Pres. Biden is pursuing some of the largest spending proposals in U.S. history, which should be sparking concerns about inflation and interest rates. But most prominent bankers and economists have told us not to worry. Fmr Central Banker Mervyn King says they shouldn't be so confident. On this episode, we speak with Lord King about his concerns of coming inflation, how he thinks central banks didn't learn the right lessons after 2008, and why he thinks the industry has become too reliant on models.
Mervyn King: The people that I listen to most are the people who from time to time will say, “I don’t know.” The reason why I think experts have got into trouble is precisely because they pretended to know more than in fact they can know.
Bethany: I’m Bethany McLean.
Phil Donahue: Did you ever have a moment of doubt about capitalism and whether greed’s a good idea?
Luigi: I’m Luigi Zingales.
Bernie Sanders: We have socialism for the very rich, rugged individualism for the poor.
Bethany: This is Capitalisn’t, a podcast about what is working in capitalism.
Milton Friedman: First of all, tell me, is there some society you know that doesn’t run on greed?
Luigi: And, most importantly, what isn’t.
Warren Buffett: We ought to do better by the people that get left behind. I don’t think we should kill the capitalist system in the process.
Luigi: After the last episode, we got intrigued about trying to find out a bit more about what insiders at central banks think about many of the issues that we discussed last time. So, we were lucky enough to get an exceptional guest, Lord Mervyn King, former central governor of the Bank of England.
Bethany: We had a wide-ranging conversation with Lord King, in which we discussed everything from whether central-bank policies are creating inequality or exacerbating wealth inequality, and why we have such a reliance on models which are often faulty, to why it is that experts have lost credibility in recent years.
Luigi: We just interviewed Karen Petrou, who has this book blaming the Fed for inequality in the United States, which I personally thought was a bit aggressive. Certainly, low interest rates had some distributional effects, but I felt that declining interest rates throughout the world are not just the result of evil bankers, but are the result of a global economic situation. How do you see this from your perspective?
Mervyn King: Well, I totally share your view. I think the global decline in interest rates, both real, adjusted for inflation, and nominal, that’s a function of developments in the world economy. In particular, the fact that a large amount of savings was injected into the world capital market, and investment didn’t rise in proportion to that.
I think the challenge for central banks, and one which I don’t think they’ve been very good at handling, is the following. Any monetary policy, any movement of interest rates, can be seen to have some redistributive effects. The reason we don’t care too much about these redistributive effects is that sometimes interest rates go up and sometimes they go down. To the extent that this is a cyclical pattern, the long-run redistributive effects are a wash.
What we’re seeing uniquely in the last 15 years is that not only have interest rates come down almost to zero, but there’s been no cycle. There’s been no moment when interest rates rose significantly. That’s where people now get worried about the redistributive consequences of it.
The criticism I would make of central banks is not that they created these redistributive effects, but that they haven’t been very successful in persuading governments or populations that the way out of these problems is not in the hands of central banks. There are many reasons why a world economy can grow very slowly. It’s very rarely the result just of monetary policy.
I think that central banks have given into the temptation to be seen as the only game in town. If you become the only game in town, then you’re going to be blamed for everything that happens. We’ve ended up in a situation where, whenever there’s a piece of bad news, central banks inject lots of liquidity into the economy. I think that’s a mistake. You need to ask the question, what is going on here? Does this particular downturn justify a monetary policy response or not? And it’s not the case that every downturn should be followed by a monetary policy response.
Bethany: Do you think central banks understand the risk posed to them by this expansion of what they view as their purview?
Mervyn King: I’m sure that when they sit in the bath at night or in the morning, that they must worry about this a bit, but they probably feel under tremendous political pressure. I’ll give you the example that I feel worried about: the number of central banks that have now gone very public in embracing actions against climate change. It seems to me a very good example of almost a political move to say, “Well, that’s where the political sentiment lies, we must follow it.”
There are times when an organization should say, “Yes, climate change is very important, and it is important that we take actions to deal with it, but it’s not something for this organization. Governments have to be responsible for the tax measures, whether it’s a carbon tax or other measures. Other organizations should be asked to deal with it. If we jump on the bandwagon of dealing with climate change, we will lose our focus as an organization. Talented people in the organization will spend more time worrying about climate change than they will about monetary policy.”
That’s exactly what went wrong before the financial crisis, in which many of the regulators took their eye off the ball of prudential regulation, because they were focused on other aspects of consumer protection, investor protection, insider trading, and so on. All of those things are very important. But the image in my mind is a group of small boys and girls playing football, soccer in your language. If you watch youngsters playing soccer, what you see is a group of 15, 20 kids all running in the same direction, chasing the ball. They will move around on the pitch together. That is what is wrong with policy. When you watch a professional soccer game, you don’t see all the players chasing the ball. They’re spread out on the pitch, each doing their own thing. That’s what policymakers need to do as well.
Bethany: Why do you think central banks have embraced this expanded role that they’ve been handed?
Mervyn King: Well, if I go back to the financial crisis, I know very well. Tim Geithner told me on more than one occasion that we have a big advantage in the UK in having a parliamentary system. Because under a parliamentary system, the executive can take their actions very promptly in just the same way as the central bank can move quickly. I think in the US, the reason why there’s pressure on the Fed to do things, often from the Treasury, is that the Treasury knows it can’t act quickly, because Congress will not act promptly. That delayed response means that the pressure for an immediate reaction is thrown at the Fed.
Now, my own feeling is it’s better to go to Congress and say to them, “Look, these are the measures that we think need to be taken. They ought to be ones decided by you, the elected politicians. You need to make an early and quick decision on it. If you can’t and we feel that the economy will collapse unless we take some action, this is what we would propose to do, and you’ve got to give us backing to do it. Otherwise, we’ll make quite clear publicly that you were responsible for pushing the economy into a collapse.”
I know it’s easy for me to say that, but I do think at times my whole experience of dealing with politicians is that they are only too willing to get other people to take responsibility for actions. But in the case of the US, we can’t really afford to let them do that. Otherwise, they will turn around afterwards and take away from the Fed the powers to do what they felt they had to do in an emergency. That’s exactly what happened after 2008.
The powers of the Fed now to deal with a financial crisis are reduced compared with where they were in 2008. That’s not a good outcome. The right outcome is for everyone to agree ex ante that in a crisis, the powers that can be taken are allocated between the Treasury and the Fed in well-defined ways. So, everyone knows who can do what, and there’s no reason why that can’t be decided in advance.
Luigi: One worry you did not mention is the fact that Chairman Powell has been fairly aggressive in saying he’s not going to raise interest rates in the near future. With trillions after trillions of fiscal packages brought in by the Biden administration, do you start to be worried about inflation like Larry Summers?
Mervyn King: After the financial crisis, when QE started, what was so striking was that the amount of money in the economy, the broad money supply, did not grow rapidly at all. The QE actions by central banks were there to prevent the amount of money in the economy from contracting, from falling. So, the QE was to prevent a repetition of the Great Depression.
Fast forward to 2020, the commercial banking system was not contracting. The amount of money created by central banks is added to the amount of money in the economy in total, such that the growth rate of even M2 in the United States is growing faster now than I think at any point since the Second World War, and monetary growth rates in Europe and in the United Kingdom are at double-digit levels. Now, that doesn’t guarantee that we’ll get future inflation, but it certainly should make people ask the question, what’s going on here?
I would have thought there’s a bigger inflationary risk than many people in financial markets now have priced in, although, clearly, financial markets have actually adjusted to the view that there is a bigger inflation risk. I think the danger for central banks is that if they appear to give a commitment not to raise interest rates for a long while—and they can’t know today whether that’s a sensible policy path at all—but if they give that impression, then if inflation were to pick up, they would be faced with an invidious choice between either saying, “Well, when we said that, we obviously got it wrong,” which may damage their credibility, or holding rates too low for too long, in which case inflation and inflation expectations will pick up, and their credibility will be lost in that way.
Bethany: Is part of the problem, in a simplistic layperson’s summary, that we, in a sense, are taking the wrong lessons from the financial crisis, or perhaps lessons that were applicable then but are not directly applicable now? In the sense that one lesson of the financial crisis was, go big, throw everything at the wall that you possibly can, and something will stick, and we’ll get our way out of it. A second lesson from the financial crisis was, wow, everybody worried that this expansion of the Fed balance sheet and everything we did was going to lead to inflation, and it didn’t. So, therefore, inflation is dead, and we’re never going to have inflation. Is that too simplistic a summary? Or does that—
Mervyn King: No, I think it’s extremely accurate. The first point is one where I think people have failed to learn the lesson. If you do nothing before a crisis and then suddenly find yourself in the middle of a crisis and throwing money at it, going big, is the right response. But, of course, the lesson you should take from that is that you’d better do things before the crisis. The thing you should be doing is actually setting out a clear ex-ante framework within which the central bank will provide liquidity and trying to constrain in advance the way in which, or the extent to which, central banks may have to expand their balance sheets. So, that’s one lesson that we should have learned. Instead, the idea that you always go big, whatever the circumstances and whatever is going on here, is not the right answer.
The second one, you’re absolutely right. We didn’t see inflation after the financial crisis. I think that’s because a number of people in the area of monetary economics looked only at the monetary base, the amount of money created by central banks, and said, “This has risen by a phenomenal amount. We’re bound to get inflation.” Actually, what mattered was what was happening to the broader money supply in the economy and the link between the monetary base and that broad money supply completely broke down—understandably, given the state in which the commercial banking sector found itself.
But we are now seeing very rapid growth rates of the broad money supply. I think what we should be asking ourselves is, what does that mean? What is actually happening now? What is happening to that liquidity? Clearly, part of it is maintaining and boosting asset prices. I think what central banks need to recognize and be much clearer on is that, at some point, either asset prices have to come down or incomes have to rise fast in nominal terms, so that the ratio of asset prices to incomes can return to a more normal level.
Because where we are today is a consequence of very low, long-term real interest rates, which I don’t think are easily compatible with a well-functioning market economy, so there will have to be an adjustment. The rate at which money supply’s growing now does raise the issue. There are many other reasons to be worried about inflation in the future, is to say, “Well, maybe the way through all this is that governments will just tolerate higher inflation for longer than they are currently pretending.” Of course, the Federal Reserve’s new average inflation targeting approach increases that risk. Because if inflation were to pick up now, the Fed would welcome that. They would say they want inflation to be above the 2 percent target for a while. But they’re going to be deeply reluctant, I think, to say for how long and for how far about 2 percent.
It’s quite easy to get in the mindset where you say, “Well, it’s good we’ve got some inflation now, let’s explain that it won’t last for very long. We’ll come up with one-off explanations.” Then you find yourself in a position where inflation expectations become deanchored from the 2 percent target. Then you’re on an inflationary path, which is very difficult and costly to reverse. We just don’t know enough about the economy to pretend that we can control inflation that precisely. We’ve seen that. I mean, central banks have been desperately trying to get inflation up to 2 percent with no success. Why on earth, therefore, do we think that they can manage to let it go to 3 percent or 3.25 percent for six months and then bring it back? We can’t do it that precisely.
Luigi: But even though, with everything you said, it seems that the case for inflation is actually pretty strong, what is the case against inflation? If you were to argue in the opposite direction, that you shouldn’t worry too much, what would be your best bet for that?
Mervyn King: I think I would argue that once you allow inflation to rise above the target level that for many years you’ve been saying is your definition of price stability, that people believe that there is no anchor to your policy reaction function, that you will allow inflation to rise. Initially, you may only want to go up to, say, between 3 percent and 4 percent. But if people’s inflation expectations have deanchored from the target, it may slip to 4 percent to 5 percent before you know where you’re going.
We know from past experience that the cost of lowering inflation once it’s gotten to that level, particularly if people expect it to remain at that level, is very high. You have to have a recession. I think the cost of another recession would be high. My guess is, actually, people would find it very difficult to implement policies until inflation got to a level where everyone agreed, this is awful, this is a return to the 1970s. But we don’t want to go back there, because we had deep recessions in trying to combat inflation. There’s no need, there’s absolutely no need, to go to that level, because the central bank has a fairly narrow remit and focuses on the control of inflation.
Luigi: It didn’t sound like an argument saying that I shouldn’t worry about inflation. It seems that you make an even more compelling argument that inflation is coming.
Mervyn King: You’re too young, Luigi, to remember the bad old days of inflation.
Luigi: No, I came from Italy. I remember in the late ’70s and early ’80s, Italy reached 22 percent a year. So, I do remember inflation.
Mervyn King: Right. The UK got to 25 percent and even the US was well into double digit-figures. That really messed up many parts of the economy. The burden typically falls on those with the weakest bargaining position to adjust their wages to reflect higher prices and lower-income families. So, I think if you’re worried about redistribution and all the things that people talk about today, inflation is a very bad path down which to go.
Bethany: It’s fascinating to me, as the noneconomist here, that whenever you have something like inflation that is this very basic concept in economics, and yet it’s still something we don’t really understand and can’t control. There are so many things like that across the world, these very basic concepts that were created by humans, but still not under human control.
Mervyn King: One of the problems with the way economics analyzes these issues is that it has phrases. It has variables in models, like GDP or inflation, or the degree of tightness in the labor market. But, actually, it’s quite difficult for these simple ideas to be linked to a single observable variable. So, your inflation rate is going to be different from my inflation rate, because you buy a different basket of commodities than I buy. People understand this at one level, but the whole concept of GDP is undergoing revision as we realize that we’re no longer in a world largely of manufacturing. We’re in a world of services, and many of those services are provided free of charge. Measuring GDP is not quite as straightforward as we once thought. That doesn’t invalidate the use of the data and the ideas. But it does mean to say that we ought to think a bit more carefully about simply assuming that because in a model you’ve got a variable called Y, which is GDP, and Pi, which is inflation, that these things correspond to absolutely unique, measurable variables.
Bethany: Let’s get started with this question about economic models versus reality, and how it is that the economist profession has become so reliant on models.
Mervyn King: I think it’s a wish to make the subject appear as close to physics as possible. But I think the difficulty, it was well illustrated in my experience, with the financial crisis, the models that had been making quite reasonable predictions suddenly completely failed. The question was, why? It was fairly easy to spot the reason why, that many of the aspects that were crucial to the financial crisis, namely the banking sector, were missing from the models.
I came away from that thinking it was much more important that we would sit around and ask the question on our monetary policy committee, what is going on here? What’s actually happening? Then models can be used in an ancillary way to give you insights, to help you think through a difficult problem. But what models are not are a good description of the world. It’s not like science or modeling the path of a rocket being launched from Earth to Mars, where you know the laws of nature, they don’t change, and they don’t depend on what we believe about them, none of which applies in the world of economics.
Luigi: I agree about the cost of models, but there is also a cost of not relying on models and being too influenced by the fad of the moment. Because the models help you think through a problem and force a consistency that you don’t have when you reason without a model. So, in my view, the risk of going outside the model is that you are very much influenced by the pressures that are around without thinking through what the consequences are.
Mervyn King: It’s certainly true that one mustn’t just be influenced by the fad of the moment. Models do have the ability to ensure that things can add up, if you’re thinking of a macroeconomic forecasting model. But the trouble is they can also make you fail to realize that you’re adopting key assumptions which are essentially driving the whole result.
I’ll give you one example. I was asked to write a report on the performance of the monetary policy committee at the Central Bank of Sweden, the Riksbank. We looked at the differences of view on their committee over a five-year period. They had very sharp differences of view. Interestingly, they were all based on the same model, but they had different approaches.
One group said, “Well, if we keep interest rates lower for longer, then we’ll have to raise them in the second or third year of the forecast horizon.” The other group said, “No, if we raise them now, we won’t need to raise them as much by the end of the second or third year.” The interesting thing was that both of these groups made fairly precise forecasts, and they were both completely wrong. The reason they were completely wrong actually had nothing to do with the internal workings of the model, but it had to do with their assumptions they made about the rest of the world. So, they had to feed in assumptions about what was going to happen to interest rates in the euro area. The economists in the Riksbank said, “Well, we don’t know a lot about the euro area. Why don’t we ring the European Central Bank and ask them what they think the path of interest rates will be?” That turned out to be completely wrong. That really undermined completely their forecast. The big question they should have asked themselves is, what is going on in the euro area?
So, models can be used, but only if you stay outside them and ask yourselves, what are the big questions we need to understand? What does the model help us to understand about those questions? If it doesn’t, then you shouldn’t put too much weight on the numbers coming out of the model.
I think another very good example comes from COVID-19. A year ago, the epidemiologists said, “We’ve got models of epidemics. We can model this.” So, people said, “Oh, what’s going to happen to the number of cases and the number of deaths?” They produced forecasts and they were hopeless, just as bad as economic forecasts are.
What the epidemiologists had to do to make forecasts was to make assumptions about parameters like the mortality rate of the virus, or for how many days you could have the virus and be infectious without developing any symptoms. But we simply didn’t know what these parameters were. We had no idea. That didn’t prevent the modelers from making up the numbers, putting it into the model, turning the handle, telling politicians, “We’re the experts.” The politicians said, “We will follow the science.”
Bethany: It’s been clear since the financial crisis how models can lead us astray. Why do you think it is that we as humans just don’t want to look at things as they are? You’ve been arguing in your book since then that we need to do better at this. Yet, if anything, we’re becoming more reliant on models rather than less.
Mervyn King: Well, that’s a very good question, and I don’t have a simple answer to it. It’s partly that people seem to yearn after certainty. The big difficulty in preventing individuals from being exploited by other individuals, whether they are fake investment advisors or fake doctors, is that people want to meet someone who says, “Don’t worry. I understand. I know what’s wrong with you. This is what you have to do.”
I think the challenge is to try and resist that and to say, “Well, why do you think you know?” Then judge on the arguments that they give. The people that I listen to most are the people who from time to time will say, “I don’t know.” That’s the answer I want people to give sometimes, “I don’t know.” A doctor who says that is worth a great deal, because you then know that he will tell you things that he doesn’t know and be honest, if he doesn’t know.
We’ve seen a lot of criticism of experts in the media, and populism is meant to be an antiexpert movement. But the reason why I think experts have got into trouble is precisely because they pretended to know more than in fact they can know. That’s true of forecasts. It was true in the Brexit referendum debate in the UK, where the big mistake of the Remain side was to say, “Well, we are experts. We’ve got these teams of economists in the treasury and the Bank of England. We know that every family is going to be £4,300 worse off if we leave the EU.” People would say to me, “The government can’t possibly know that we’re going to be £4,300 worse off.” So, the credibility of that side was lost.
I think it’s much better for them to have said, “Look, none of us can really forecast what the outcome is. These are the arguments which, in general, lead us to believe, on our side, that it is sensible to remain in the EU. These are the risks of leaving. I’m not going to pretend to quantify it, because none of us really know.” It’s never happened before that an advanced economy left a trading arrangement like that. But by pretending to know, they lost the trust of people who might otherwise be inclined to listen to experts but won’t listen to them when it’s self-evident that they’re claiming to know more than they possibly can.
Bethany: I love that, three of the most important words in the English language, I don’t know.
Mervyn King: Yes, absolutely.
Bethany: It’s really nice to talk to you as always, Mervyn. Thank you so much for making the time.
Mervyn King: Now, keep sending me your work, Luigi. Keep doing that. You, too, Bethany, all the articles you write, I want to read. I look forward to seeing you both in Chicago at some point.
Luigi: I came out from the conversation pretty much scared about inflation in the future. What really struck me is that when I asked him to give me the best reasons not to fear inflation, he gave me other reasons to fear inflation.
Bethany: Yes, I totally agree with you. He’s one of the few former central bankers, one of the few economists I’ve heard, really raising the specter of inflation. It’s something that I think people are really worried about. I was on this call that Jamie Dimon did with investors and clients. One of their number one questions was, what’s your sense of inflation? Everyone is worried about it, but economists dismiss it. I think many economists dismiss it, sorry, Luigi. I think they dismiss it, as we discussed with Mervyn, because of a false analogy to the financial crisis.
Luigi: Yeah. To be fair, Larry Summers has been quite vocal, shouting the fear of inflation, which is very unusual for Larry. I listened to a debate between Paul Krugman and Larry Summers, and it was very stimulating. But, at the end of the day, I was surprised that even Paul Krugman was ready to say that the amount of fiscal expansion we’re doing was too much. But he said, oh, it’s too much, but he’s neutral. If things go badly, we have the tools to fight inflation.
What I got from the conversation with Mervyn is that maybe we have the tools, but first of all, we’re not sure how well they work. This is where the humility comes in. Two, we’re actually worried about the willingness to pull the trigger fast enough. The old definition of a central banker was the guy who takes away the punchbowl at the party when the party starts to get exciting. I think that you don’t see that guy around. It’s certainly not Jerome Powell, and you don’t see who it’s going to be.
Bethany: As soon as you said that, I thought, “How quaint.” I don’t think we’ve seen a central banker like that for a long time. I talked recently to someone who was involved in decision-making in the Obama administration, post the financial crisis. This person said, “We thought there was a limit as to the amount of debt that the United States could carry. That’s why our fiscal policies were somewhat more conservative.”
It’s quite amazing how you and I, the first podcast we did together, I think, or one of the first podcasts we did, was on MMT. It seems like we’ve gone to MMT without anybody officially saying that we’ve gone to MMT. I think, I remember in reading Stephanie Kelton’s book, that the key question for me that emerged out of that was, well, if inflation is the thing that makes you stop this, how do you know when too much inflation is too much? How do you stop it in time? I’ve always thought that inflation was one of those things that, once you see it happening, it’s too late. Is that an oversimplification of it?
Luigi: No, I don’t think it’s an oversimplification. I think it is possible to stop it, but as with many economic phenomena, it’s easier to stop them if you catch them early. But your incentive to catch them early is very limited, especially in the current political climate.
We know that the way to stop or slow down inflation is that you have to raise interest rates. In a sense, you have to create a small recession. I don’t think there is a lot of appetite anywhere in the political system to create this small recession. I’m not sure there’s a lot of appetite at the Fed, either, to do that. So, I don’t see the political will. In fact, I see the opposite. I see a bit of condescension that says, “A bit of inflation will help resolve a lot of problems, starting from student debt to overindebted households and the overindebted state, and so on and so forth.”
Bethany: How the Fed stopped worrying and learned to love inflation. Maybe that should be the title of our episodes.
Luigi: Yeah, except that I’m not so sure I have an answer to the question, how did Jerome Powell fall in love with inflation? In part, to be fair, there was a growing trend among macroeconomists at central banks that said that maybe the 2 percent inflation target was too low. First of all, there’s nothing magic about two versus three, or one, or four. The most important thing is to have stable inflation, not very variable.
Also, central bankers realized that if you start from two, an interest rate that is already low, your margin to run monetary policy is very limited because there is what is called zero lower bound, that you cannot bring nominal rates below zero. In fact, in some cases with the reserves, you can get slightly below zero, but certainly deposits of individuals are not below zero, and so on and so forth. So, that limits tremendously the ability of central bankers to stimulate the economy by cutting interest rates.
If you start from a high level of inflation, then interest rates on a nominal basis tend to be higher. So, you have more room to cut them when you have a crisis. Ironically, the problem is that nobody wants to say, “We moved to two to three, or two to four.” Because once you start to change the target, what prevents you from saying, four to five, five to six, seven to eight, and then keep going? The part that scares me is what Mervyn said, that, “Yes, we don’t really know how we can control it.” So, it’s not obvious that you can have an epsilon of inflation more and then be fine. It could be much higher.
Bethany: A question on this topic, because I know it’s one a lot of people raise, and it ties into one of Mervyn’s other points that is very near and dear to my heart, which is if you’re measuring the wrong thing, then your measurement is, I’ll put it politely, screwed up from the beginning. Do you think our measures of inflation capture actual inflation?
Luigi: Oh, absolutely not. Now, the question is, what are we missing? Every measure is, of course, imperfect. I was talking with some colleagues the other day, and one thing that came up, certainly during this COVID period, inflation was mismeasured, in the following sense, that if you define a good as a safe flight, the price of a safe flight skyrocketed. Because before, a safe flight was a normal commercial airline, and now a safe flight is a private plane. So, if you measure inflation in that way, you had an explosion. If you wanted to buy a Tesla, you had to wait for, I don’t know how many months, in order to get it. So, shortage clearly creates a hidden inflation, because you would like to raise the price, but you don’t see the transaction. So, that’s one way in which it is measured.
The other is that, and Mervyn was mentioning it, that inflation is different for different people. We end up measuring this as the CPI index, which is a basket that the average or median person consumes. But inflation for different people are different objects. Imagine that you read a lot of books. For you, what is very relevant is the price of books. If the price of books goes up, then for you, inflation is very high, versus the price of dating services.
The baskets are different, and in particular, the issue that is important is we see asset pricing increasing tremendously, and asset prices don’t enter directly into the CPI, they enter only indirectly through rental prices of units. Even the rental price of units I don’t think is measured that well. So, certainly, the current measure of inflation underestimates inflation in asset prices, and that is a relevant issue.
Bethany: Anyone raising children and trying to buy a house during the period since the financial crisis, trying to buy a house in an urban area, would argue that there was a great deal of inflation in the world. If you think of what’s happened to the price of education, both at private schools and at colleges, and then what’s happened to home prices, particularly, first of all, in desirable cities, but now in the pandemic and in rural areas. Some people’s experience of inflation would be quite profound, I would think.
Luigi: Absolutely. That’s the reason why you have to be careful in saying that inflation was not there. I think inflation was there in asset prices but not in consumption prices.
Bethany: Yep. What did you make of Mervyn’s argument about models and lack of faith in experts, that too much faith in models is a way of pretending to have certainty that we can’t possibly have? Then that pretense of certainty actually erodes credibility. I thought his example of Brexit was fascinating.
Luigi: Yes. I think that he brings a lot of healthy skepticism. There is a risk in not having an intellectual framework to analyze the data, but there is also a risk in excessive use of this mental framework. I always tell the story. There is a very famous novel that all Italian students are forced to read in high school. It’s not very known outside of Italy, but in Italy it’s very famous. In this novel, a traditional scholar of philosophy in the 17th century is faced with the explosion of the plague, and he’s trying to classify the plague in his mental framework, in his philosophical framework. He’s an Aristotelian and says that everything in the world is either matter or accident. He cannot classify plague as a matter, he cannot classify it as an accident, so he concludes that the plague does not exist, and sure enough, he dies of plague. I think that that’s a very useful reminder of the fact that we don’t want to be too enslaved by our own model.
Bethany: Yeah, I think all of that’s really interesting. I remember starting at Goldman Sachs way back when, as a young analyst, having to do all sorts of spreadsheet models and being so horrified to realize that all of these fancy ways of valuing companies that sounded so scientific, like a discounted cashflow analysis and a study of comparables, that it all just depended on the inputs. You could basically make the outcome say whatever you wanted to just by changing some small assumptions on the way in. It seems to me that where we’ve gone wrong with models isn’t so much in their use, because I think you’re right and we do need a framework. But it seems to me that where we went wrong in epidemiology with COVID is almost similar to where we went wrong in the financial crisis with value at risk. That was the failure to look at what was actually happening in the real world and say, “Is this lining up with what our model predicted? If it isn’t, then let’s adjust our model and figure this out.” I think that’s where the loss of credibility comes in.
Well, first, definitely the pretense of certainty when you can really have no such thing, and a model offers no such thing, but then the failure to say, “Aha, what’s happening in the real world is not aligning with the model we devised, let’s make a new model.” I think you saw, I remember Goldman Sachs saying in the wake of the financial crisis, and whether you like what they did or not by selling all of their bad securities to their clients, the reason they did it was because they looked at the market and said, “Our value-at-risk models aren’t working anymore. We’re looking at the market and where these securities are trading, and the models are breaking down, it’s not working.”
They looked at the real world and said, “How is the real world lining up with the models? It isn’t, we need to get out.” I think that seems to me part of the scale that’s been lost, is the ability to say, and maybe it comes back to Mervyn’s point, being able to say, “I don’t know.” Because being able to admit that your model isn’t working requires saying, “I didn’t get it right. I have to make some changes.” Maybe we’ve all lost that capacity as human beings.
Luigi: Absolutely. But I think it is challenging, because let’s continue with the epidemiology analogy, because I think that our listeners can follow us more easily. Remember the first weeks, and even the first month of the pandemic, the data that that were arriving were very noisy. You knew relatively well the number of deaths, but you didn’t have really a good sense of the number of people infected. It was oscillating by a factor of three or four, and clearly by a factor of three or four, you could feed any model depending on how you choose your assumptions, et cetera. That’s the reason why I thought it was brilliant, and they should have done more, that in Italy—and I wrote for ProMarket a small piece about this—in Italy, an epidemiologist at the beginning of the explosion, he isolated a small town of 4,000 people and tested everybody in that small town.
That information was incredibly valuable, because it was the first to prove with a good degree of precision that half of the people are asymptomatic. At the beginning, people did not realize this that even the asymptomatic could spread the disease. So, until you understood these two simple facts, I think that any kind of measure of prevention was useless. I think that that creativity in taking models, understanding what the critical parameters are and collecting the data to feed those critical parameters, and then making decisions based on that. That I think is a real skill, and clearly Mervyn King was suggesting that that’s what you should do. In some cases, we have seen that in the pandemic, but in many other cases, we didn’t.
Bethany: It’s like any scientific inquiry, uncertainty as a key part of it. You have to acknowledge your uncertainty and then be willing to adjust when the real world shows you that your assumptions were wrong. I think, for some reason . . . Maybe it’s partly because the pandemic was just so frightening that we were all grasping for a certainty that we couldn’t achieve. So, we were overly reliant on models. Sometimes I think some of our reliance on models comes from an effort to manage uncertainty. The model makes you feel better about the world, because it pretends to a kind of certainty that the world can’t actually have. If that’s true, then the more uncertainty there is, the more we would be driven into the arms of models as a refuge from the feeling of total terror that you have otherwise, existential terror.
Luigi: But I wonder, what are the effects of the diffusion of Twitter accounts among scientists? Because in the old days, scientists were not talking very much with journalists, and there was a lot of filter and so on and so forth. Now they take to Twitter, they fight with each other on Twitter. The risk is that they get entrenched in their positions and they don’t get very flexible. So, what you need to do is recognize you don’t know and be very open to change your mind based on data. But if you go out of your lane in front of a lot of people early on by stating, “This is a super lethal disease, or this is nothing and is going to pass tomorrow,” it is very difficult to walk back from those words. You lose the best minds, because they’ve become radicalized and engage in a shouting match in the public sphere rather than in a serious intellectual discovery.
Bethany: I think you raise a really important point. I think once you’ve staked out a position publicly, it becomes much harder to back off that point, particularly if you’ve gotten tens of thousands of followers because they like the point of view that you’ve staked out. You have a lot of incentives never to back off it. I actually wish that the scientists were fighting with each other on Twitter. What I see far more often is they just cancel each other. Instead of saying, “Aha, you raised this point. I don’t think it’s right. Here’s why.” They say, “How stupid are you?” And “I’m blocking you.” It’s really quite atrocious.
To me, that shows the underlying problem, which is they’re not confident enough to actually have a discussion than to have an argument on the merits. So, they make their decisions and their pronouncements based on outrage. That attracts more followers because that’s music to the ears of many, but unfortunately it doesn’t make Twitter a forum for the advancement of scientific beliefs or discussion.