realize when you think about it, when you ask the question, what's going on here? If the answer to the average number of sexual contacts is, for example, 100 per year, it makes a hell of a difference whether it's 100 contacts with the same person or 100 contacts with 100 different people. Because if it's the latter, the disease will spread very rapidly across Southern Africa.

And if it's the former, it won't spread at all. So forget your complex black box model. What you need to know is something about the pattern of sexual contacts in the population. And when they did that, they discovered that the disease obviously was spread by lorry drivers driving around Southern Africa, having large numbers of contacts at each place they went. And so that was the key parameter. And that was driving the result. And so the model itself was far too complex. What was really needed was someone to ask the question, what's going on here?

How is the disease being spread? Who is spreading it? Is it being spread by some groups rather than other groups? If so, what should we do about it? And that is also, of course, true with thinking about COVID-19. And the models that May and Anderson constructed were very good for teaching purposes because they showed you the shape of an epidemic. It takes off slowly, suddenly accelerates away, reaches a peak and comes back down again. But you can't use it to make predictions unless you've got good information about the key parameters. And they give examples from data which were collected from previous epidemics of various

kinds, both animal and human, to show how to use the models. But the trouble when it comes to COVID-19 was that there were key parameters that we simply didn't know about. We didn't know the fatality rate from the virus. We didn't know how quickly it was spread, that we learned about that. The virus occasionally mutated. We certainly didn't know the human response to the measures that might be taken of a lockdown kind, and we certainly didn't know how people would react when those lockdown measures were relaxed and people could, in principle, go back out again. Would they or would they not? But all of these parameters were fundamentally crucial to making quantitative predictions. So the models in many ways were misused because people used them to make predictions. People then said, in the UK, the government said,

we're doing this because this is what the science tells us we must do. But the science doesn't tell us what we must do. It helps to inform our judgment. And the judgment which governments are supposed to take is partly to understand the uncertainty behind the models, but also partly to recognize that there are issues other than the spread of the disease that need to be taken into account. And this current crisis is one in which governments have the unenviable task of navigating between two very risky sets of rocks. One is the rocks created by the death rate from COVID-19

and the other are the rocks caused by the enormous economic cost of shutting down the economy. And both of these are very costly. And really, since we don't know very much about either, the process that we are going through always had to be one of trial and error. And I think governments would have done better to have been open about that and said, look, we don't really know enough about the virus to be confident as to what will happen. We don't really know about the size of the economic cost, but we're going to try various things. And if we learn, as we will as we go through it, then we can adjust our policy course. And if we learn, we get to a point where

we think infections aren't rising very quickly, we can reopen the economy. And if there are localized outbreaks, as you've been experiencing in Melbourne, then you can shut down that one part of the economy again. And in a way, over time, learn more about this to make a better policy trade-off between these two competing, highly undesirable outcomes. And that's the way most decisions are made in practice. We know something, but we never know enough to be able to quantify the results and hand it to some expert to tell us what we must do. That, in essence, is what radical uncertainty means. I want to talk about technology as a source of radical uncertainty. Last year, I had your friend David Tuckett on the podcast and I quoted Karl Popper's lovely little syllogism

from the preface of his book, The Poverty of Historicism, where Popper writes something like, the course of history is largely determined by new technologies. We can't predict the technologies of the future since if we could, we would already have invented them. Therefore, the future is fundamentally indeterminate. And David said that actually you can think of this as like the flip side of free will,

where if people are allowed to come up with new ideas and freely collaborate they'll produce technologies which will be unpredictable and i've been reflecting on that a lot recently in preparation for this conversation with you mervin and at the moment i think that technology gives rise to radical uncertainty in three different ways. Firstly, there's this lovely section in Nassim Taleb's book, Anti-Fragile. I think the chapter is called History is Written by Losers or something like that. And it's like an attack on the teleological view that research departments in large corporations or at universities is what drives fundamentally science and innovation. And Taleb makes the case that actually the history of technology

is the history of bottom-up non-teleological tinkering by practitioners and amateurs and that it's a story of trial and error and he I think he mounts a like a revisionist history of the industrial revolution as like you know one of the main examples in his argument and that got me thinking that if technology is fundamentally a bottom-up phenomenon most of the time you just never know what's going to pop up. Even if you could tabulate every project being worked on in the proverbial garage of every entrepreneur around the world right at this minute, that still wouldn't give you a good handle on the future for the two further reasons.

The first is that technologies drift or find new applications. And your co-author, John Kay, has a brilliant little book called Obliquity, which is a term he coins to describe how an innovation can find new applications and and like a classic example is uh viagra which was originally uh intended to cure hypertension and anginas until during the research trial um the experimenters found that it had a very surprising side effect aspirin is another example, I suppose. So, technologies can kind of find new applications which we can't predict. And then I guess the third thing is coming back to David's point about free will collaboration

or what Matt Ridley calls ideas having sex. You know, every innovation is essentially just a recombination of pre-existing innovations. Entrepreneurs meet and speak with each other. Scientists meet and speak with each other and they come up with new innovations. And that means that there's a super additive function, which is by definition unpredictable since we can't predict all of the possible combinations that might arise. Do you want to react to any of that? I've just sort of spewed my thoughts out. No, I mean, I agree with all the points you've made. And indeed, when Frank Knight, you know, wrote his famous 1921 book, Risk, Uncertainty, and Profit. It was all about entrepreneurs being able to make a profit solely because they were producing something that hadn't existed before, whether it's a product or a process, etc. And, as you say, when this happens, then there are all kinds of consequences to this which can't be predicted.

So either the products can be used in uh day-to-day applications in our everyday life that we're not aware of that were produced as a by-product of sending people into space formula one motor racing has produced a whole series of things that are used in ordinary cars. So there is this drift, as you point out. So the two things I think are important coming out of this are one, that new ideas, new technology are what drives higher living standards. And they are inherently unpredictable. They are a classic example of radical uncertainty.

And if that is the case, it is a bit odd for the economics profession to try to sideline the existence of radical uncertainty when that is the sole reason we have increases in living standards instead of putting radical uncertainty at the heart of the economic research project. The second thing I'd say is the phrase that you quoted from Matt Ridley, ideas having sex, is incredibly important

because it's what distinguishes the human race from the animal kingdom, I think. If you take one human, they will be hard put to survive anywhere. Many animals can run faster than humans. There are animals that can fly faster than humans. There are animals that are stronger than humans. There are animals that can fly faster than humans. There are animals that are stronger than humans. There are even computers that can do mathematical calculations faster than humans. But what the human race can do is to collaborate with other humans to achieve things which are quite remarkable. So the example we give in the book is that even though humans can't fly, we can collaborate with large numbers of people around the globe, many of whom

we have never heard of or met, to produce an aeroplane that can fly from London to Sydney. And that's a remarkable achievement. It's collaboration which enables humans to create prosperity really. This is the Adam Smith's division of labor, the pin factory and so on. And you see it everywhere we look and I think think behind all of this is that the use of probabilistic calculus, which was invented in the 17th century to deal with problems of card games, applies to a very narrow range of phenomena. But radical uncertainty applies to most phenomena that affects both personal and business decisions

and underpins all decisions about the creation and adoption of new technology. And it should therefore be, as I said, put really right at the heart of the economic research project. Yeah, I completely agree with you, Mervyn. I think it's astonishing that the economics profession sidelined or overlooked the concept of radical uncertainty. But how did that even come about? So,

it's really post-Second World War phenomenon, although there was a key debate in the 1920s and early 30s among philosophers, mathematicians,

much of which took place in Cambridge, England, and indeed within the walls of King's College, Cambridge, where John Maynard Keynes was inclined to believe, along with Frank Knight in Chicago, in the importance of radical uncertainty. But a young, brilliant philosopher, Frank Ramsey, who died at the age of 26, and wrote two or three papers that spawned whole fields in economics, came up with the proposition that if individuals had subjective probabilities,

that is, their own view about the probability of an event occurring, and if they had these beliefs about probabilities defined over all possible future events, and if those beliefs obeyed the normal laws of probability, namely that if you have an exhaustive set of outcomes, the probabilities must add up to one, then you could then demonstrate that people would behave in such a way as to optimize their expected happiness or utility where you weight the utility they would get from a particular outcome by the probability of that event occurring and that knocked the winds out of the sail really of of Keynes and other people arguing that you couldn't attach probabilities and there was

radical uncertainty, so they didn't use that phrase. Keynes of course came back to, I mean, Frank Ramsey then died either 1929 or 1930. Keynes came back in 1936 with a general theory where he emphasized crucially the importance of radical uncertainty. But economists after the Second World War went back to Ramsey and said, look, this is great. If we can assume this, then we can assume that people will maximize their expected utility. That means we have something we can optimize. And if we can optimize a function, then we can calculate how people will react and respond to changes in tax rates, changes in prices. We have a genuine positive theory of human behavior based on

economics. And this was such a seductive idea that it took over the subject. And Milton Friedman basically asserted that people behaved as if they had probabilities defined over all these different events. No evidence was adduced to support that assertion, but nevertheless, he asserted it using the analogy with a billions player, saying that people who play billions don't understand Newtonian mechanics, but through long years of practice, they behave as if they did understand mechanics. Actually, all it shows is they practice playing billions a lot. And the as as if assumption became pretty dominant in economics until people started to realize that they were big flaws because we had nothing to say about

big crises the financial crisis you know bob lucas said the the greater and genuinely great american macroeconomist did many fascinating things in macroeconomics he argued that we solve the problem of macroeconomics we now had a deal with recessions he said that only a few years before the great financial crisis um and of course that was followed by the great recession we hadn't solved the problem at all because unexpected things can happen. And we weren't prepared for it. We hadn't made the financial system sufficiently resilient or robust. So radical uncertainty is really very important. But it's not something that people want to put in economic models because as one economist said to me but if we can't quantify uncertainty we can't attach probabilities what are my

graduate students going to do what what can they assume people optimize well people don't optimize they ask the question what's going on here and do the best in present circumstances that doesn't give you much by way of a general theory that you can use to make predictions. And in that sense, economics is much closer to engineering than it is to physics. You've got to solve with particular problems that you're confronted with. And in many ways, that's what Keynes' career was. His views did change. It's famous to point out that when the facts change, I change my mind. What do

you do, sir? But actually, it's not so much that the facts were changing as such, but the answer to the question, what's going on here, was changing. So we weren't facing the same phenomenon that we were facing 10 years before. So he looked for different kinds of answers. And that's what we have to do, I think, when confronted with particular challenges in the economic sphere, not to pretend that it's just another sequence of events described by a model and a probability distribution that's been generating the past. This could be something very different from the past, and we need to think about what it is. There's so much I want to ask you.

I just want to dwell on Frank Ramsey for a moment. I think I can see a book of his sitting there on your desk behind you. It's a book of essays by – well, there are two books here. One is a biography of frank ramsey that was produced really is that cheryl mizak's biography exactly but the other is the um the foundations of mathematics and other logical essays which were the essays papers by ramsey edited by braithwaite and published posthumously. This one here.

Exactly. Very impressive. So, I tried to read Truth and Probability, which is where these ideas have their genesis. He wrote it in 1926, but it was published posthumously in 1931 in the Braithwaite papers. And he was an extraordinary young man. So, he wrote Truth and Probability when he was 23 years old and effectively won, as you say, Mervyn, won the intellectual debate against his friend Maynard Keynes. And that's interesting in and of itself. And Cheryl Mizak's biography is great for people who are interested in Frank Ramsey. It's called A Sheer Excess of Powers, which I think is taken from a quote

by Schumpeter about Ramsey. And you, so while Ramsey won the intellectual argument for a while, and you point out the economics profession was willing to give him the debate because it assuaged their physics envy, you and John Kay argue that Ramsey was wrong and the reason he was wrong is because in real life we don't always have to take bets to use Warren Buffett's phrase we don't have to swing is that sort of the essence of your rejection of Ramsey's yesivist view? Yes. I think the attraction of Ramsey's argument was that if you assume that people have subjective probabilities, then you can very quickly get to a very attractive outcome in which you've got a theoretical apparatus that allows you to assume that people will maximize expected utility

and all kinds of wonderful models and ideas flow from it. The trouble is, and it's very often the case, you can see it with people writing economic papers today. It's the very first line of the paper, the very first assumption, because people don't have probabilities attached to all conceivable events. They simply do not bet. And there are very good reasons for that. That is, if people say well you know what odds

would you be prepared to bet on an outcome the rational response to most of this is i'm not going to bet at all because i suspect that you know far more about this outcome than i do and you will frame this bet in such a way that i may think i've got a logical probability, but actually there is something I don't understand about the nature of the bet, which means that I will get this wrong. And the example we'll be given in the book is, I think, a nice one, which is that it comes from the musical Guys and Dolls,

where the hero is talking to his best friend about gambling. They're being persuaded to bet on what a young woman will or will not do. And the hero says, I'm going to tell you a story that my father told me. My father told me when I was very young that he didn't have any money to send me to a good college like Harvard or Princeton but he was going to give me a good piece of advice which was worth far more than any degree from Harvard or Princeton he said this is the advice son, one day you're going to meet a man

and he's going to say look here's an unopened pack of cards, the seal is all around it you can see I bet you that I can get the queen of spades out of there and behind my ear and when i do that uh i'm going to squirt if it a cider in your ear and the hero says you do not take this bet my son you do not take this bet because if you do as sure as anything you'll have cider in your ear and that was a statement of the

bet looks a sure thing how could anyone open a pack of cards with a seal on it and get the queen of spades out and how can the queen of spades card shoot cider into your ear but you do not take that bet my son because that bet is being offered because there is some trick that the person offering it has to make sure that you will think the event has actually come true so So it doesn't make sense to take bets on a whole range of things that you don't understand and don't know about. And so there's a terrible philosophical flaw in the assumption that people have subjective probabilities which they can attach to every conceivable future event. And once you drop that assumption, then you're in a world of radical uncertainty, and none of the arguments that Ramsey used, or the arguments

that subsequent theorists used to justify the assumption that people maximize expected utility hold true. I was going to say, I'd rather a cider in the ear than in my mouth, but I'd take your point. And in this world of radical uncertainty, we make decisions according to narratives, which is a key part of the argument that you and John make in the book. And my concern is that when people hear the word narrative, they're going to think that you're being non-rigorous. It's a kind of a fuzzy idea. What have you learned about the science of narrative in your own research over the last however many years? I think it's a mistake to try to be too rigorous. So one of the essences we're trying to get across

is that in a world of radical uncertainty, you should be content with fuzzy language, like it is likely that. It is likely that there will be a pandemic at some point in the future. That is the best we can do. And to try to be more rigorous about it, it's very tempting, leads you down the path of quantification,

which is completely misleading and potentially very dangerous. So the fact that you can't be rigorous, the question is what does it mean to be rigorous very often it means to quantify and i think that's a mistake there are certain problems and decisions where you can be rigorous in the sense of thinking clearly about something asking the question what is going on here? Trying to understand the nature of the problem, get a feel for the relative orders of magnitude. That's being rigorous.

Being precise in that situation is not being rigorous. It's being misleading. It's spurious precision. Yeah. yeah but i guess my question is more about so narratives themselves don't need to be rigorous but we can study narrative in a rigorous way and is there any way of distinguishing between good narratives and bad narratives? Well, the obvious way is to interview people and talk to them and see whether there are internal contradictions in their narratives and then pointing out to them gently that perhaps there's a contradiction in your narrative. And most people at that point will say, well, yes,

I see the point now. Maybe I ought to rethink my narrative because you want a narrative to be internally consistent and coherent. In essence, that's what happens in a court of law, in a trial, that the two sides are presenting competing narratives and you're trying to find contradictions in the narrative of the other side. So I think that's the way one does it and this does go back to the question which economists often put to me which is well you know what do we do in a world of radical uncertainty what is research? Well it's many of the very abstract models that have been produced are actually very helpful in understanding problems. So whether it's David Ricardo on trade or George Akerlof on understanding

why the informational asymmetries between buyers and sellers may cause a market to break down. These are models which give you deep insights into the world. And so models can produce insights, but what they don't produce is a precise quantitative description of what will happen. And I think therefore what's important is to distinguish between economics in the work that it does to provide insights and in the way we should be involved in making decisions where we simply confront radical uncertainty about the future, where we need to ask the question, what is going on here, and not resort to spurious precision. Several months ago, you had lunch in New York with Nobel laureate and former guest of this podcast, Bob Schiller. And Bob has been thinking a lot about narratives too. He had a book out last year called Narrative Economics. What are the big points of difference between you and Bob when it comes to narratives? And do you think he's coming around to your view?

So I think the similarity is that we both give weight and importance to narratives as influencing behavior. I think the difference between us is that he often regards narratives as a failure of humans. That is, it's an example of a cognitive bias, that they're using a narrative which has been spread like an infectious disease, and it's irrational. So he would, I think, tend to describe the influence of narratives on economic behavior as examples of irrationality. We would describe them as examples of rationality, that is, if you don't live in a world in which you can maximize expected utility, the world's more complex than that, you confront radical uncertainty, that it's wholly rational to ask the question, what's going on here, to construct a narrative, and use that narrative to help you make your decision. And be prepared always to challenge the narrative and amend the narrative, revise the narrative when necessary.

But nevertheless, that's going to guide the decision. That is not at all irrational. It's wholly rational. And I think that's the major difference between us. Got it. What does it mean to be rational? Well, that I think is the essence of what our book is about. Got it. What does it mean to be rational? had such a pernicious influence because once you've adopted that, you can then use a series of what look like plausible axioms to generate the result that people will maximize expected

utility or something rather like that. In other words, optimizing behavior can be defined and that is what you mean by rational behavior. And anything that deviates from that is, by definition, irrational. I think it's highly misleading because, as we argue in the book, when you're in a world of radical uncertainty, those axioms don't apply. You cannot assume that it's rational, therefore, to maximize expected utility. And being rational is much, much more difficult. And all you can do

is to ask the question, you know, what is going on here? What am I confronted with in this decision? And you draw on your experience, your insights, which you may have taken from other people and the past to make a good decision, which is not one that you will ever say afterwards you know i really took the best decision i could take you said you know i took a pretty good decision it it was good enough to uh to cope with that situation and that's really what we're confronted with all the time and that's what being rational is what you know it's not rational to optimize in a world where optimizing is not the rational thing to do or can't be defined and i think this is a great mistake the universal use of optimizing behavior and economics is a classic example of looking under the lamp

in the light to find something. You search in a very narrow area because that's where the light is. That's where your axioms tell you you can optimize. But outside that, which is almost certainly where the thing you dropped is there somewhere, is a world of radical uncertainty to which optimizing behavior doesn't apply. Yeah. uncertainty to which optimizing behavior doesn't apply yeah yeah it reminds me of um herb simon's famous characterization of bounded rationality as a pair of scissors and one blade was uh cognition the other blade was environment um and and i guess in more modern times this is similar to

gigarensa's idea of ecological rationality. Yes. That we can't talk about rationality outside of the context of the real world in which people make decisions, which is a large world, not a small world, to use Savage's distinction. But I just want to throw a quote at you, Mervyn, a quote of Daniel Kahneman's. This is from the journal Behavior and Brain Sciences from 2000. So, quoting Kahneman,

all heuristics make us smart more often than not, end quote. I guess my problem is I don't really see what the key clash is in the great rationality debate um it seems to me that kahneman toversky always conceded that heuristics made us smart and that they were adaptive but occasionally they led us down the garden path and the critique of good gigarensa's school of thought is more like aesthetic it's saying you know you're giving heuristics a bad name people can't be that dumb um but i i just don't understand what the fundamental clash is so i think the fundamental clash is that behavioral economics still identifies rational behavior with optimizing behavior,

maximizing expected utility. And any deviation from that is regarded and described as a bias, a human bias, which in some way is weak and we ought to try and find ways of correcting for it.

So Kahneman, for example, a couple of years ago said that we should give as many decisions as possible to computers rather than humans, because computers don't suffer from the same biases that humans do. I think this is a terrible mistake, because many of the biases which are identified are not biases. They are certainly deviations from optimizing behavior. But in a world of radical uncertainty, that may well be entirely rational. Many of the biases that have been identified, there's not time to go through too many examples of them, but the example of the experiment in which people fail to notice.

People are asked to watch a short video of two teams passing a ball to each other, and the question is, at the end of this 60-second video, how many passes of the basketball between the players in white shirts, did you count? And at the end of it, people are asked the number, and most people get it pretty much right. And then they're asked the question, did you see the person in the gorilla suit walk across the screen? And more than half people say they didn't.

Then they're asked to watch the video again, and what they're shown is the video. And their astonishment midway through this video someone slowly walks across the screen wearing a big gorilla suit. How could you miss them? And this is described as a bias. We see what we want to see or we fail to notice key things. But it's not a bias because if the purpose of watching the video is to count the number of passes made, then that's all you're looking for. And one of the great talents of humans is the ability to blot out anything extraneous to the task at hand.

We congratulate people who are good at doing that, not tell them that they're biased. When Steve Smith scores a century, because he concentrates unbelievably hard on every single ball, which is why he never gets out, people don't say to him, did you notice that little plane fly over in the middle of your innings? Of course he didn't. He was completely focused on the cricket. And that's his great talent, an unbelievable ability to take each ball as it comes and to concentrate on each delivery, never misperceiving it. And the fact that he doesn't notice other things at the same time is a great source of strength, not weakness. So you can't define these biases

without knowing the context in which people are behaving. And I think the great weakness of behavioral economics is that it's all based on lots of artificial experiments with no genuine context against which to appraise behavior. Much of the behavior that's described as biased is, in fact, entirely rational. And humans are so good at that. How could it be that if we've identified, you know, well over 100 biases that humans are susceptible to, apparently, how can an animal, a creature with so many biases, turn out to be the dominant species on earth we have to be doing something right if we're the dominant species and of course what we're very good at doing is coping with completely unpredictable events things that we didn't imagine would happen we adapt we respond

we can we use our imagination we communicate with other. These are things which computers do not do. And I would not think it wise to hand the discussions about the future relationship between the United States and China in the South China Sea to a computer. They certainly can do mathematical calculations faster than humans can, but that's only a subset of the kind of decision-making that has to be made to survive and cope in the world in which we live. Yeah. I think the other problem with K species, in part the product of group selection and things which are seemingly irrational at the individual level might actually scale up to be collectively rational.

Yes. So we should never talk about rationality without referring to the scale. Yes. Given your flattering account of human rationality, where does this leave speculative bubbles? So it's never easy to identify speculative bubbles while they're going on, because there is always the argument that I'm confident I can get out before the bubble bursts. Therefore, it's rational to join it, because I will make a lot of money before it comes to an end. and people can construct a narrative where they say to themselves you know there is a speculative bubble but nevertheless i'm i'm more confident than other people that i can identify when's the right moment to get out so i will get in and i think a lot of people who work in hedge funds believe that for example they can i've seen examples where people have said to me,

I agree with you that a particular bank is in trouble and it can't easily survive, but we're very confident that we can manage to get out of it within three hours of any piece of information arising that would lead us to cast doubt on how long it can continue. So if you think you can get out early, it's rational to continue to be in it and the bubble will go on. A lot of people never go into a speculative bubble because they are nervous about the fact that it is a speculative bubble. So I know it's a question of narratives and eventually something happens that negates the validity of the narrative that some of the people believed in and then once you that happens then the thing can unwind fairly quickly so it's a question i think of what are the narratives that people are telling themselves or using in order to explain why they wish to hold those assets.

Yeah. So those narratives could be individually rational but collectively irrational. Yes. Yeah. Do you think... The point you make about collective rationality is a very, very important one. That Nassim Taleb has always emphasized the importance of survival.

And as a human race, we want to survive. And it may well be that there are certain things that we do, which are individually irrational, but collectively rational. So people who volunteered to fly planes in the Second World War and the Battle of Britain, their life expectancy was very short. You could argue that they were doing something that was individually irrational, but collectively highly rational. And we develop ethical standards in order to encourage people to behave in that sort of way. Yeah. I'm cautious that I don't want people listening to your ideas on radical uncertainty to leave

this podcast episode feeling nihilistic. How would you encourage the average person to harness radical uncertainty and use it for their benefit? Well, there are two ways of doing it. One is that think carefully about the distinction between risk and uncertainty. Risks are bad things. So try to imagine what things could derail the plan of your life or the plan for the next six months.

So what could go wrong in terms of, let me give you a trivial example. If you're going on a foreign holiday, what could go wrong? Well, it will make absolutely no sense for you to leave it to the last minute to go to the airport because if there is congestion on the roads and you miss the plane, then you will be very unhappy. You'll miss your holiday. So don't take risks in that sense. Now, there are bigger things, obviously,

when it comes to providing for a pension, to give consumption in retirement. But ask the question, what are the things that could derail this sort of path I want to live along? What are the big things that could go wrong? Take out insurance against that or take actions that will minimize the likelihood of those risks occurring. That's the first thing. But the second thing on uncertainty is to embrace it, because all the good things in life actually come from uncertainty, whether it's the new products

that we can find that we never expected to be able to use. I never expected when I went to Cambridge to study and to work on that first computer model of the British economy, I never expected that I would have one day in my hand a telephone, which was in itself amazing enough, but also a computer that was many thousands of times more powerful than a computer for the whole of Cambridge University when I started as a student. These things are tremendous. And all the other elements to uncertainty that we should embrace, you meet new people that you didn't know they existed. You met them.

They come into your lives. They can transform your lives. You go to a place that you never visited before and you discover something wonderful that you hadn't really imagined and you want to go back there again. You discover a new restaurant. You discover a new book or a piece of music. It's the uncertainty which is the spice of life. And that, I think, is the thing to hang on to. Serendipity is a wonderful thing. That's the good thing. Risk is not a good thing.

So think carefully about ways of avoiding risks to how you would like your life to pan out. But then for the rest of it, embrace the uncertainty and allow serendipity to take its course. Mervyn King, thank you so much for joining me. Pleasure, Joe. Thank you. Thank you so much for listening. I hope you enjoyed that conversation as much as I did. For show notes and links to everything we discussed, go to my website, www.josephnoelwalker.com. That's my full name, J-O-S-E-P-H-N-O-e-l-w-a-l-k-e-r.com you can also find me on twitter my handle is at joseph n walker until next time thank you so much for listening

ciao