The Sumner critique says that the Great Recession was more ‘caused’ by failure to supply enough money (when the market was clearly anticipating almost zero inflation) to keep NGDP on a level growth path, suggesting that the main path toward resolving this problem would be figuring some way to further convince the Federal Reserve to adopt NGDP level targeting. Also, as I understand it, even some of the people making money off shorting the housing market were quite loud about it—it’s not obvious that a lack of transparency is the problem. To make money you need to time your short correctly which the now-famous winners may have done by coincidence, and this is always the problem with bubbles. For that matter, it’s not obvious that the housing market would still have collapsed if the Fed had committed to an NGDP path.
Which, FWIW, is what happened. The question is, “why didn’t it happen sooner?” with perhaps a follow-up of “what’s the soonest we can reasonably expect these things to happen?”
Convincing the US Federal Reserve and the European Central Bank to adopt more intelligent monetary policy (something like NGDP targeting) would be big win for the world.
I’ll admit that I’m tragically under-read on the NGDP thesis, but it seems like a restatement of the old Phillips Curve nonsense that caused so much of the wreckage of the 1970s. It basically amounts to “inflate when times are bad to spike employment”, which does work in the short-term, but the long-term second-order effect of increased inflation expectations saps all the value out of it quickly.
Is there something I’m missing here, or is this just failed policy from 50 years ago with a new label on it?
First, I think ‘flexible inflation targeting’ is still the conventional wisdom, so I don’t think people have abandoned “inflate when times are bad to spike employment”.
Second, the case for NGDP targeting is more subtle than that. Market Monetarists claim that an NGDP targeting provides approximately the correct amount of money for the economy. A key difference is that one targets a time-path for the level of NGDP rather than the rate of growth.
There’s definitely something you’re missing, its not just a dressed up old theory. I wish I had a great starting point to start reading about this stuff, but I don’t. If you’re very interested I can try to find some especially good articles. This post of mine (and the ones it links to) tries to explain the process of monetary disequilibrium, which I think starts to give you some intuitions for why NGDP targeting might be a good idea (but is certainly not the whole story).
So the problem was that authorities didn’t (do enough to) prop up one of the 20th-century correlates of economic shock resistance? If only more money had entered the economy in the form of massive, cheap loans to much the same intermediaries that failed to make their business plans resilient against such shocks, the economy could have reorganized quickly into sustainable modes of production?
When NGDP/NGDI, the total amount of money the economy spends, goes down below trend, so does RGDP. If the Fed had e.g. bought government bonds, the amount of circulating money could have been prevented from going down (without any need for “fiscal stimulus”, by the way) without making loans to any particular trading companies (no need for bailouts! let them go bust!), and the bonds wouldn’t be monetized, they would be sold again as velocity picked up. I recommend http://themoneyillusion.com/ - your viewpoint here is too distant from sane macroeconomics, and too close to crazy populist economics, for me to tackle all the individual problems. But what you’re saying is more or less literally, exactly what people were saying about banks needing to take their medicine just before the Great Depression. Seriously, look it up, that was what made people realize that monetary velocity slowdowns needed to be made up by (temporary) monetary supply increases.
When NGDP/NGDI, the total amount of money the economy spends, goes down below trend, so does RGDP.
Only under “all else equal” assumptions, which are weakest in this situation. Any exogenous forcing of higher trading would imply that the extra trades have lower consumer surplus (perhaps negative relative to the no-forcing baseline).
In other words, if what you’re saying is true, then threatening to murder anyone who hoards more than 1% of their income (relative to a baseline year, etc) would also solve the problem. But then the cost of such a policy would be too large relative to its benefits for the most diehard anti-Sayer to ignore.
I recommend http://themoneyillusion.com/ - your viewpoint here is too distant from sane macroeconomics, and too close to crazy populist economics, for me to tackle all the individual problems.
You’re acting like I haven’t already been to the site or thought about these issues, when I actually followed it for the entire time Sumner was making his case, hence why I could make comments like in this thread—which I, for my part, recommend.
FWIW, sane macro doesn’t have the implications that interest rates can be forced to zero without cost, or any money not immediately spent is pure waste, or that the temporary surge in economic indicators you get from looting the rich is evidence of policy success.
But what you’re saying is more or less literally, exactly what people were saying about banks needing to take their medicine just before the Great Depression.
So? They were just as correct then: people suddenly expected (and expect now, less suddenly) to have their cake and eat it too with banking. Specifically, they wanted their banks to provide them with instant, guarateed access to their money, while also investing it in illiquid, return-generating ventures. That inevitably leads to situations where banks can’t redeem the investments for cash, and yet people do want it in that form. That’s not the basis for a sane economy.
Yes, we’d all be better off if the short-long trade was outlawed and banks issued bonds on a liquid market, possibly insured by private counterparties, to their customers, instead of claiming that their money was available on demand. But nobody is actually doing that, or has any incremental incentive to adopt it so long as governments supply free insurance, and we have to consider what second-best options are available. Money has no intrinsic value (I assume we both take this as axiomatic) and the utility ‘money’ provides to civilization comes from increasing the number of positive-sum trades. When people attempting to use money as a guaranteed store of value keep that money instead of spending it, the velocity of positive-sum trades goes down. This doesn’t mean that they’re evil hoarders, though I do think that preferring money as a store of value generally indicates something wrong. But it does mean that supplying more money is a positive-sum move because it increases the number of positive-sum trades occurring, so long as there is significant unutilized capacity.
To the degree that money is used as a store of value, the money supply available for ‘positive-sum’ trades decreases. Let us say that the supply of goods and services on the market stays the same, then with less money available to potentially purchase theses goods and services, the price of the goods and services decreases; microeconomics supply and demand curve. This incentivizes people who are not holding money as a store of value to participate in more positive-sum trades.
Of course, people might end up taking their store-of-value money and investing it, allowing the creation of capital goods that make more efficient production possible. But that’s another story.
Hence the standard belief that reluctance to lower nominal prices, or delay in lowering nominal prices, is key (along with nominal debt contracts) to explaining the observed fact that deflation is destructive of RGDP, which is why Scott Sumner’s blog is called “The Money Illusion”.
This doesn’t mean that they’re evil hoarders, though I do think that preferring money as a store of value generally indicates something wrong.
Well, that’s the core of our disagreement (indeed, my disagreement with anyone who buys into either the Sumner or Krugman view).
Specifically, on what basis can you even begin to make a case that there is “too much” hoarding without a framework for tabulating its benefits? That’s like saying that someone isn’t “eating healthy enough” while only counting the benefits of eating healthy.
And yet (as best demonstrated by the Landsburg link), any time such proponents are asked how to tabulate the relative benefits of hoarding to hoarders against its social costs, the answer is somwhere between silence and “assume we want to get people spending”.
That should be a much bigger red flag for you in the opposite direction.
I don’t think you’re engaging your opponents strongest arguments here. Yes most proponents (even economists or scott sumner) of active monetary policy can’t articulate why it might be a good idea that is solidly grounded, but other people can (myself, nick rowe, other market monetarists or the austrian monetary desequilibriumists), and you should engage the arguments they make.
I know you know of a way to tabulate the benefits because we’ve had extensive discussions about it. The benefits are the marginal utility of holding money (convenience etc.), the costs/benefits are the changes in money holdings you impose on other people when you trade or don’t trade with them.
I don’t think you’re engaging your opponents strongest arguments here. Yes most proponents (even economists or scott sumner) of active monetary policy can’t articulate why it might be a good idea that is solidly grounded, but other people can (myself, nick rowe, other market monetarists or the austrian monetary desequilibriumists), and you should engage the arguments they make.
Hold on—Sumner is the one everyone refers back to when advocating monetary policies he likes, he’s blogged about it for years, at tremendous length, starting from the crisis, he’s become (supposedly) influential in policy circles, and you’re saying I’m attacking a weak point? Sumner is exactly who I should be engaging, which is why it’s all the more saddening that his arguments fail simple checks:
Doesn’t this mean the Fed should be loaning to any ol’ person who can put up the collateral at 0%, not just large banks?
Why is it bad to “hoard” money for a year, but not five days? Why 5% NGDP growth and not some other number?
Why is it a “crisis” when interbank short-term loan rates “spike” from 4% to 6%? That just means their borrowing cost went up in the parts per million. (His entire response to this was something like “yes, ‘for want of a nail’ and all that”.)
If more exchanges are good, why aren’t hyperinflation scenarios (in which people treat money as a hot potato) a social optimum? (Or lesser scenarios that make people trade more than they otherwise would, like bans on home cooking.)
I know you know of a way to tabulate the benefits because we’ve had extensive discussions about it. The benefits are the marginal utility of holding money (convenience etc.), the costs/benefits are the changes in money holdings you impose on other people when you trade or don’t trade with them.
No, those weren’t the main benefits of holding money, and your citing of them means I somehow didn’t communicate the benefits to you. On your specific illustrative examples of convenience in our past discussion, I found them to completely assume away the important social benefits of hoarding.
In short, every example you gave was a case in which people knew in advance what they were going to spend the money on (e.g. having money for the bus). But these are emphatically opposite of the cases where money is most important and uniquely valuable—i.e., when you don’t yet know what you wish to redeem the money for and prefer to keep its option value.
The main benefit of holding money is that it signals the higher social demand for option value, which in turn is indicative of “discoordination”, or the lack of confidence that people can find stable comparative advantages. Money is only the extreme end of a scale that includes goods of increasingly multiple use—but in such scenarios, anything if valued if and to the extend that it will be useful in a broader array of situations.
You can suppress that signal, but only by worsening the economic allocation problem, just as you can suppress spiking oil prices, but only by shifting around the inefficiencies.
So I don’t think anyone, including you, has really engaged with the benefits of hoarding and so don’t have a satisfactory framework for evaluating whether there’s “too much”.
I phrased that badly. You’re not engaging a weak point so much as not steelmanning, which is what you should be doing. Fix your opponents arguments.
Yes, everyone refers to Sumner. He is the popularizer of market monetarism. Yes, Sumner doesn’t produce a defense of his views that is solidly grounded in economics. He is none the less vastly better than most people talking about this.
You should say things like ‘The best arguments for Sumners views are made by Nick Rowe and others (link to work or something) and are sometimes called ‘monetary disequilibrium theory’, however I think that this theory misses the important effect of Discoordination which works like this …’.
Basically no one discussing monetary economics (including the Austrians and others against active monetary policy) manages to ground their arguments in solid theory except Nick Rowe, a couple other market monetarists and the monetary disequilibriumist Austrians. The quality of debate is bad, but that doesn’t mean its OK for you not to engage the strongest arguments.
I found them to completely assume away the important social benefits of hoarding.
And that’s fair enough. I didn’t get that you were referring to that.
However,It is highly misleading to say
I don’t think anyone, including you, has really engaged with the benefits of hoarding and so don’t have a satisfactory framework for evaluating whether there’s “too much”.”
Because monetary disequilibrium is an internally consistent theory which does talk about the benefits of holding money and which can assess whether people hold ‘too much’ or not within the theory. You do not appear to deny this.
As far as I can tell, you additionally claim that this theory does not describe the important effects of Discoordination which provide additional social benefits to holding money. This is well and good! You should acknowledge what monetary disequilibrium theory does do well, and then attempt to improve upon its deficiencies. I’m actually very interested in hearing those arguments!
Your previous posts do not make this at all clear that this is what you were arguing, even to me. Since I probably know more about your views than anyone else other than you, this means they were probably also unclear to everyone else.
You’re speaking language I don’t recognize. This has nothing to do with ‘hoarding’. It’s about sticky nominal prices and fixed nominally priced debt contracts implying that when monetary velocity goes down, monetary supply should be increased to maintain the velocity of positive-sum trade at full capacity. Nothing you’ve said contradicts this, at least not in any language I recognize.
I would say your language is at least as unrecognizable if you’re going to propose measures to stop people from hoar… not spending money fast enough, while saying the problem “isn’t about that” (another red flag term), but instead “really about” this other new thing you just brought up.
Regardless, about that thing: yes, it certainly sucks when you can’t retroactively change the numbers in a contract you signed, and don’t understand how it embrittles your business plan to guarantee a stream of payments like that. However, sane policies should not favor those who failed to plan against eventualities, nor are sane economies predicated thereon.
I’m afraid that you do have to employ the concept of hoarding (or some isomorphic one) when you want to claim that welfare-enhan… positive sum trades are maximized when NGDP grows at n% like clockwork, regardless of how unmoored the economy has become, for some value of n that Sumner pulled out of thin air.
FWIW, sane macro doesn’t have the implications that interest rates can be forced to zero without cost, or any money not immediately spent is pure waste,
Remember our conversation! Sometimes negative interest rates are appropriate. If on the margin its impossible to make more investments that have a positive return, and people still want to transfer more wealth to the future, then negative rates reflect the actual social costs and benefits of holding money.
Any exogenous forcing of higher trading would imply that the extra trades have lower consumer surplus (perhaps negative relative to the no-forcing baseline).
Also, I don’t understand why you insist on talking about adding more money to the economy as ‘forcing of higher trading’. Remember that money is a product, and that it can be correct to increase or decrease the supply in responses to changes in demand.
Remember our conversation! Sometimes negative interest rates are appropriate.
Even so, the case you describe wouldn’t require central banking to bring about zero interest rates, and yet their action is needed to do exactly that.
Also, I don’t understand why you insist on talking about adding more money to the economy as ‘forcing of higher trading’.
Then why do you understand why it’s talked about in exactly that manner, i.e. used to prop up total nominal expenditures? How do you differentiate e.g. quantitative easing from from the murder policy I described, or the “ban on home cooking” that I described in past discussions with you. Both get people to spend more money.
Because one attempts to solve a shortage of money by adding more money and one tries to solve the same problem by much worse means?
But the dynamic is exactly the same! People are making all trades in which the stuff they get is better than enduring the “punishment” (arrest or loss of savings). This means the fraction of trades that only happened because of this policy are not, like other trades, Pareto-optimal. The fact that the means are better or worse among them (in gross terms) does not matter; they all are undermining the basis on which we can conclude that trades are positive sum. If your solution assumes away this loss of utility, you can get away with anything.
I’m confused why we’re having this argument again, I thought I had more or less convinced you on these issues.
What gave you that impression? I thought we reached pretty intractable points, in that our core disagreement hinges on the question of the economic role of liquidity in goods, and the extent to which the market can thereby signal “discoordination” (a term you also claimed doesn’t operationalize).
2 - Yeah, you’re right, I was misremembering. We weren’t able to come up with a concrete description of discoordination (which is a new concept that is unknown in the academic literature including austrian as far as we know) that made sense to me, and I lacked the intuition about it you have. IIRC I was able to convince you that ignoring discoordination effects, minimizing monetary disequilibrium is the thing to do, and NGDP targeting does a decent job of that. You also convinced me there was a case where it wont (when there’s a big shift in market to non-market activities or vice versa). Is that right?
1 - Given our disagreement in 2, I think you should still agree with me here since you haven’t brought up any discoordination related arguments as far as I can tell.
To get pareto optimality you have to ignore any monetary externalities (i.e when I buy buy something, I increase the quantity of money you have and my decision does not take this into account) or assume the quantity of money is optimal. And if we’re doing this then: given the quantity of money, we’re going to get pareto optimality no matter what. In other words, the trades in the no-monetary-change and monetary-change case are going to have the same pareto optimality. All trade in both scenarios is voluntary, so given the quantity of money each trade is welfare improving in the same sense in both cases. So our basis for concluding the trades are positive sum is unaffected.
The effects of a monetary change occur through an externality.
There actually is something a central bank can do when it’s up against the zero lower bound, but it’s risky. The interest rate won’t always be zero, so if the central bank can credibly promise to create high levels of future inflation by keeping interest rates low even after the economy recovers—to “credibly promise to be irresponsible”—it can affect real interest rates today. But good luck getting investors to believe a central bank is going to keep the proverbial printing presses going even beyond the point where it starts doing more harm than good.
It’s much easier under these circumstances to do the same thing with fiscal policy rather than monetary policy: have the government borrow and spend to put idle resources to work directly, increasing NGDP by increasing real GDP. (This is how World War II caused the end of the Great Depression.)
I think you’re actually just not understanding Krugman right. He often writes misleadingly, so that’s understandable. The Fed’s normal tools perhaps can’t be used to do that (but they probably can), but the Fed’s normal tools are stupid and it’s easy to produce inflation with slightly different tools (negative interest rates on reserves, level targeting, targeting the forecast etc., credible promise to keep interest rates low). This isn’t some new theory, Krugman has a famous paper about how the Japanese Central Bank could do exactly that.
Having a target and achieving the target are definitely different things, but achieving an NGDP target is not difficult because central banks have approximately infinite control over a single future nominal variable (e.g. NGDP, nominal interest rate or price level).
One of the big lessons from central banks response to this recession is that interest rates and inflation are a very misleading and confusing ways to talk about monetary policy.
It’s much easier under these circumstances to do the same thing with fiscal policy rather than monetary policy: have the government borrow and spend to put idle resources to work directly, increasing NGDP by increasing real GDP. (This is how World War II caused the end of the Great Depression.)
If I’m understanding my Sumner right, Krugman is just plain wrong about this. Central banks can decrease interest rates, promise to keep future interest rates low, engage in quantitative easing, charge negative interest on reserves, and print physical money and drop it out of helicopters. “There is no zero bound” is a market monetarist slogan and I have to say it sounds a tad plausible from over here.
On this point, I’d just have to chalk it up as “beyond my current expertise”. (Part of Krugman’s argument is that although “unconventional” monetary policy is possible, political and other considerations make it much harder to do.)
Holy crap: that article has “get to the point”:fluff ratio of about 5%. I appreciate the link, but just want to warn readers they need to skip past a lot of meandering about scenes from NYC to get the answer to the question.
As a blanket presumption, that’s dumb and you know it.
Certainly printing money is a near costless way to solve economic problems at some times otherwise we wouldn’t use money at all. Sumner doesn’t advocate printing money at all times and places, he advocates printing or destroying money until there is the correct amount (which he suggests is when NGDP is on-trend).
As a blanket presumption, that’s dumb and you know it.
I think you’re confusing “presumption” with “categorial principle” or “axiom” or something. A presumption can be overridden, given sufficient evidence; my question was whether or how strongly EY takes a presumption against it. If you agree with the concept of criminalizing counterfeiting, you agree with that presumption.
Certainly printing money is a near costless way to solve economic problems at some times otherwise we wouldn’t use money at all.
Does not follow. Not all moneys arose in a Pareto-optimal fashion, so their use not evidence printing money being costless.
The problem isn’t that the Fed can’t reach (from low to hit) NGDP targets; the problem is that anyone can scheme to raise NGDP by conducting dummy transactions that nevertheless go on record in economic accounting as higher NGDP. At which point NGDP loses its evidential and causal power through Lucas/Goodhart effects.
Can there even exist a way to fix this without abandoning GDP-like indicators in favor of something closer to human level? Shifting resources around among enterprises, at higher and higher volume, is not the kind of economic activity worth wanting.
You are exactly correct. Policy must switch to a less gameable metric in order to avoid situations like that. But it’s important to remember that the problem isn’t just the potential for schemers, but that, even without schemers, the extra trades you’ve forced to goose NGDP are even less indicative of economic goodness than the normal “GDP = good” assumption requires!
And I don’t think anyone can come up with such a metric without solving the (holy grail level) problem of microeconomic foundations for macro.
Wait, now I’m confused. Micro-foundations for monetary economics exist, and I know you know, because we’ve talked about them at length. Maybe you mean that there’s more to macro beyond monetary economics?
I’m referring the commonly-known problem of deriving macro results from micro, which we also talked about in those exchanges, and during which you rejected the claim that failing to so derive results is a reason to reject the macro conclusions. If you remember differently, give a link.
I think we discussed that we haven’t seen the AS/AD models derived from micro principles. I’m not sure this can’t be done, its just not commonly discussed, and I don’t see a good use for those concepts so I haven’t tried to find a derivation.
The Sumner critique says that the Great Recession was more ‘caused’ by failure to supply enough money (when the market was clearly anticipating almost zero inflation) to keep NGDP on a level growth path, suggesting that the main path toward resolving this problem would be figuring some way to further convince the Federal Reserve to adopt NGDP level targeting. Also, as I understand it, even some of the people making money off shorting the housing market were quite loud about it—it’s not obvious that a lack of transparency is the problem. To make money you need to time your short correctly which the now-famous winners may have done by coincidence, and this is always the problem with bubbles. For that matter, it’s not obvious that the housing market would still have collapsed if the Fed had committed to an NGDP path.
Yes, people who short are usually loud about it. The sooner people agree with them, the sooner the bubble pops.
Which, FWIW, is what happened. The question is, “why didn’t it happen sooner?” with perhaps a follow-up of “what’s the soonest we can reasonably expect these things to happen?”
Convincing the US Federal Reserve and the European Central Bank to adopt more intelligent monetary policy (something like NGDP targeting) would be big win for the world.
I’ll admit that I’m tragically under-read on the NGDP thesis, but it seems like a restatement of the old Phillips Curve nonsense that caused so much of the wreckage of the 1970s. It basically amounts to “inflate when times are bad to spike employment”, which does work in the short-term, but the long-term second-order effect of increased inflation expectations saps all the value out of it quickly.
Is there something I’m missing here, or is this just failed policy from 50 years ago with a new label on it?
First, I think ‘flexible inflation targeting’ is still the conventional wisdom, so I don’t think people have abandoned “inflate when times are bad to spike employment”.
Second, the case for NGDP targeting is more subtle than that. Market Monetarists claim that an NGDP targeting provides approximately the correct amount of money for the economy. A key difference is that one targets a time-path for the level of NGDP rather than the rate of growth.
There’s definitely something you’re missing, its not just a dressed up old theory. I wish I had a great starting point to start reading about this stuff, but I don’t. If you’re very interested I can try to find some especially good articles. This post of mine (and the ones it links to) tries to explain the process of monetary disequilibrium, which I think starts to give you some intuitions for why NGDP targeting might be a good idea (but is certainly not the whole story).
Thanks Eliezer. I’ll look into this.
So the problem was that authorities didn’t (do enough to) prop up one of the 20th-century correlates of economic shock resistance? If only more money had entered the economy in the form of massive, cheap loans to much the same intermediaries that failed to make their business plans resilient against such shocks, the economy could have reorganized quickly into sustainable modes of production?
When NGDP/NGDI, the total amount of money the economy spends, goes down below trend, so does RGDP. If the Fed had e.g. bought government bonds, the amount of circulating money could have been prevented from going down (without any need for “fiscal stimulus”, by the way) without making loans to any particular trading companies (no need for bailouts! let them go bust!), and the bonds wouldn’t be monetized, they would be sold again as velocity picked up. I recommend http://themoneyillusion.com/ - your viewpoint here is too distant from sane macroeconomics, and too close to crazy populist economics, for me to tackle all the individual problems. But what you’re saying is more or less literally, exactly what people were saying about banks needing to take their medicine just before the Great Depression. Seriously, look it up, that was what made people realize that monetary velocity slowdowns needed to be made up by (temporary) monetary supply increases.
Only under “all else equal” assumptions, which are weakest in this situation. Any exogenous forcing of higher trading would imply that the extra trades have lower consumer surplus (perhaps negative relative to the no-forcing baseline).
In other words, if what you’re saying is true, then threatening to murder anyone who hoards more than 1% of their income (relative to a baseline year, etc) would also solve the problem. But then the cost of such a policy would be too large relative to its benefits for the most diehard anti-Sayer to ignore.
You’re acting like I haven’t already been to the site or thought about these issues, when I actually followed it for the entire time Sumner was making his case, hence why I could make comments like in this thread—which I, for my part, recommend.
FWIW, sane macro doesn’t have the implications that interest rates can be forced to zero without cost, or any money not immediately spent is pure waste, or that the temporary surge in economic indicators you get from looting the rich is evidence of policy success.
So? They were just as correct then: people suddenly expected (and expect now, less suddenly) to have their cake and eat it too with banking. Specifically, they wanted their banks to provide them with instant, guarateed access to their money, while also investing it in illiquid, return-generating ventures. That inevitably leads to situations where banks can’t redeem the investments for cash, and yet people do want it in that form. That’s not the basis for a sane economy.
Yes, we’d all be better off if the short-long trade was outlawed and banks issued bonds on a liquid market, possibly insured by private counterparties, to their customers, instead of claiming that their money was available on demand. But nobody is actually doing that, or has any incremental incentive to adopt it so long as governments supply free insurance, and we have to consider what second-best options are available. Money has no intrinsic value (I assume we both take this as axiomatic) and the utility ‘money’ provides to civilization comes from increasing the number of positive-sum trades. When people attempting to use money as a guaranteed store of value keep that money instead of spending it, the velocity of positive-sum trades goes down. This doesn’t mean that they’re evil hoarders, though I do think that preferring money as a store of value generally indicates something wrong. But it does mean that supplying more money is a positive-sum move because it increases the number of positive-sum trades occurring, so long as there is significant unutilized capacity.
To the degree that money is used as a store of value, the money supply available for ‘positive-sum’ trades decreases. Let us say that the supply of goods and services on the market stays the same, then with less money available to potentially purchase theses goods and services, the price of the goods and services decreases; microeconomics supply and demand curve. This incentivizes people who are not holding money as a store of value to participate in more positive-sum trades.
Of course, people might end up taking their store-of-value money and investing it, allowing the creation of capital goods that make more efficient production possible. But that’s another story.
Specifically, then they wouldn’t be using money as a store of value anymore since they wouldn’t be holding money but securities.
Hence the standard belief that reluctance to lower nominal prices, or delay in lowering nominal prices, is key (along with nominal debt contracts) to explaining the observed fact that deflation is destructive of RGDP, which is why Scott Sumner’s blog is called “The Money Illusion”.
Well, that’s the core of our disagreement (indeed, my disagreement with anyone who buys into either the Sumner or Krugman view).
Specifically, on what basis can you even begin to make a case that there is “too much” hoarding without a framework for tabulating its benefits? That’s like saying that someone isn’t “eating healthy enough” while only counting the benefits of eating healthy.
And yet (as best demonstrated by the Landsburg link), any time such proponents are asked how to tabulate the relative benefits of hoarding to hoarders against its social costs, the answer is somwhere between silence and “assume we want to get people spending”.
That should be a much bigger red flag for you in the opposite direction.
I don’t think you’re engaging your opponents strongest arguments here. Yes most proponents (even economists or scott sumner) of active monetary policy can’t articulate why it might be a good idea that is solidly grounded, but other people can (myself, nick rowe, other market monetarists or the austrian monetary desequilibriumists), and you should engage the arguments they make.
I know you know of a way to tabulate the benefits because we’ve had extensive discussions about it. The benefits are the marginal utility of holding money (convenience etc.), the costs/benefits are the changes in money holdings you impose on other people when you trade or don’t trade with them.
Btw, I don’t remember if you read the write up I did of my simple mathematical model of monetary disequilibrium.
Hold on—Sumner is the one everyone refers back to when advocating monetary policies he likes, he’s blogged about it for years, at tremendous length, starting from the crisis, he’s become (supposedly) influential in policy circles, and you’re saying I’m attacking a weak point? Sumner is exactly who I should be engaging, which is why it’s all the more saddening that his arguments fail simple checks:
Doesn’t this mean the Fed should be loaning to any ol’ person who can put up the collateral at 0%, not just large banks?
Why is it bad to “hoard” money for a year, but not five days? Why 5% NGDP growth and not some other number?
Why is it a “crisis” when interbank short-term loan rates “spike” from 4% to 6%? That just means their borrowing cost went up in the parts per million. (His entire response to this was something like “yes, ‘for want of a nail’ and all that”.)
If more exchanges are good, why aren’t hyperinflation scenarios (in which people treat money as a hot potato) a social optimum? (Or lesser scenarios that make people trade more than they otherwise would, like bans on home cooking.)
No, those weren’t the main benefits of holding money, and your citing of them means I somehow didn’t communicate the benefits to you. On your specific illustrative examples of convenience in our past discussion, I found them to completely assume away the important social benefits of hoarding.
In short, every example you gave was a case in which people knew in advance what they were going to spend the money on (e.g. having money for the bus). But these are emphatically opposite of the cases where money is most important and uniquely valuable—i.e., when you don’t yet know what you wish to redeem the money for and prefer to keep its option value.
The main benefit of holding money is that it signals the higher social demand for option value, which in turn is indicative of “discoordination”, or the lack of confidence that people can find stable comparative advantages. Money is only the extreme end of a scale that includes goods of increasingly multiple use—but in such scenarios, anything if valued if and to the extend that it will be useful in a broader array of situations.
You can suppress that signal, but only by worsening the economic allocation problem, just as you can suppress spiking oil prices, but only by shifting around the inefficiencies.
So I don’t think anyone, including you, has really engaged with the benefits of hoarding and so don’t have a satisfactory framework for evaluating whether there’s “too much”.
I phrased that badly. You’re not engaging a weak point so much as not steelmanning, which is what you should be doing. Fix your opponents arguments.
Yes, everyone refers to Sumner. He is the popularizer of market monetarism. Yes, Sumner doesn’t produce a defense of his views that is solidly grounded in economics. He is none the less vastly better than most people talking about this.
You should say things like ‘The best arguments for Sumners views are made by Nick Rowe and others (link to work or something) and are sometimes called ‘monetary disequilibrium theory’, however I think that this theory misses the important effect of Discoordination which works like this …’.
Basically no one discussing monetary economics (including the Austrians and others against active monetary policy) manages to ground their arguments in solid theory except Nick Rowe, a couple other market monetarists and the monetary disequilibriumist Austrians. The quality of debate is bad, but that doesn’t mean its OK for you not to engage the strongest arguments.
And that’s fair enough. I didn’t get that you were referring to that.
However,It is highly misleading to say
Because monetary disequilibrium is an internally consistent theory which does talk about the benefits of holding money and which can assess whether people hold ‘too much’ or not within the theory. You do not appear to deny this.
As far as I can tell, you additionally claim that this theory does not describe the important effects of Discoordination which provide additional social benefits to holding money. This is well and good! You should acknowledge what monetary disequilibrium theory does do well, and then attempt to improve upon its deficiencies. I’m actually very interested in hearing those arguments!
Your previous posts do not make this at all clear that this is what you were arguing, even to me. Since I probably know more about your views than anyone else other than you, this means they were probably also unclear to everyone else.
You’re speaking language I don’t recognize. This has nothing to do with ‘hoarding’. It’s about sticky nominal prices and fixed nominally priced debt contracts implying that when monetary velocity goes down, monetary supply should be increased to maintain the velocity of positive-sum trade at full capacity. Nothing you’ve said contradicts this, at least not in any language I recognize.
I would say your language is at least as unrecognizable if you’re going to propose measures to stop people from hoar… not spending money fast enough, while saying the problem “isn’t about that” (another red flag term), but instead “really about” this other new thing you just brought up.
Regardless, about that thing: yes, it certainly sucks when you can’t retroactively change the numbers in a contract you signed, and don’t understand how it embrittles your business plan to guarantee a stream of payments like that. However, sane policies should not favor those who failed to plan against eventualities, nor are sane economies predicated thereon.
I’m afraid that you do have to employ the concept of hoarding (or some isomorphic one) when you want to claim that welfare-enhan… positive sum trades are maximized when NGDP grows at n% like clockwork, regardless of how unmoored the economy has become, for some value of n that Sumner pulled out of thin air.
Remember our conversation! Sometimes negative interest rates are appropriate. If on the margin its impossible to make more investments that have a positive return, and people still want to transfer more wealth to the future, then negative rates reflect the actual social costs and benefits of holding money.
Also, I don’t understand why you insist on talking about adding more money to the economy as ‘forcing of higher trading’. Remember that money is a product, and that it can be correct to increase or decrease the supply in responses to changes in demand.
Even so, the case you describe wouldn’t require central banking to bring about zero interest rates, and yet their action is needed to do exactly that.
Then why do you understand why it’s talked about in exactly that manner, i.e. used to prop up total nominal expenditures? How do you differentiate e.g. quantitative easing from from the murder policy I described, or the “ban on home cooking” that I described in past discussions with you. Both get people to spend more money.
1) Fair enough, I think targeting US bond interest rates is probably a bad policy too.
2) Because one attempts to solve a shortage of money by adding more money and one tries to solve the same problem by much worse means?
(I’m confused why we’re having this argument again, I thought I had more or less convinced you on these issues).
But the dynamic is exactly the same! People are making all trades in which the stuff they get is better than enduring the “punishment” (arrest or loss of savings). This means the fraction of trades that only happened because of this policy are not, like other trades, Pareto-optimal. The fact that the means are better or worse among them (in gross terms) does not matter; they all are undermining the basis on which we can conclude that trades are positive sum. If your solution assumes away this loss of utility, you can get away with anything.
What gave you that impression? I thought we reached pretty intractable points, in that our core disagreement hinges on the question of the economic role of liquidity in goods, and the extent to which the market can thereby signal “discoordination” (a term you also claimed doesn’t operationalize).
2 - Yeah, you’re right, I was misremembering. We weren’t able to come up with a concrete description of discoordination (which is a new concept that is unknown in the academic literature including austrian as far as we know) that made sense to me, and I lacked the intuition about it you have. IIRC I was able to convince you that ignoring discoordination effects, minimizing monetary disequilibrium is the thing to do, and NGDP targeting does a decent job of that. You also convinced me there was a case where it wont (when there’s a big shift in market to non-market activities or vice versa). Is that right?
1 - Given our disagreement in 2, I think you should still agree with me here since you haven’t brought up any discoordination related arguments as far as I can tell.
To get pareto optimality you have to ignore any monetary externalities (i.e when I buy buy something, I increase the quantity of money you have and my decision does not take this into account) or assume the quantity of money is optimal. And if we’re doing this then: given the quantity of money, we’re going to get pareto optimality no matter what. In other words, the trades in the no-monetary-change and monetary-change case are going to have the same pareto optimality. All trade in both scenarios is voluntary, so given the quantity of money each trade is welfare improving in the same sense in both cases. So our basis for concluding the trades are positive sum is unaffected.
The effects of a monetary change occur through an externality.
Didn’t we already have a long debate about this? Didn’t I mostly convince you that NGDP stabilization was a good idea?
If I’m understanding my Krugman right, declaring an NGDP target and actually achieving that target are two different things. The Fed’s normal tools can’t actually produce any inflation when the economy is depressed and interest rates are at zero. - increasing the money supply will just lead to more cash under Apple Computer’s proverbial mattress instead of increased NGDP.
There actually is something a central bank can do when it’s up against the zero lower bound, but it’s risky. The interest rate won’t always be zero, so if the central bank can credibly promise to create high levels of future inflation by keeping interest rates low even after the economy recovers—to “credibly promise to be irresponsible”—it can affect real interest rates today. But good luck getting investors to believe a central bank is going to keep the proverbial printing presses going even beyond the point where it starts doing more harm than good.
It’s much easier under these circumstances to do the same thing with fiscal policy rather than monetary policy: have the government borrow and spend to put idle resources to work directly, increasing NGDP by increasing real GDP. (This is how World War II caused the end of the Great Depression.)
I think you’re actually just not understanding Krugman right. He often writes misleadingly, so that’s understandable. The Fed’s normal tools perhaps can’t be used to do that (but they probably can), but the Fed’s normal tools are stupid and it’s easy to produce inflation with slightly different tools (negative interest rates on reserves, level targeting, targeting the forecast etc., credible promise to keep interest rates low). This isn’t some new theory, Krugman has a famous paper about how the Japanese Central Bank could do exactly that.
Having a target and achieving the target are definitely different things, but achieving an NGDP target is not difficult because central banks have approximately infinite control over a single future nominal variable (e.g. NGDP, nominal interest rate or price level).
One of the big lessons from central banks response to this recession is that interest rates and inflation are a very misleading and confusing ways to talk about monetary policy.
This is likely an incorrect understanding.
If I’m understanding my Sumner right, Krugman is just plain wrong about this. Central banks can decrease interest rates, promise to keep future interest rates low, engage in quantitative easing, charge negative interest on reserves, and print physical money and drop it out of helicopters. “There is no zero bound” is a market monetarist slogan and I have to say it sounds a tad plausible from over here.
On this point, I’d just have to chalk it up as “beyond my current expertise”. (Part of Krugman’s argument is that although “unconventional” monetary policy is possible, political and other considerations make it much harder to do.)
My understanding is that Krugman is not wrong, he just writes correctly but very misleadingly for reasons that are not totally clear.
Oh, I think his reasons are pretty clear. They basically amount to politics.
I found the article “The Deflationist: How Paul Krugman found politics” educational on that point.
Holy crap: that article has “get to the point”:fluff ratio of about 5%. I appreciate the link, but just want to warn readers they need to skip past a lot of meandering about scenes from NYC to get the answer to the question.
Plausible enough to overcome the presumption against “printing money is a costless way to solve economic problems”?
As a blanket presumption, that’s dumb and you know it.
Certainly printing money is a near costless way to solve economic problems at some times otherwise we wouldn’t use money at all. Sumner doesn’t advocate printing money at all times and places, he advocates printing or destroying money until there is the correct amount (which he suggests is when NGDP is on-trend).
I think you’re confusing “presumption” with “categorial principle” or “axiom” or something. A presumption can be overridden, given sufficient evidence; my question was whether or how strongly EY takes a presumption against it. If you agree with the concept of criminalizing counterfeiting, you agree with that presumption.
Does not follow. Not all moneys arose in a Pareto-optimal fashion, so their use not evidence printing money being costless.
The problem isn’t that the Fed can’t reach (from low to hit) NGDP targets; the problem is that anyone can scheme to raise NGDP by conducting dummy transactions that nevertheless go on record in economic accounting as higher NGDP. At which point NGDP loses its evidential and causal power through Lucas/Goodhart effects.
Can there even exist a way to fix this without abandoning GDP-like indicators in favor of something closer to human level? Shifting resources around among enterprises, at higher and higher volume, is not the kind of economic activity worth wanting.
You are exactly correct. Policy must switch to a less gameable metric in order to avoid situations like that. But it’s important to remember that the problem isn’t just the potential for schemers, but that, even without schemers, the extra trades you’ve forced to goose NGDP are even less indicative of economic goodness than the normal “GDP = good” assumption requires!
And I don’t think anyone can come up with such a metric without solving the (holy grail level) problem of microeconomic foundations for macro.
Wait, now I’m confused. Micro-foundations for monetary economics exist, and I know you know, because we’ve talked about them at length. Maybe you mean that there’s more to macro beyond monetary economics?
I’m referring the commonly-known problem of deriving macro results from micro, which we also talked about in those exchanges, and during which you rejected the claim that failing to so derive results is a reason to reject the macro conclusions. If you remember differently, give a link.
Macro results are definitely derivable from micro principles: http://goodmorningeconomics.wordpress.com/2012/04/07/the-backrub-economy-a-simple-mathematical-model-of-monetary-disequilibrium/ .
I think we discussed that we haven’t seen the AS/AD models derived from micro principles. I’m not sure this can’t be done, its just not commonly discussed, and I don’t see a good use for those concepts so I haven’t tried to find a derivation.