Scott Sumner is the Ralph G. Hawtrey Chair of Monetary Policy at the Mercatus Center at George Mason University and blogger at themoneyillusion.com and econlog and author of two books: The Midas Paradox and more recently, The Money Illusion.
Scott came to prominence because his work on the Great Depression (published in Midas Paradox) gave him analytical superpowers for understanding the Great Recession in real time in 2008 and 2009 and beyond and in retrospect. We've seen the monetary policy establishment move closer and closer to the views Scott has been trying to talk them into for over ten years.
Scott is one of those people who understands some very deep things about a very challenging subject. We can all learn from Scott!
In the conversation we cover:
- How might he design crypto monetary policy?
- What matters more, revenue or wages?
- Where does monetary policy end and fiscal policy begin?
- Why isn't the fed more politicized?
- How does Monetary Policy really work? How are inflation expectations set and how do they really matter?
- NGDP targeting and how Scott's view has changed on it
Surprise, It's Insurance mailing list
Social Science of Insurance Essays
My guest today is Scott Sumner, the Ralph G. Hawtree, Chair of monetary policy at the Mercatus Center at George Mason University. Scott is an influential blogger on monetary policy at econ log and money illusion, and has written a new book called The money illusion. We'll be covering that and much more today. Scott, welcome.Scott Sumner:
Thank you for inviting me, David.David Wright:
So I'd like to start actually in an area that you haven't commented on much before. But that's cryptocurrency. And to me, if I think about programmable money, that offers a lot of interesting possibilities to the control of monetary policy in an economy, and so here, we'll have to assume that the monetary policy in an economy is represented in or the money in the economy is a cryptocurrency. So programmable money if it gives us infinite control over the money supply, because you can program whatever algorithms you'd like to add or take away money to whomever you care to. And you can actually be omniscient about what's going on, you can see all the transactions in economy and measure them and track them by software. So you know everything I can do anything, how would you design the monetary policy for a cryptocurrency?Scott Sumner:
That's a very good question, I probably would continue to favor something kind of like nominal GDP targeting. In this scenario, you describe it, it should be easier actually to accomplish that, because you wouldn't have the information like we have now or we don't know, for weeks or months, what the data is going to look like on you know, the level of nominal spending in the economy. So either cryptocurrency or actually any kind of electronic money that gave you real time reading on the level of nominal spending should make it easier to stabilize nominal GDP and the actual techniques used would be pretty much the same as under traditional monetary policy, you'd adjust the quantity nudged up or down to keep nominal GDP on target. Now, you may not want to necessarily target the current level of nominal GDP, there's a pretty good argument that the economy be more stable if you target expectations of nominal GDP, say one or two years forward. And that's especially true in a scenario like COVID, whereas probably optimal for nominal GDP to briefly decline, rather than try to artificially prop it up last spring. But you'd like it to return back to the trend line, not too distant future. So in that case, you you create futures contracts, and you'd sort of pegged futures contracts linked to nominal GDP, and use monetary policy to stabilize the value of those futures contracts.David Wright:
You know, one of the things as I was pondering this world, I started kind of asking all kinds of inconvenient questions about like, so what is normal GDP, and you have an interesting definition, what it mean anything is pretty standard definition of it. It's all the spending that happens in an economy. But when you think about like cryptocurrency exchanges, it's simply sending money to somebody. And nominal GDP, I think tracks a certain kind of exchange, like a special kind of exchange, which are commercial exchanges. Right. So if I just, I don't know, I just gave you money. Scott, just handed you $100 A bit. Things. That, is it. That wouldn't be nominal GDP? I don't I don't know like, you know, what every transaction be counted? Or would you need to designate certain special transactions?Scott Sumner:
Most transactions are not part nominal GDP. Let's start with the financial system. transactions in the financial system are far larger than GDP. But they're not counted, because they don't represent production of goods and services.David Wright:
So normally buying assets, you mean like,Scott Sumner:
yeah, trading, securities trading, foreign exchange, trading bonds, all that stuff. So the amount of transactions in the economy far exceeds nominal GDP, by the way, in a trivial sense. Also, when you buy and sell used products, that's not really part of nominal GDP either, because those were produced in an earlier period, like garage sales, things like that. But um, so the question is, then why do we care about nominal GDP? Why don't we just care about all transactions? Why don't we want to target that? Well, there's two reasons for caring about nominal GDP one is stability and nominal GDP helps stabilize the labor market. And second, it helps stabilize the financial system. Because nominal GDP is both the sort of total gross flow of revenue in the economy and the gross flow of income. So if nominal GDP goes down sharply, there's less revenue in the business world to pay wages and you get high unemployment. If nominal GDP falls sharply, you have less total income in the economy to service debts, and you have a lot of debt defaults and a financial crisis. That's not necessarily true when there's a drop in other types of transactions. So you can have a drop of transactions in the foreign exchange market. That's very large. Just for some reason, people aren't trading foreign exchange as much, but that wouldn't cause high unemployment and debt defaults in the way that a drop in nominal GDP would. So there really is a have a specific reason to focus on that kind of transaction as being important for the health of the economy.David Wright:
So we would need in our little universe here, we would need to carve out certain kinds of transactions. So in the kind of the jargon of the crypto world, we'd need two kinds of wallets at least. Right? We need a kind of wallet that receives, I guess, commercial transactions, right? So receives revenue and another kind of wallet, that maybe that same wallet pays wages. So because you're saying revenue and wages are special, and other transactions are not. I got that, right.Scott Sumner:
Yeah. So here, here's how I would put it like, there's really two questions here. If, if we're using the current techniques for stabilizing nominal GDP, current types of monetary policy, we don't really need to know which transactions are for which purposes, because they don't know that now with US dollars, the Fed doesn't know how many dollars are spent on this or that. Instead, they directly monitor the variable they're trying to stabilize, right? So they estimate and GDP. However, if you want to do a next level monetary policy, where instead of reacting with a long delay to data, you actually see real time data on nominal GDP, then you do need to know where this stuff is spent. So you'd need to have real time data on how much is spent on retail stores, or how much is spent in the service sector and things like that. And I don't know enough about cryptocurrency to know how hard it would be in a centralized system to have that kind of data. But I mean, you could certainly imagine a future of some sort where all money is electronic. And there's an enormous amount of real time data on the economy, so that we would have at least a pretty good estimate of current nominal GDP. How precise that would be I don't know, like, if you break the economy down, you have some firms that are large and sort of easily monitored, like you could imagine Walmart, providing everyday data and how much electronic money is spent on their firm, right. But then maybe a little mom and pop store, you don't have that information, unless you have some way to centrally monitor it by having, you know, all transactions tagged in some way. You know, like if you if you have some sort of future economy like 1984, where the government is watching everything, and even every little small business has a code number associated with the business and every transaction shows up with that code number and so on. I mean, I could imagine a scenario where you could monitor in real time, how much is spent every single day on nominal GDP type transactions. But that would require a lot of data collection,David Wright:
It would and well, one of the nice things about neither you nor I being software developers, and we can assume that job is easy. And say, well, let's just make lets us make up that we can do all this. Now what now? How would we design the ideal system? I'm wondering about revenue and wages about whether they are equal in your model, or in the model we're kind of building here, or whether you would put more weight on one or the other because, at least in the popular conception, I care more about wages, I think then I care about my company's revenue, right or any other revenue? Is that is that conduct follow through and that being more important variable to track? Or target?Scott Sumner:
Yeah. And before I answer your question, let me back up so I don't get in trouble with your readers, I actually view the system I'm talking about as a dystopia. I'm not advocating. Now to answer your question. So you're, I think you're kind of right. But for the wrong reason, like, it probably is true that the wages part of nominal GDP is more important than the profit part. Because the wages part is probably going to be more effective in stabilizing the labor market. We worry more about unemployed workers than unemployed machines. So, um, but it's not so much because you know, we think wages are important to us that that's the variable that we might want to stabilize. As an analogy. The reason most people don't like inflation is because they don't like to pay more when they go to the store. But that's not why the Fed tries to control inflation, they try to control it because they think it leads to a more stable macro economy. They see inflation as zero some some people pay more, some people receive more, right. So when we're thinking about the macroeconomic benefits of some sort of stabilization, we want to think about what the ultimate goal is and what instruments are most effective in achieving that goal. So because I believe that unemployment is one of the most important macroeconomic problems, stabilizing the growth path of total wages, is arguably more effective than stabilizing the growth path in total nominal GDP.David Wright:
Imagine this kind of world where in such a universe, the impact of our crypto world, one of the things you could do is you could have monetary policy act as a tax or a bonus on transactions. So let's say we had this special wallet that was a merchant wallet, and I spent some money there. And suddenly, the the, the algorithm recognizes, oh, the economy's running hot there, there's too much activity going here. And so they put a little put a little sales tax on my transaction to one disincentivize me from doing it, but to to actually drain some money out of the economy. Right. And so you can actually enact a contractionary monetary policy through the point of sale. Right. And so that that analogy, like you react to that makes monetary policy and kind of this perhaps dystopian, idealized sense, a tax or a bonus on consumption. Is that make sense to you? Or?Scott Sumner:
Yeah, I would actually call that a fiscal policy, though, maybe it'd be a fiscal policy implemented via the monetary system. But if you're if you're actually taxing sales, to discourage sales, that's, I would consider that more of a fiscal policy. But I mean, you're right, you could, in an all electronic money system there, there's ways in which government can do policy more quickly and maybe effectively than in the current regime. Now, something that's similar to that, but is not fiscal policy is adjusting the interest rate on money. either positive or negative, it can be either one, right, so if you want people to spend more, one way to do that is to increase the velocity of money by essentially taxing money, discouraging people from holding money, yes. And if you want people to spend less, you can subsidize the holding of money by paying interest on money, so that people are more content to just sit on their money holdings rather than spend the moneyDavid Wright:
now, so let me keep going on this for a sec. Because if I think that I think of the goal of monetary policy is actually not say Ngdp. Right. So to me NDB stabilization or the path of Ngdp growth, then, if in my kind of concept of like this, I agree with I mean, I agree with the the idea that you know that the link between the word tax and fiscal policy, but doesn't necessarily need to go to like a treasury department or be used for the revenue doesn't need be revenue for the government, it could just get extinguished. Like, you just delete the money that comes in through this policy, right, it could just go away. And there, we wind up draining some Ngdp from the pool directly without having to do anything other than just go away. The difference is, instead of draining it from a savings account, you're draining it from a transaction from consumption instead of savings right? Now, I don't know the implications of this. That's why I'm talking to an economist about it. But I think that if you see what I'm kind of trying to get out here, I'm trying to like, see if we could move the transmission of monetary policy away from the asset market and into like the consumption market without having governments necessarily be involved in a fiscal sense.Scott Sumner:
Yeah, I'm, I'm thinking out loud, that might be harder than you think. Because, um, you know, the Federal Reserve, for instance, is part of the consolidated government budget, right. So like, if the Federal Reserve pulls currency out of circulation, and burns it, it's really no different from the Federal from the Fed pulling currency out of circulation and storing it in a room. Because to the Fed, currency isn't really money, right, something that they've created, right? And so let me give you an example that I've thought about and see if it relates to what you're saying. So some people have told me they don't like the fact that the Fed injects money into the economy by, say, buying bonds, why not directly injected and where it benefits people? Right? So then I go through this thought experiment where I compare it to, like, suppose the government said, From now on, we're not going to buy bonds to inject new cash into the economy, we're going to pay government worker salaries. By doing so. Well, if you actually follow through with that, it's really no different. So by paying the salaries, they don't have to borrow as much money to pay the salaries because you're paying a newly created money. So the actual amount of debt held by the public is exactly the same in those two scenarios, whether the Fed buys the bonds or whether they pay the government workers directly. So what looks like a very different system is really pretty much the same system. Now a helicopter drop is different from paying government workers with newly created cash. Because when you contemplate a helicopter drop, you're contemplating two things at once, and easy money policy and fiscal stimulus. Right that's different from paying government workers that were going to be paid anyway. Okay. And if you're doing both of those things at once, what you'd really want to compare it to is buying government bonds and simultaneously doing that same welfare program through traditional means. Okay, you see what I mean? So helicopter drop can be broken down into two parts, you could drop enough to give everybody in the country $1,000 from a helicopter. And that's both money creation and its fiscal stimulus. Or you could do this, you could write checks to everyone in the country for $1,000, like we recently did in our COVID stimulus, borrow the money to do that, and have the Fed buy those bonds with newly created money. Okay, my point is, those are functionally identical policies. In the helicopter drop, we just dropped the money out of the helicopter, the government, the Treasury is not involved at all in terms of the bond market, right? It's all the Feds doing this. But in the traditional way of doing it, where we send checks from the Treasury for $1,000, to everyone, we get the money for that by issuing treasury bonds, and then we have the Fed buy those treasury bonds. Well, the Fed is just buying back, it's the government's debt. If the Treasury issues bonds, and the Fed buys those bonds, it's as if the bonds never existed in the first place. Like if I took out a mortgage on my house, and I became a counterfeiter, printed some money and bought back my mortgage from the bank. It says, if the mortgage didn't exist, I bought the house with counterfeit money, right. So I mean, the whole thing of the debt isn't really fundamentally what's going on. In that case, it's really about the the Money Creation and the fiscal stimulus that's occurring. So I think that people tend to worry too much about the fact that the government traditionally injects money or removes money by working with the bond market, it's really just because the Fed thinks it's convenient. It's not because they're trying to favor the financial system over real people. And, again, the decision to give new money directly to real people, is an interesting decision. But it's very different if you do it by pre existing obligations, like paying government worker salaries with newly created money, right? There's nothing fundamentally interesting about that. A helicopter drop is very different in the sense that you're actually doing fiscal stimulus to you're not just putting new money in the economy, but you're giving people money they wouldn't have otherwise received. Right, so that that's a fiscal stimulus, like a welfare program. So I think it's really important to be clear, in our minds, the distinction between how the money is put into the economy, and whether doing so is associated with a welfare program, or is just an equal swap for something else, like buying a bond of equal value or something.David Wright:
So I, I actually have the opposite, kind of I want to see concern. But surprise, I'm actually surprised this isn't a bigger issue. Because the political issue, this is a political conversation we're having here. And I want to be clear, I agree with your your technical assessment, of course, that, you know, when the Fed creates money, it's not just it can create all the infinite amount of money, it has to somehow get it out there and have to choose the transmission mechanism, right typically goes through asset markets, but it could go through, as you say, put in a couple of examples there or through some kind of fiscal stimulus or some magical thing that is fiscal without the government totally understand those distinctions. But the, I'm surprised that we don't see more political back, like think about compare this to say, like stock buybacks, right. And in the stock market, people kind of freak out about that, because they see this as being an artificial increase in the stock price by by a company buying its own shares and extinguishing them or just sitting on them. And to me, like the result on the asset market being stock is the similar kind of effect that the Fed has on interest rates, when it buys, say treasury bonds or whatever it buys. Right? It's sort of adjusting. It's actually in its rhetoric, and we're explicitly targeting the outcome, the market outcome of an asset by then buying and selling in that asset market. That is like I am, you know, that's a complicated kind of thing. Actually, it surprises me that people don't freak out politically about that by misunderstanding it more often. Because you could easily come up with this argument saying, well, you're just benefiting the fat cats on Wall Street by by making them rich in order to get the money back into the economy. Why are you picking favorites there? To me that political, the political question is, Can you can you decide on... or can you get the transmission mechanism to match a political agenda be it egalitarianism, everyone gets $100, or being let's pick different workers as the winners or let's pick hedge fund managers as the winners? It seems to me that we don't have that conversation as often as I would have expected.Scott Sumner:
I think there's a very simple reason for that. Monetary policy is just very, so esoteric issue. It's hard for people to understand. Most people just don't pay a lot of attention to it. And I think in our political system, most people look at outcomes. So to the extent people are paying attention to monetary policy is when they grumble about inflation or various outcomes of monetary policy. There are definitely a fair number of people, including maybe most of your listeners that do know something about monetary policy. And there are different camps different points of view, there's, you know, the Austrian School critique, there's the MMT critique, there's a lot of heterodox theories of monetary policy. And there's people that are critical of the Feds structure, the fact that, you know, bankers have a role in the policy deliberations at the regional banks. So there's there there is sort of at a, at a smaller level, a political element to this, but it doesn't really reach a mass audience, because the average person just finds all this confusing. If you talk to most people about monetary policy, they probably think in terms of the Fed adjusting interest rates, right, they don't even think in terms of the money creation as being central. And I think most people probably don't even understand the distinction between a gold standard and fiat money. You know, they, in the back of their minds, they probably sort of still think our money is backed up with something. You know, and well there's an interesting question as to whether, you know, our money is a liability of the government, you actually see economists debate that issue. But I mean, these these things are kind of tricky issues. So most people, I think, just look at the outcomes. And if they're unhappy with about inflation, they vote for different politicians, and so on. But I think it's very hard for most people to understand, you know, the critique of, you know, the Fed buying government bonds, rather than injecting money into the economy in some other way. Oh, here's another example. When I talk to average people about money injection into the economy, it quickly becomes apparent that people aren't distinguishing at all between fiscal and monetary stimulus. Right. So they've read a lot about, you know, a big stimulus bill. And they think that's money creation, but fiscal stimulus, you know, even in the $2 trillion fiscal stimulus coming along, doesn't create any new money at all, it just moves the existing money around in the economy. So people aren't even distinguishing, really, between fiscal and monetary policy, typically, when they think about these issues, and I'm talking about average people, the 90%, maybe 10% of the people follow the news closely enough, where they do have some knowledge of how the Fed operates. And some of these issues that are criticized. And, you know, at the political level, the issues you raised, I think, are starting to rise, mostly because the Feds footprint is getting larger. So they used to just buy treasury bonds. Right. Now, with all these recent crises there, they're going to mortgage backed securities and get they've gotten a little bit into the corporate bond market. There's debates about whether the Fed should have a climate policy favoring loans that don't you know, worse in our environment, there's discussion of inequality, should they buy municipal bonds? You know, there's, there's a lot of awareness of the Feds role becoming larger. In recent years. As a result, I think it is becoming a little bit more of a high profile political issue than it was when things were more calm, say before the Great Recession. There was a long period where people in both parties seemed happy with fed leaders like Greenspan and you know, just thought things were going along fairly well and didn't want to rock the boat. But when you have problems like the great recession or the COVID crisis, then then it becomes more of a high profile issue. And also, again, the Feds footprint is so much larger, the monetary base, that is the amount of cash and bank reserves they've injected in the economy has become so large, that they've had to buy many more assets, even unconventional assets than in the past. And that large footprint has made them more controversial,David Wright:
would you so I think you appeared a moment ago to be a little ambivalent about the transmission mechanism as as my I'm describing anyways, as markets, is it fiscal policy? Is it something else? But I'm wondering whether there are technical arguments to prefer one or the other of a transmission? I mean, I don't offend people maybe has looked into this. The because to me, you're, you're at least alluding to the possibility of distortion in certain markets, by the Feds reliance on asset purchases, and exposing it to a certain political vulnerability. And on the political point was one thing or just came to mind that I'd like to at least make the point at least Economists aren't immune to political preferences. Right? I mean, even if the subset kind of doesn't matter how small the subset of the population is that understand that it still surprises me. There's there's this is not more politicized, and as you say, probably getting more politicized now. But we, you know, if we could make it, if we could somehow maybe that could be an objective of Fed policy would be defined the least politically volatile or risky manner of transmitting policy. So I wonder if I could get you to come down and maybe with a preference for how you might design? Again, we assume that we can do anything in our crypto world, we'll come back to that. Would it be better for the Fed to do helicopter drops? Or would it be or by the Fed? I mean, or Kryptos, you know, AI or something like that? Or should they intervene in asset markets? Is there any reason to prefer one or another of these mechanisms.Scott Sumner:
So I'll put my cards on the table. I'm more of a small government free market economists. So I would prefer the Feds footprint to be as small as possible. And for instance, that's one reason why I oppose the payment of interest on bank reserves. That's one policy that's made the Fed balance sheet bigger. And in terms of distortions in the economy, I'd prefer they just buy back Treasury debt. The Fed is functionally part of the federal government in a fiscal stance, so all the profits from the Fed basically go back to the Treasury beyond what they need for them to operate. And so when they buy back Treasury debt to issue new money, they're they're not distorting the economy very much there. That's the least distortionary way to inject new money into the economy. I'd rather they not pursue other goals. I mean, climate goals are probably valuable. But I'd rather elected officials pursue those goals. Fed policy makers aren't even necessarily government officials, the ones that the regional banks are not part of technically part of the government. So I'd prefer the Fed go back to the pre 2008 system where the Fed balance sheet was relatively small, open market operations where they injected new money were very, very tiny relative to the size of the economy. Just to give you an idea, before 2008 98% of the monetary base was currency, literally the stuff in your wallets. Only 2% of the monetary base was electronic bank reserves, reserves held in deposit at the Fed. Now, most of the money is bank reserves. We still have a lot of currency out there. But what's really ballooned is bank reserves. So the Fed has become much, much larger, I mean, many orders of magnitude larger in terms of its impact on the banking system. And this is a policy that I think was unfortunate because it's it's inevitably going to lead to growing politicization of the Fed. And, you know, arguments about what it should be buying and what, and so on. So I really liked the the pre 2008 system better, it was less controversial, it seemed to work pretty well. And it had a much smaller footprint on the economy.David Wright:
That's a phenomenal lead in Scott to some of the more interesting ideas that you've taught me through your blog. Because one of the most interesting things, and there are many is that the the Fed's balloon and balance sheet was a failure, but not necessarily a failure in the sense that it, it bought too much well, is that that had to write. And so you know, you've talked about how the, the absence of a sufficient trust in the Fed actually doing what it says, forces it to act more than it would otherwise. I wonder if you could talk a bit about this idea of if the Fed actually could credibly commit to a certain price level target, they wouldn't need to buy anything. And they in fact, in certain ironically, certain scenarios would actually sell to, to generate the right kind of outcome, even though they had to do quite the opposite after doesn't it?Scott Sumner:
Right, so, um, when I talked about the small fed balance sheet back before 2008, I was referring to the monetary base, which was mostly cash at the time. And in most developed countries, including the US that would be less than 10% of GDP, the size of the money created by the Fed. That's the total stock not annual flows, which are even smaller. Now, if you look at a country like Japan, their monetary basis 140% of GDP, massively larger than before, mostly extra bank reserves, and mostly because the Bank of Japan has let inflation and interest rates fall to such a low level. That's Sitting on bank reserves is actually a profitable investment. So in my mind, you want a economy where things are growing fast enough in nominal terms that people don't want to just sit on cash and hoard lots of cash or bank reserves as an asset. You want the monetary base to be less than 10% of GDP, not 140%. Now, what's kind of ironic and confusing I call this the Alice in Wonderland world is things are never as they seem in monetary economics. So on the one hand, on a day to day basis, injecting more money into the economy is expansionary, it's inflationary, it tends to push up inflation. But on the other hand, as inflation rises, people don't want to hold as much money. So the demand for money as a fraction of GDP is much less when you have higher inflation. So the example I like to use if the Bank of Japan were to raise the Japanese inflation rate, from near zero to say, five or 10% a year, people in Japan wouldn't want to just sit on a lot of zero interest, cash and bank reserves any longer. So the demand for money would fall sharply. Or if you prefer, the velocity would increase very, very sharply. Instead of holding 140% of GDP in base money, the Japanese would probably hold less than 10% of GDP. So a successful inflation policy in Japan, while it might require some monetary injection in the short run, would actually overtime result in a much smaller money supply as a share of GDP than their current monetary system. And this Alice in Wonderland world also applies to interest rates, where things are never as they seem so an expansionary monetary policy might involve in the short run a reduction in nominal interest rates. But if it creates inflation, that inflation will eventually push interest rates higher. So if you had a truly inflationary policy in Japan, a five or 10% year inflation, interest rates would actually go up, it wouldn't be tight money, it would be inflation that's pushing interest rates. So. So these things are often very confusing, because people like to just sort of look at the money supply or look at interest rates and think they know everything they need to know about the stance of monetary policy. But you really need to look at these things in the context of the demand for money to fully understand what they're telling us.David Wright:
And that is approximate by velocity. Right? So how much money is spent? Right?Scott Sumner:
So velocity is kind of the inverse of the demand for money. So let me see if I could give you a numerical example. If I'm, if the amount of cash in the economy, let's say is one divided by 52, of people's incomes, are of nominal GDP, 150 seconds of nominal GDP, then to support all those transactions, the cash would have to be spent 52 times a year. Right? So let's say you had an economy where GDP was 52 billion, and you had 1 billion in money. Right? So velocity is 52, the money has to be spent 52 times a year to support all those transactions. But your demand for money as a share of GDP is one over 50 to roughly 2%. Right? So you're holding a stock of money. That's roughly 2% of your annual income. So what we call the demand for money, how much money you want to hold, say, as a share of income is really the inverse of the velocity circulation. Does that make sense?David Wright:
Yes, no, no, I love it. i i Here's kind of one thing that I was puzzling over. Really literally, just before we started, trying to kind of work out my thought, normally, I like to try and do this before asking the question, but we'll see how this goes. So one of the one of the points you make in the book, and I've seen it elsewhere, is that productivity growth, actually is deflationary. Right. And one of the things that confuses me a little bit about that is this concept of I want to hold so much money in my bank account or my hand, I want to have this much cash. And if productivity growth increases, suddenly, I wind up spending less for the current consumption level, let's say there's no last. Since this week, I bought so much stuff next week, productivity goes up by 10%. Everything costs, I get everything cost 10% Less. Let's say that's the way it happens. Now, suddenly, I have more money in my hands. That suggests to me that I would want to spend that money, right. So I'm drawing an analogy here to another explanation, which is that we have this sort of reserve level of currency, we want to hold a certain amount of money and everything else we spent. Right. So if I were to then spend that money, that would be inflationary, I would think. So I'm wondering about how productivity growth fits in fits into this. If I'm thinking about it, right.Scott Sumner:
No, in that case, you're you're not thinking about right because when you spend money, it doesn't mean there's any more or less money in the economy. What we really need to think about is how much money there is in total in the economy. So then that's determined in this assumption by the Fed. Here, I'm talking about fed created money like cash and bank reserves. So if you have a certain stock of Fed created money, that's determined by the Fed, if you spend more, somebody else receives more. So the minute you have less, somebody else has more money in the economy. But for any given stock of money, productivity growth is going to be deflationary. Because what's going to happen is the, you'll get more output and less inflation for any given level of nominal GDP. In terms of the aggregate supply and demand model, this would be like the aggregate supply curve shifting to the right, which will result in a new equilibrium with a higher real GDP and a lower price level. Okay. Now, this this fallacy of composition is really central to monetary economics, what's true for the individual is not true for the group. So, to me, the most fundamental insight of all in monetary economics is that the Fed controls the nominal stock of money in the public controls the real stock. So we all feel like we control how much money is in our wallets, right? We just feel like the Fed can't tell us to hold more or less money. And that's a very strong instinct, we have like, we don't want to hold so much money, we can spend it put in the bank, whatever get rid of it. But at a collective level, that's not true. Like collectively, the public can't get rid of excess money holdings. When they try to, if the Fed puts more money in the economy than people want to hold, and they try to get rid of excess cash balances, what will happen is their attempt to get rid of this extra cash by spending, it will push up prices, until prices are so high that we again want to hold that extra money, because now we have to pay more for things. And so in that new equilibrium, the public has determined the real amount of money that's in the economy, and the central bank has determined the nominal amount of money in the economy. So each each group is sort of right in its own way the Fed thinks that can control the money stock, and it can control the nominal money stock. The public thinks it controls its money holdings, but it actually controls its real money holdings, not its nominal money holdings. And the interaction of those two groups is what makes monetary policy had macroeconomic effects. So the Fed puts money in the economy that people don't want to hold. And a new macroeconomic equilibrium is achieved, where they then do want to hold that money because prices and output of change, to make the public want to hold that extra money.David Wright:
So what I'm kind of hoping is happening is in my mind, and listeners minds, is that this is getting kind of weird, because if you think about the world, were back to your Japan example, or you know, your point there about, if you increase inflation expectations, people want to get rid of their money, because the money is getting less worth less, right? Why would I hold on to this thing that's depreciating by 10% a year to get rid of it by spending it? And then that, you know, then the Feds got to pull that money out of the system? Because now suddenly, we're spending all this extra money in this scenario. And, you know, and so the central bank has to like, you know, soak up all that extra cash, lest inflation, you know, go way up, right? overshooting, right? And so they're gonna have to start when people get in their heads that, Oh, I gotta get rid of this stuff. It's very easy to kind of like overdo it. Right? And so they got to pull the pull the money out that the interesting problem, then is what makes them think that it's going to get 10%, you know, that the inflation is going to come? How do you convince them?Scott Sumner:
Well, first of all, they don't always make the right decision in the 1960s and 70s. They didn't pull the extra money out, and we had, you know, 10% inflation during some years in the 70s.David Wright:
Do you know why?Scott Sumner:
Well, they had some they had they had bad models they had, they were making two or three intellectual mistakes. It wasn't bad intentions. I don't think they had a Phillips Curve model that convinced them that they could permanently lower unemployment by having a little bit more inflation. So that was one mistake. A second mistake was they misjudge the natural rate of unemployment. It actually rose somewhat in the late 60s and 70s. So they were aiming for an unemployment target. That was unreasonable. They were chasing a goal that was on achievable. They misjudge the stance of monetary policy because they thought nominal interest rates measured whether money was easy or tight. And since interest rates were fairly high in the 70s, they thought they already had a tight money policy. But in fact, interest rates were high because of the high inflation and monetary policy was not tight or contractionary. So there's there's three right off the top three intellectual errors the Fed made, and that we're all corrected in The 80s. And they brought inflation back down with a different regime a different approach for targeting inflation. And really ever since the 80s, central banks have known how to keep inflation under control. In fact, you could say there's actually been two big learning curves. In the last 100 years of macroeconomics, we had the Great Depression. And then central banks learned how to prevent something that deep, although I argue they made a smaller, similar mistake in the Great Recession, but much smaller mistake. But they learned how to prevent great depressions by doing enough stimulus to prevent, you know, 25% unemployment, then they made the opposite mistake in the 60s and 70s. And we have very high inflation, some people call that the great inflation. And the 80s, they figured out how to avoid that not just the Fed, but all the major central banks of all the major developed economies, none of them really have high persistent inflation. The only countries in the world now that have high inflation or developing countries with severe political problems, Zimbabwe or Venezuela, places like that. So we know how to, you know, control high inflation, we know how to prevent deep depressions, now we're engaging in the fine tuning part of it like, Okay, how do we prevent smaller recessions? How do we prevent smaller outbreaks of inflation? So we're still making mistakes, but we're making smaller mistakes than earlier decades.David Wright:
And so this, the that makes sense, yes, the, the important thing that I want to kind of get to, and I interrupted your train of thought, but is how these expectations are actually set. And here's the thing that kind of like, still bothers me about this. So you know, there, there's this new, you've talked about it elsewhere, but the, the Princeton School of Economics where you had this, this, this revolution, and really, really understanding or starting to really study and understand how expectations of expectations of future fed action or of outcomes the Fed trying to go for really, actually govern voluntary policies, this other thing, right. So think exists in between, like, it's just like a human, like, it's a social fact, right, of the somewhere between communication, you know, the economy's understanding exists, this idea of where of what we think needs to happen, or what the Fed will do to, you know, that's, it's not it's not measurable by money, right. And, you know, we can, you can measure it in certain ways, and you talk about using forecasts and stuff, which I do want to get to, because I think it's amazing, but this this kind of social fact of of, of the expectations, that creates what, though? So if is it just the aggregation of many, many businesses now having this sort of expectation, they adjust all the little micro decisions that they're making to suit this new reality? Because they believe in it? Is that actually the real transmission mechanism for monetary policy as opposed to all this crazy technical stuff about asset markets?Scott Sumner:
Well, I think expectations are very important. You know, how they're formed is, is an interesting question. There's a school of thought I adhere to, which is unfortunately called rational expectations. But um, it sort of implies that people are like the term implies people are like, Spock on Star, star trek, no, machine like thinkers that can rationally see the future. But rational expectations actually should be called consistent expectations. That is, what we're really assuming here is that the public's expectations are consistent with our model of the economy. So if we think a certain policy will produce this outcome, we're assuming that the public also has that view, because we have no reason for them to assume anything else. In other words, if we have a model where the model predicts, the policy will produce 4.5% inflation, it would be weird to put in the model a public that believed inflation was maybe 9% or 1%. Like why would you do that? It's not that I think the public will exactly have 4.5% inflation expectations. But I also believe in this concept of wisdom of crowds, that is when you aggregate the forecasts of a very large number of people. Often the results are surprisingly accurate. You've probably heard of like the Jelly Bean experiment where you ask 100 people how many jelly beans are in that jar? Well, nobody really has a clue, right? When you look at a jar, how do you how can you guess, and yet when you average all the guesses, it's actually pretty accurate to the number of jelly beans that are in the jar. So um, I also believe that certain expectations are more important than others. So let's take a money supply announcement. The Fed makes a money supply announcement. The financial markets instantly react the Stock Market jumps 5% within minutes of the announcement, and this happens sometimes, like I can cite examples 5% Jump in the stock market within minutes of a Fed announcement. I don't think the public actually looks at the Fed announcement and has some kind of new Keynesian model the economy and figures out what that'll do to, you know, the economy. Rather, the public looks at the stock market jumping 5% and takes its cue from those people. So in this sense, expectations are formed at two levels, insiders, specialists that follow Fed policy closely, try to figure out what its implications are for the economy, their their conclusions, drive asset prices quickly in response to those fed announcements, the public sees the movement and asset prices and draws inferences from that. So in other words, the transmission mechanism is very complex. And I even what I just sketched out, I'm sure is a gross oversimplification of what actually goes on. I also believe that central banks are relatively easy to read. So a lot of times I'm asked like, well, just because the Fed promises to do something, why should people believe it? Right. And I don't think they necessarily will believe it, I think what people will believe is, they'll believe what the Fed is actually going to do. Right? So during the 1970s, the Fed promised to do something about inflation, and they didn't. And they were not believed for very good reasons, they weren't really sincerely committed to doing something about inflation. But if a central bank really was committed to reduce inflation, or create inflation, and was going to do, this is an important phrase, whatever it takes to do that, I believe they will be believed. Like the public will believe them, I think of central banks is like this person I know, whose face is very easy to read. You look at her face, you can tell her emotions, happiness, anger, it's just right there on the face. Right, no hidden agenda, no hidden motives. And, you know, the Fed is a big public institution with a lot of people within the institution that speak publicly have different points of view, it's not at all hard to figure out what's what's going on, in in, quote, the mind of the Fed. One reason the economy did so poorly in late 2008, is people correctly understood that the Fed was not going to do whatever it takes to prevent a deep recession. So you know, there's a lot of discussion about, you know, forward guidance and how the Fed can shape expectations to stabilize the economy, and so on. But most important of all, the Fed must ultimately be determined to do what it promises. That's the hard part. Once it's decided it will do whatever it takes to achieve some longer term objective, I believe the public will come along, they'll, they'll believe that objective, that'll be credible. Now, if they've been dishonest in the past, it might take a little while to establish some credibility. But I think the expectations are, on average going to be roughly what the Fed is, is sincerely going to do.David Wright:
Now, the thing that I still still kind of bothers me is, if you think about it, the actual the actual outcome of NGDP, or of you know, let's call it wages, the real wage, let's put it this way. The real wage, which is nominal wages, adjusted for is divided by the price level, right, you know, some adjustment for the price level? Right? Those are those are, those are private company decisions, both of those things, the prices and the wages. So companies who pay people and the government's I suppose, who are a big part of the economy, they have to make decisions about wages and prices, that result in an outcome that is then counted as in GDP and inflation. So somehow, these organizations have to have an expert. So to me, like that's why I kind of come back to this, you know, that the expectation of the Fed actually results somehow in decisions by individual companies to do things. And those decisions, surely are the real ultimate transmission mechanism of monetary policy, right? Be it whether they're forced to or something else, you know, you right now, we're in an economy. We're recording this in October 2021, which is pretty hot. Right now on GDP growth. You wrote a blog post this morning or yesterday, I think you said we kind of reestablish the we've plugged the nominal hole from the beginning of COVID. And, and so but like so companies are actually Making decisions now that are resulting in higher than GDP? How is why does like asset markets or predictions of a company's will do, right? So I get that. So that's like, you know, the experts, as you say, are saying, well, here's what I think the company's gonna do, and they're gonna make more nominal profits. And that's going to make their nominal asset value go up. And then the public is going to anticipate the company's decisions as well, which is going to get more money and things that are cost more. And so both of these things are, you know, it's a mixed bag. But why isn't the comp. Why aren't the company's actual actions, the micro decisions there featuring more prominently in your at least your description of the model?Scott Sumner:
Well, that's that's a huge question. And it's a very difficult question to answer, because this kind of works on several levels. On the one hand, let's start with the long run. So in the long run, the Fed is determining nominal values in the economy, nominal wages, nominal prices, you know, even the the allowance, I might pay my daughter in nominal terms that's controlled by the Fed, not by me. In real terms, all these things are controlled by agents in the economy. Right. So that's the long run, though. Why is this so confusing, because in the short run, it's not true. And we tend to observe life in the short run, we experience life in the short run, right. So in the short run, we control both the real and nominal wages, real and nominal prices, the real and nominal allowance, I pay to my daughter, those are all controlled in the short run by agents in the private economy. Somehow, we have to have a system where or a model where we can transition from the short run to the long run. And the so called transmission mechanism, how that changes is not easy to understand. It I mean, it seems to be partly related to this notion of disequilibrium, at least in some industries. So if you have, you know, expansionary monetary policy where existing prices and existing wages, there's not enough goods, there's not enough workers, then you have at the micro level, people bidding up wages and prices. So those nominal wages and prices start to rise to reflect the new money supply. But you know, wages might be adjusted only once a year. So the process is very confusing. Now, if you had a a perfectly flexible price economy, then these adjustments would take place right away. And we wouldn't see any real effects at all. And in fact, I sometimes like to talk about the example of currency reform. This is very far from the minds of most Americans, because we don't live in a country that has had currency reforms. But if you've lived in Mexico or another country, you may have gone through this. So after a while, they've had so much inflation, they'll take 100 Old pesos and go to one new peso, right? Well, after the currency reform, basically, all nominal values in the economy go down 99%. Instantly, like in one day, you go from the old pesos to the new pesos. Everything is 99% cheaper. in nominal terms, in real terms, nothing has changed. Now, if we ask, Well, why doesn't that huge deflation, that huge drop in prices create all sorts of economic problems. The reason is simple. With a currency reform, all wages, prices, and even debt contracts are 100%. Flexible, like everything's adjusted immediately. Not only do you do 100 Old Mexican pesos for one new Mexican peso, but your salary immediately drops from, you know, 100,000 pesos a week to 1000 pesos a week or whatever. Your mortgages adjusted 100 to one, right. So if all contracts are immediately adjusted, then that kind of monetary policy change only shows up as nominal changes, nothing real matters, nothing real changes. But what actually happens in the US economy is we have this kind of complicated process where that's still the long run equilibrium, like if the Fed doubles the money supply other things equal, in the long run, the only thing that changes is all wages and prices will double. That'll be the effect of doubling the money supply. But because wages and prices and debt contracts are sticky, are negotiated in nominal terms and are adjusted infrequently. The short run effect is for all sorts of real things to happen disequilibrium occurs in the economy when you have a monetary shock, and that disequilibrium sets in motion things that in the long run will produce the desired nominal change. So in that long run money is neutral, it has no real effects. You just have the nominal wages and prices adjust. But because of short run rigidities in the economy, you get disequilibrium. And that creates bidding up process of wages and prices are bidding down. If you have a deflationary monetary shock, you force wage cuts, the force price cuts you that's when you usually have unemployment occurring, because it's hard to cut wages. So you know, in terms of the the actual transmission mechanism, that I mean, there's like a billion transmission mechanism, because monetary policy creates all these little disequilibriums all across the economy, when the money supply is adjusted. It's just working through the housing market, it's working through the stock markets working foreign exchange rates change, wages become under pressure, if there's unemployment, or if there's a shortage of workers wages go up. That's just the million areas where we're you're getting little incremental changes in response to this monetary shock. But again, all of these incremental changes are only occurring because of stickiness. It's weird. Like, if wages and prices were 100%, flexible, like in that Mexican currency reform, you just have the whole adjustment occur once everything nominal would be adjusted, and nothing real would change in the economy when that occurred. Another currency reform examples, the euro, like nothing real changed in Europe, when they went from French francs to euros or German march to euros, just a different accounting system, like changing the length of yardstick. But because we have stickiness in prices, for these gradual changes in monetary policy, that's why you get these messy adjustments to monetary shocks.David Wright:
That kind of touches on the aspiration of my crypto example, right? Because if you're it is a very simple thing, you're saying the downward stickiness is the kind of the original sin here. That's the thing that makes all this stuff be weird, then in a world where you could arbitrarily make those adjustments through monetary policy or some kind of like you could you could design a mechanism that would correct for that at the ground level, perhaps you could, you could actually fix that in a more elegant way. But we don't have to keep going on about that. The, the, the thing that I am curious about, like, I'm gonna come back to this question of the expectations as it interacts with what you're saying that this equilibria because when I kind of have this, this sort of urge to say that, I feel like in an expectations driven monetary policy regime, the actual money that's created is kind of pulled, or rather there's a there's a, like the change in the demand for money, and in the thing, because of expectations of monetary policy, right? Or have expectations of inflation, and NGDP. Right, so that changes kind of how much money the public wants, because they're going to create that reality through these disequilibria. They're, they're making up these disequilibria, right? The feds not going around. And just in the marketplace, they're saying, Here's what the world's gonna be. Everybody says, holy cow, now we're out of equilibrium, but it's just a change in their mental state like they are they are making that happen themselves, right? It's always like,Scott Sumner:
let's be clear, the Fed is not trying to create disruptive disequilibrium. So they have their goal of like 2% inflation target, and they're trying to match the public's changing demand for money in such a way that inflation will stay close to 2%.David Wright:
Yes, that's what I'm getting at here. Because the public, let's say the public suddenly gets in this bad macro equilibrium, where their demand and money has changed. For some reason the Fed wants to set reset their expectations so that, you know, they reset their demand for money to some other level. Right. So now we kind of have this three step process. But the Fed hasn't done anything yet. All they've done is change their expectations. Right? It's like, you know, Powell goes out there as the current Fed chairman. So and then we'll say, here's the new expectations, public public says Done, now they go about busying about their lives, acting on those expectations, which are creating these little micro disequilibria. And then the Fed now has to just watch that and hand them the right amount of money. Right? Right. So it's got to like react to what it will see as the new velocity given its new expectations. And so to me, like, the Fed is setting up the expectation that it's going to be as a highly reactive institution, not screw up the amount of money that gets in there. Because if they overdo, you're gonna do it, you're gonna wind up in the wrong place, and you're gonna mess up there. You know. So how do they do that? How do they know how much money you actually put in the system?Scott Sumner:
Well, that's a good question. So there's the easiest one answers. Probably the very, very short run fluctuations are mostly handled via interest rate targeting. Right so let's say there's more money needed at Christmas. Okay, you people are doing more shopping. And this is true by the way there is there's somewhat more money in circulation in December than the other month. And how does the Fed not know how much money to inject during December? Well, if they're targeting interest rates, then if there wasn't enough money, interest rates would sort of start to rise as banks are trying to get their hands on more money to meet customer demand for withdrawals from ATMs to go shopping, whatever. So to prevent interest rates from rising, the Fed will inject money in the economy holding interest rates stable during December. And that keeps the money supply matching, the demand for seasonal changes. These are the easiest to address. They're kind of predictable, seasonal changes, like Christmas shopping. And one reason we know this is true is that before the Fed was created in 1913, interest rates did move around seasonally. I think late in the year, interest rates tended to go up as I recall, because there wasn't the Fed to meet the demand, right? So there were these seasonal fluctuations in interest rates. Now the seasonal thing is easy to deal with. And we're not even trying to stabilize the economy over the year, we were happy to have GDP a little higher in December than January, right. So the bad kind of fluctuations are deep recessions and overheated economies with high inflation. For those they need models. And it's not enough to peg interest rates. Because if you peg interest rates for a long period of time, and the equilibrium interest rate is changing, and you're not changing the target rate at the Fed, then interest rates will get out of equilibrium and the economy will spiral off course, if interest rates are too high, it'll spiral downward if they're too little spiral upward. So the Feds developed techniques, things called like the Taylor rule for adjusting interest rates. I'm personally not a fan of interest rate targeting but most economists disagree with me. But you know, the Bank of Singapore targets exchange rates as a way of stabilizing their economy. And the Fed is constantly looking at all sorts of data to try to sort of predict where things are going. It's sort of like imagining someone driving down the highway, and it's kind of foggy, and they're trying to stay in their lane. But they can't see to well, they're looking for any indicator that they might be drifting over the centerline or off to the shoulder. And then they they adjust the steering to try to stay in the lane. And the and it's even worse, if the car just slowly to the turn of the steering wheel, right, you'd like a car like a Ferrari that responds quickly. So the Feds got this steering wheel, they turn monetary policy, but the economy is like a big ocean liner, it doesn't move right away in response to the monetary policy. So it's it's often difficult for the Fed to provide the amount of money that allows them to hit their macroeconomic targets in an appropriate way. Nevertheless, at least when interest rates are above zero, they did seem to be sort of figuring this out to some extent in recent decades before the zero interest rate environment kicked in. We they were making less bad mistakes than in earlier decades. Then when interest rates fell to zero, and they were no longer able to adjust interest rates as a technique for steering monetary policy. We made some significant mistakes around 2008 and nine. But I mean, there's no simple answer your question, again, in the short run, targeting interest rates will meet little seasonal fluctuations in demand for money. But to control the business cycle, you need something more than that you can't just stable interest, stabilize interest rates forever. The Fed, by the way, generally adjusts its interest rate target every six weeks if you want to sort of timeframe on that. So that's good enough for to handling Christmas shopping and so on, but not good enough for long run macroeconomic problems. And they look at all this data, they have forecasts, they have models, which are not really very good models, as your listeners probably know, economists are notorious for not being able to predict recessions very well. And they do the best they can. Now my preference, I guess we'll probably get into this is that they should let markets steer the monetary policy but anyway, that's basically what what they do. They look at a lot of data and try to figure out as best they can, what kind of interest rate adjustments will keep inflation close to 2%.David Wright:
You did very accurately predict my next and perhaps final move for a conversation, which is the real thing is actually as you say they look so I would if I may characterize your comments there. They look at market prices. When they move in to decide how much money to put into and use it interest rates, and you also say these sort of economists' models, which I'm sure are very good, maybe, maybe not. But we know the market prices are well, I mean, I, I learned much of my faith in markets, market prices from you, actually, Scott over the last 10 or 12 years, how long you've been blogging. And, and what we know is that markets are really the best forecast for what's about to happen. And interest rates are a kind of whether they are marked as a market price they were forecast. And but there are probably better ways or better aggregates to develop markets for that would give the Fed real good guidance on exactly how much money to put in the economy. So maybe it's a good time to talk about another amazing idea that I learned from us NGDP futures, what would those be? And how might the Fed use that?Scott Sumner:
Right? So the it doesn't have to be NGDP futures. By the way, if you target inflation, you could have inflation futures. And we actually have something that's a little bit similar to inflation futures already. So I think it's worth mentioning that this isn't completely a hypothetical idea. We have something called the tip spread in the bond market, the difference between the yield on nominal bonds and indexed bonds. And that's a rough estimate of the amount of inflation people expect over various time horizons. So we do have market indicators of some of these variables. I favor targeting nominal GDP, we do not have any kind of direct measure of the market forecasts of nominal GDP, we can kind of put it together by looking at a lot of market indicators and try to infer what the markets seem to be expecting about nominal GDP. But I've proposed we've actually created a nominal GDP futures market and not just create one, but have the Fed stand willing to take unlimited positions, short or long at certain price points. Okay. So let's say the Feds goal is 4% nominal GDP growth. So you're gonna have a nominal GDP futures contract, where the maturity value depends on the actual growth in nominal GDP. And one idea would just have the Fed fed pegged the price of that contract at 4%. That at 4% growth, that might be a little bit extreme. So more recently, I proposed what I call a guard rails approach. And it would work kind of this way, people that thought nominal GDP was going to grow faster than target faster than 4% would take a long position on these futures contracts because they'd expect to profit from the higher than expected Ngdp growth. Those that were bearish on the economy expected slow growth would take a short position on the contracts. So the Fed would had guardrails at 3% and 5%. The Fed would tell the market, we'll take a short position on any 5% contract. Like we think nominal GDP growth will be less than 5%. If you disagree with us bet with us on this contract, you'll profit if it's more than 5% will profit is less than 5%. The Fed would also tell the market will take unlimited long positions on a 3% contract. So we think nominal GDP growth will be more than 3%. If you think it's going to be less than 3%. Go ahead and take a short position, we'll take a long position on that contract. And these would be sort of warning signs for Fed policy. So if the market was all over on one side or the other, it would suggest monetary policies probably off course, we're not going to get 4% growth like in late 2008, everybody in the market would have been taking short positions expecting less than 3% and GDP growth. I compare it to the beeping noise a truck makes backing up it's kind of like warning you you might hit something now you can still back up if you want to you can ignore that beeping noise. And the Fed can ignore those contracts and say, Well, we think we're smarter than the market. And we're just gonna do this policy anyway. But of course, if the Fed is wrong, and there's a deep recession or high inflation, they're gonna have to go to Congress. explain to them why they thought they were smarter than the market right? And why they were wrong. So I think what would actually happen is fed policy would be continually adjusted, in what I call a whatever it takes approach, they would do whatever is necessary to keep nominal GDP expectations for future growth within that three to 5% range, which is consistent with macroeconomic stability. So essentially, the market would be guiding like Think of it this way the market would be telling the Fed what sort of interest rates and money supply it needed. Just like a country that has a fixed exchange rate system, the market tells that central bank what kind of interest rates, it has to have to keep its interest rate fixed. I don't know, if you follow like Bretton Woods or today, Hong Kong's dollars fix the US dollar, they don't control their interest rate the market does, there's a certain interest rate that is necessary to keep the exchange rate fixed. Singapore, they target exchange rates. So the market tells them what the interest rate will be given their exchange rate target. And under my system, basically, the Fed would no longer target interest rates or exchange or money supply to any significant extent, they would essentially be following market directions on what kind of interest rate or money supplies necessary to keep expectations of nominal GDP growth on targetDavid Wright:
a question I have about the futures and kind of wondering, but there's ever since you first mentioning these for as long as I've been reading you the if let's say the Fed adopts a short position of 5%. And everybody knows, except maybe the Fed that the inflation is going to be 6%. And let's say, you know, let's just go the, you know, the absurd extent here. Let's say everybody in economy shorts. The rather goes long on these futures, that's five and a half percent. And the Fed then pays out. They're putting more money into an overly stimulated economy, because one of the things that I was wondering about like, was whether the there's something special about the the payout of the contract as a specific monetary stimulus.Scott Sumner:
It could be Yeah, it in my preferred scheme, it doesn't have any money supply implications. They're just paid with money that's already in the economy. The Treasury writes them a check, not not newly created money, right.David Wright:
It's only the information value of the forecast the actual paying the contracts isn't isn't part of the mechanism.Scott Sumner:
Now that my original version of the proposal, the money supply would actually be directly adjusted by these purchases and sales and so on. But I kind of like thought that's unnecessarily complicated, or I think it's easier to just think in terms of these payouts not affecting the money supply at all. And rather, the market is just providing information. And it's also sort of providing an incentive for the Fed not to want to miss the target because the the Fed does not want the embarrassment of losing a lot of the Treasury's money. So in other words, what I think because central banks are conservative institutions and don't like to take risks. They'd rather adjust monetary policy until the short and long positions are roughly balanced. They'll keep nudging interest rates and money supply up and down until about the same number of people are taking short and long positions on nominal GDP growth.David Wright:
I we're out of time. I'm sorry, Scott. I have to end it there. My guest toda is Scott Sumner. Scott, thank you very much for being on the sScott Sumner:
Enjoyed it. Very good question.