Take that, Cantillon effect fans
Hot off the wire, once again Fed stimulus raises interest rates and lowers bond prices:
Treasuries fell after the Federal Reserve said it will buy $45 billion a month of U.S. government debt, expanding its asset-purchase program while linking its main interest rate to unemployment and inflation.
Only the Fed can make the price of something fall by purchasing more of it. Ironically, the Fed wants to lower long term interest rates.
But as Bob Dylan once said:
there’s no success like failure
Tags:
12. December 2012 at 10:33
So in the future when the Fed wants to start lowering rates, should they start selling securities?
12. December 2012 at 10:33
So the Fed should sell treasuries then right?
12. December 2012 at 10:37
You are really beating the living daylights of this straw man, Scott, providing ‘brilliant’ arguments against assertions that no one ever claimed were entailed by the fact of Cantillon’s recounting of imperfect prices adjustments.
Extended golf clap.
This is just genius stuff, Scott.
Ascribe views to who classes of economists who don’t hold those view, then ‘expose’ how stupid those economists are with refutations of views they don’t hold.
That’s what good economic blogging is all about, isn’t it?
12. December 2012 at 10:51
So Greg, you are saying that in your previous comments, you never said that after QE bond prices go up???
btw, did the Fed also kind of commited to an Evans rule?
12. December 2012 at 10:51
@ John
“So in the future when the Fed wants to start lowering rates, should they start selling securities?”
Possibly, yes.
Just like when the US got downgraded, demand for Treasuries went up.
12. December 2012 at 11:03
@Ssumner
Scott, that’s not a pure measure of a Cantillon effect. A pure measure would be if the Fed bought bonds vs. the Fed buying another asset class. If there are no cantillon effects, it should not matter what asset class was purchased, so long as the impact on expected NGDP was the same.
The above examples shows that the income effect from expectations of the Fed’s activities dominates the substitution effect.
Also, theoretically, the market should absorb the substitution effect all at once. However, if you look at the history of QEI and QEII, the immediate impact was a drop in Treas rates, but over the actual implementation we saw Treas rates rise as the Fed actually executed its huge volume of trades (there are, at these magnitudes, limitations to arbitrage). Krugman has commented on this in the past. However, it’s not clear the cause – is it because of the actual flow of Fed money into bonds (which the markets failed to arbitrage due to magnitude), OR is it because over the duration of QE1 and QE2 it became obvious that it wasn’t enough and the Fed wasn’t going to immediately extend the programs.
I certainly don’t know. But the Fed did not answer the Cantillon effect question today.
However, the Fed DID basically declare an NGDP target today. Remember, it is not LEGALLY allowed to do so (it needs to frame its actions according to its dual mandate), but it came about as close as you can to doing so within these legal boundaries.
12. December 2012 at 11:09
The thing to remember is that when the Fed buys bonds there are obviously sellers. But it is possible for there to be other sellers, who are selling to buyers other than the Fed.
How can the price of bonds go down when the Fed is buying them? Investors are selling more bonds than the Fed is buying.
The impact of money creation on spending, output, and prices all create an incentive to sell bonds for reasons other than the Fed buying them (and offering a higher price of them.)
12. December 2012 at 11:11
Scott
One definition of “crazy” is to repeat the same thing over and over and think the result will be different “this time”. That´s the Fed with the “we want to lower long term rates” spiel:
http://thefaintofheart.wordpress.com/2012/12/11/they-should-say-we-want-long-term-rates-to-rise/
12. December 2012 at 11:17
John, not quite. The Fed’s policy led to a rise in the _natural_ interest rate. When we talk about monetary easing as “cutting rates”, this should be understood as making the market rate lower, compared to the natural rate.
12. December 2012 at 11:27
It’s like the debate over the Cantillon effect did not even occur.
The debate was not whether t-bond prices rise or fall over time with additional inflation, relative to the baseline of no additional inflation.
The actual debate was whether it matters if the inflation enters the t-bond market first as opposed to other markets first.
If the Fed announced that the additional $45 billion inflation per month was instead going to be used to purchase gold, or tacos, or cars, then while t-bond prices would likely still have fallen, they would not have fallen by as much as they would have if the Fed spent $45 billion a month directly on the t-bonds themselves, i.e. what just took place.
The question “it matters where the money enters the economy first” requires us to stay in the context of money entering the economy somewhere, when considering alternatives, not just comparing additional inflation with no additional inflation at all, such that events like this are used to “prove” false an argument that nobody who accepts the Cantillon effect is even making.
You accepted that if the Fed were to start buying gold, or tacos, or cars, then that it would make those prices rise higher than they otherwise would be if the Fed bought t-bonds instead. But does that mean if buying gold or tacos or cars was followed by falling gold, taco and car prices over time, that you would be wrong? No, you would not be wrong. We would just say that the prices fell by less than they otherwise would have fallen had the Fed bought t-bonds instead!
The Cantillon effect is at root a counter-factually grounded, that is, logically grounded, concept. It doesn’t make predictions of price trends of specific goods over time after an inflation event occurs. It only says that the economy is differently affected (i.e. individuals are differently benefited) if money inflation enters the economy at location X rather than location ~X. It does NOT say “the price of X will rise by Z% over time if the inflation enters at location X”, nor does it say “the price of ~X will rise by Z% over time if the inflation enters at location ~X.”
————————–
Imagine you owned a Sumner Bond. Imagine the Fed announced it is going to inflate somewhere in the economy. Ask yourself this: Given that your Sumner Bond is going to fall in price no matter where the Fed injects the money, will the fall be the exact same no matter where the Fed injects the money?
12. December 2012 at 11:32
Stats Guy,
Why does the Fed typically purchase securities when they want to lower the Federal Funds rate then? That’s what all the textbooks describe.
12. December 2012 at 11:33
Personally, I think it would be foolish to believe the fall in t-bond price would be the exact same no matter who received the new money first. Owning money affects an individual’s behavior. An investor of t-bonds isn’t going to price t-bonds exactly the same no matter where the Fed injects the new money. There is no constant mathematical formula of “Fed injects $X dollars, therefore t-bond prices rise f($X)%.”
12. December 2012 at 11:33
Statsguy, As you may know I’ve recently been bombarded with comments from Cantillon effect fans assuring me that when the Fed buys bonds interest rates on bonds fall. I certainly understand that the effect might have been slightly different if they’d bought MBSs instead, although of course any differences would be trivial compared to the macro effects of monetary policy.
The Fed most certainly did not institute an NGDP target today, nor will it do so for the foreseeable future. Today’s low interest rate promise could be kept with Japanese style growth over the next 20 years.
Bill, Fed policy is different because when they buy things they also increase the money supply and lower its value. When I buy something I don’t increase the money supply, so I only influence the value of the thing I buy, not money.
Marcus, Your prediction that long rates would not fall was correct, and most of the commenters here were wrong.
BTW, I’m sure Bernanke would argued that it’s real rates that matter.
12. December 2012 at 11:36
The Bloomberg article also includes this:
“The Fed is losing some of its credibility as an inflation fighter,” Gary Pollack, who helps manage $12 billion as head of fixed-income trading at Deutsche Bank AG’s Private Wealth Management unit in New York. “They will allow inflation to go above the long-term target. That’s disappointing for the market.”
What is this Gary Pollack talking about?, Deutsche Bank of course, and does he make any sense?
Did the market show disappointment today??
12. December 2012 at 11:38
Stats Guy,
You and Scott don’t seem to be making a complete point. Short-term rates can go down and long term rates can go up (steepening yield curve). Also, short-term rates can rise while long term rates go down, inverted yield curves like this often happen before recessions. The fact that long and short term interest rates can move independently have nothing to do with Cantillon Effects. Cantillon Effects have to do with the fact that if everyone had their money doubled like in Hume’s famous thought experiment, not every price would double instantly and not everyone would be affected in the same way by the eventual doubling in prices. It’s not complicated stuff.
12. December 2012 at 11:39
StatsGuy:
Just like when the US got downgraded, demand for Treasuries went up.
Including the Fed’s demand, right?
Scott, that’s not a pure measure of a Cantillon effect. A pure measure would be if the Fed bought bonds vs. the Fed buying another asset class. If there are no cantillon effects, it should not matter what asset class was purchased
Bingo. The “problem” is that we can never observe that counter-factual “other asset” class, given they buy the “asset” class in question.
If there was a way to observe multiple futures, where each future is changed by only one thing, then we could statistically and empirically confirm the Cantillon effect. But alas, we only have this world to observe, which means some people are going to have to use a little more imagination than looking at t-bond prices after an inflation announcement, and thinking they’ve made a huge discovery.
12. December 2012 at 11:41
ssumner:
Statsguy, As you may know I’ve recently been bombarded with comments from Cantillon effect fans assuring me that when the Fed buys bonds interest rates on bonds fall.
…from what they otherwise would have fallen by had the Fed bought other assets instead.
I certainly understand that the effect might have been slightly different if they’d bought MBSs instead, although of course any differences would be trivial compared to the macro effects of monetary policy.
Then you understand the Cantillon effect! Cantillon didn’t care about macro aggregates when he argued his case.
12. December 2012 at 11:49
This is pretty good evidence. Weird how QE gets reported as being done to drive down interest rates across all types of assets (read UK economic press, or City economists) if it doesn’t in fact do that though. Do things work differently for UK and US?
12. December 2012 at 11:53
Scott:
“Only the Fed can make the price of something fall by purchasing more of it.”
No, not even the Fed can do that. Forward guidance dropped bond prices; QE is still doing nothing!
12. December 2012 at 11:56
I know I can’t prove this now, but if the Fed’s announcement was subsequently followed by a temporal nominal rise in t-bond prices, then I would not have said “Take that Cantillon deniers!”
I would have said that the rise in prices were higher than they otherwise would have been had the Fed bought something other than t-bonds instead. I wouldn’t say the Fed made the t-bonds prices rise absolutely.
This whole dispute seems to be grounded on the rather avoidable misunderstanding that one side was thinking temporally and nominally, while the other side was thinking counter-factually and relatively. I thought everyone was thinking counter-factually, but it seems that wasn’t the case.
12. December 2012 at 12:24
John, You said;
“Cantillon Effects have to do with the fact that if everyone had their money doubled like in Hume’s famous thought experiment, not every price would double instantly and not everyone would be affected in the same way by the eventual doubling in prices. It’s not complicated stuff.”
I’ve never seen people use the term ‘Cantillon effect’ in that way. That’s nothing more than mainstream econ. So it would be silly of Austrians to complain about the mainstream ignoring Cantillon effects.
Ben, It doesn’t always happen that way–it partly depends on whether the announcement was more or less than expected, and partly on whether the monetary injection is expected to be permanent.
12. December 2012 at 12:28
It’s comic gold to watch you provide what you think are counter examples to economics that you don’t know, Scott.
12. December 2012 at 13:08
Scott, coming on the heels of MF and Greg Ransom’s protests, you will surely downplay what I have to say. But anyway: Wasn’t this announcement exactly what the market was expecting the Fed to say? Or at the very least, you should be saying, “Aha! Since the Fed promised to buy more bonds than I think the market was expecting, we see bond prices fell–Murphy can’t explain that.”
When it comes to your own framework, you are always careful to be nuanced like this. Yet when it comes to blowing up Irish economists…
(Also, in case you say an expectations-based approach is your turf, and not something the Austrians can use: OK, but then this announcement means nothing, right? It’s not as if the Fed actually bought trillions of dollars worth of Treasuries today, right?)
12. December 2012 at 14:18
I think where people get lost is that you can imagine a scenario where the Fed buys enough of a particular security and the price of the security would go up faster than the inflation expectation pushes it down. But, I gather in practice, it is likely that private buyers would simply create synthetic securities to obtain a particular duration, rather than fight the Fed for the last bits, and this would spread that inflation effect over the other securities and hence no Cantillon effect.
All of this holds true unless you have a large group of significant market participants who are forced to hold a particular type of security or are holding such securities for reasons other than inflation, in this case you wouldn’t see market participants repricing for inflation expectations to the same extent. Here, we can imagine a case where money gets pushed out, but instead of searching for yield it gives up and lands as excess reserves and the price level doesn’t change because you have only printed the money and stuffed it in a warehouse and yields go down despite increased inflation expectations.
Why would you give up on yield? a) because you don’t have as much price risk in cash; and b), if you are holding risk free or nearly risk free assets, you clearly don’t have good positive NPV projects in which to invest and it is the lending mechanism which moves the excess cash from those who can’t find a good place for it to those who will. And, we continue to have excess reserves because the lending mechanism is broken.
12. December 2012 at 15:04
@ John
I never understood the macro textbooks, so I’m the wrong person to ask.
“Cantillon Effects have to do with the fact that if everyone had their money doubled like in Hume’s famous thought experiment, not every price would double instantly and not everyone would be affected in the same way by the eventual doubling in prices.”
Maybe I don’t get it, and maybe ssumner does or doesn’t either, but I think we both agree that doubling money is hard to accomplish without changing the distribution of wealth, because some assets are linked to nominal money and some are not. But this would occur NO MATTER WHERE you inject the money.
The issue is whether or not WHERE you inject the money matters. All today’s reaction shows (to the extent that it was unanticipated, which is hard to know) is that the NGDP boosting effect is much larger than the substitution effect created by expected demand. This is something that I certainly agreed on in previous posts.
Scott certainly agrees that injecting new money has distributional effects in general, he just doesn’t think it matters much how it gets injected.
Or, maybe I’m just clueless. Very possible.
12. December 2012 at 15:07
@ Major:
“‘Just like when the US got downgraded, demand for Treasuries went up.’
Including the Fed’s demand, right?”
The Fed did not announce or indicate new intention to create money immediately after the first debt cap debacle – stocks sunk, Treas spiked. I’ve argued, in the past, that Treas behaved like a giffen good.
http://baselinescenario.com/2011/08/15/money-as-the-ultimate-giffen-good/
12. December 2012 at 15:22
Bob, The news reports indicated that the policy was more expansionary than expected. Most economists did expect another 45 billion a month, but a significant number expected less. So the “new information” was admittedly modest, but in the expansionary direction. It would be easy to find dozens of similar examples throughout history, if you don’t like this one. It’s hardly a state secret that expansionary Fed announcements often make long term rates rise.
As far as MF, I don’t read him so I don’t know what he is saying. Greg obviously doesn’t agree with you. He thinks it would make no difference if I correctly interpreted the market move since I don’t understand Cantillon effects. I’d guess no one does, since everyone has their own private meaning for the term.
All I know is that lots of commenters who talk about Cantillon insisted that rates fall when the Fed buys bonds. And I insist it’s much more complicated.
Orn, Yes, there are many possibilities.
Statsguy, No, you are among the elite few who are not clueless. 🙂
12. December 2012 at 15:47
Look, I’m no expert, but I can observe the evidence of my lying eyes. And my lying eyes tell me that whenever the Fed announces a monetary expansion, interests rates respond by going up, and whenever the expansion ends, interest rates respond by falling.
I’m not sure why Bernanke and those other guys at the Fed haven’t noticed this. I can only assume one of two answers. (1) They are rejecting the evidence of their lying eyes because they don’t have the conceptual framework to process it or (2) They are lying about their intentions because they fear a lot of other people lack the conceptual framework to process what they are up to.
12. December 2012 at 16:18
“But as Bob Dylan once said: there’s no success like failure”
As Justin Bieber once said “Never say Never.”
12. December 2012 at 17:59
ssumner:
All I know is that lots of commenters who talk about Cantillon insisted that rates fall when the Fed buys bonds. And I insist it’s much more complicated.
More complicated than “When the Fed buys bonds, bond prices fall soon after”?
12. December 2012 at 18:01
ssumner:
The news reports indicated that the policy was more expansionary than expected. Most economists did expect another 45 billion a month, but a significant number expected less. So the “new information” was admittedly modest, but in the expansionary direction. It would be easy to find dozens of similar examples throughout history, if you don’t like this one. It’s hardly a state secret that expansionary Fed announcements often make long term rates rise.
Would they rise the exact same if the inflation entered the car market instead?
12. December 2012 at 18:56
“Ironically, the Fed wants to lower long term interest rates.”
Please explain this? I can’t for the life of me figure this out! If the Fed wants to increase the money supply, the Fed must encourage banks to lend. To do that banks have to get something for taking on risk!
12. December 2012 at 21:31
Scott, mind if you let me tell people what *I* think?
That’s sort of been my request all along — stop identifying “what other people think” when that involves explanatory work in economics which you are not familiar.
I’ve said several specific things about “Cantillon Effects” — I’ve said that Hayek went through this eye opening chapter in the history of economic thought for British economists for a reason — to open the door to a more profound relative prices shifting effect, the Bohm-Bawerk Effect in production expansion and the Hayek Effect in money and credit expansion.
And it served a broader purpose, to open the door to them to a different picture of the operation of price signal — a stream or flow picture of the interaction of prices rather than a static, repetitive Paretoian general equilibrium or a static, repetitive Marshallian partial equilibrium analysis or a static, repetitive Fisherian circular flow quantity theory equilibrium.
With Cantillonian price adjustment you get path dependence effects, something impossible in Pareto and Marshal and Fisher, that is unique & bumpy and easily imperfect historical streams and *not* ahistorical equilibrium perfections.
This is exactly the causal factor with explanatory oomph when you combine Bohm-Bawerk Effects with Hayekian Effects and you take seriously the factors of time, divided understanding, dispersed knowledge, imperfect relative price signal networks, genuine uncertainty, etc.
Scott writes,
“Greg obviously doesn’t agree with you. He thinks it would make no difference if I correctly interpreted the market move since I don’t understand Cantillon effects. I’d guess no one does, since everyone has their own private meaning for the term.”
12. December 2012 at 22:16
Note well.
Hayek identifies 3 stages of the development of monetary theory.
The most ancient stage is Stage 1: a mechanistic quantity theory of money, revived in ‘modern’ times by Fisher with his equation of exchange.
The next state, stage 2, involves a criticism of stage 1, and is illustrated by the stories of Cantillon and Hume describing the flow of prices changes as newly discovered gold flows through particular nodes of the economy.
Stage 3 is something different and comes *after* stage 2, and this involves extending the logic of marginal valuation to individual behavior involving money and credit, and to dynamic expectational interactions involving heterogeneous expectations, uncertainty, heterogeneous production goods, etc.
Note well that Hayek does *not* identify his own monetary economics with stage 2 economics, ie Hayek is *not* identifying his account of monetary economics and its relation to booms and busts as a Stage 2 “Cantillon Effects” explanatory story.
Scott gives us Stage 1 monetary economics augmented by magical instant and perfect expectations coordination and by random unexplained ad hoc ‘shocks’ to the normal magic kingdom.
And Scott mistakes Stage 2 monetary economics for “Austrian economics” because he’s genuinely unfamiliar with this explanatory causal framework, so its all he can imagine it to be.
There is your play guide folks.
Hope it helps.
12. December 2012 at 23:19
It seems to me the Fed is finally stepping up to the plate. They will now buy $85 billion a month in securities, half Treasuries and half MBS. And keep buying. That seems like the headline. I do wish it were $120 billion, or a rising amount monthly as suggested by Scott Sumner.
This is probably what the Fed should have done two years ago, but were too timid, and then it was election season, and Governor Perry suggested we execute Bernanke, so the Fed chickened out before the election.
13. December 2012 at 08:54
[…] I said I wouldn’t post any more on the Cantillon Effects stuff. Let me just say, Nick and Scott still think I’m an idiot. (I dispute this in the comments of their respective posts.) And […]
13. December 2012 at 09:01
Statsguy and Scott,
Now suppose that instead of the money supply doubling by everyone having twice as money, it doubles by giving one person the entire increase. Surely there will be a diversion of resources towards that person who gets the increase. How could anyone possibly ignore that? It seems so obvious. My point is that even if money were evenly distributed at the start the rise in prices would affect different people and commodities differently.
13. December 2012 at 09:05
[…] I said I wouldn’t post any more on the Cantillon Effects stuff. Let me just say, Nick and Scott still think I’m an idiot. (I dispute this in the comments of their respective posts.) And […]
13. December 2012 at 10:16
John:
Now suppose that instead of the money supply doubling by everyone having twice as money, it doubles by giving one person the entire increase. Surely there will be a diversion of resources towards that person who gets the increase. How could anyone possibly ignore that? It seems so obvious. My point is that even if money were evenly distributed at the start the rise in prices would affect different people and commodities differently.
The (alleged) escape clause to that bleedingly obvious point is that it allegedly won’t have much “macro impact.” In other words, it won’t make NGDP or price levels to be so different as compared to equivalently sized helicopter drops, or bank reserves increasing via OMO.
Yes, this is presented as a valid response. Yes, this is believed as a refutation of Cantillon. Yes, this is not a joke.
13. December 2012 at 10:47
Bob the suprise was the targets-6.5% U, 2.5% inlfation That wasn’t expected.
13. December 2012 at 11:53
As far as MF, I don’t read him so I don’t know what he is saying.
I’ve started following the same policy, and it has improved my quality of life.
13. December 2012 at 12:06
Major Freedom,
LOL. I get that though these people are completely incapable of tracing macro results back to individuals. That’s why they are so intellectually bankrupt.
13. December 2012 at 12:08
At least I got Scott to admit that a doubling of the money supply would affect different people and commodities differently instead of just raising NGDP and the price level. Once you admit that, you have to admit the Cantillon effect.
13. December 2012 at 12:18
In Hume’s thought experiment where the money supply doubles by doubling everyone’s money, the winners when prices have doubled are the people who spent money first before the price appreciation fully set in. Simple.
To illustrate the Cantillon effect, if the total money stock doubled by increasing the assets of one individual, the winners in this scenario would be that individual and the people who sell to him. The people who sell to the people who sell to him would also benefit. However, as you go farther down the line, people farther removed from this individual in the chain of spending become net losers.
This is simple F’ing logic. You can’t dispute it by looking at macro variables or specific empirical facts that are the results of massive causal chains. No natural experiments are possible. Either attack the logic or shut up.
13. December 2012 at 12:54
Scott has to deny the Cantillon Effects because it then implicates his life’s work as an inflationist as something other than beneficial. And like all bias seekers, this is too painful for him to admit. It takes a lot to seek truth at all costs; most people don’t want truth. There is simply nothing that you can do to convince him of something he doesn’t want to be true. Keynesians are the same way.
13. December 2012 at 13:35
Bear in mind Razer that there is nothing inherently immoral about the Cantillon effect, since it would exist in ANY monetary regime, even the purest free market regime. The effect occurs even in a gold standard.
Where the immorality begins is the force that compels everyone to accept them as money (because they have to pay taxes in it, which means even if they earned only gold, then they would still have to go out and acquire what the tax authority wants, which effectively monetizes what the state issues).
Note that I am not saying your argument concerning Sumner’s motivations is necessarily wrong. It may be true or false (probably true).
13. December 2012 at 13:36
Josiah:
I’ve started following the same policy, and it has improved my quality of life.
That just means your quality of life is founded on lies.
13. December 2012 at 13:47
MF, I know that Cantillon Effects are just that, effects. But if you accept that the winners are the ones who get the money first (relative to those that don’t ever touch it), and you are an advocate of inflation, then you have to accept that your policy will have winners and losers. I think this troubles people, especially when they claim to be for fairness or ‘the little guy.’
It’s just like the Keynesian or MM insistence that their inflation schemes do not cause business cycles… To admit that, or any culpability whatsoever, would admit their very life’s work is a sham. I’m still waiting for anyone of them to tell me what would change their mind? But that’s like asking a theist what it would take for him to lose their faith in God. And nothing short of having the deity show up in person and tell them that he doesn’t exist will do. Same with Keynesians and MMs. But I get it, though. It’s tough to let go of your long cherished beliefs. Damn tough.
13. December 2012 at 14:13
Razer:
MF, I know that Cantillon Effects are just that, effects. But if you accept that the winners are the ones who get the money first (relative to those that don’t ever touch it), and you are an advocate of inflation, then you have to accept that your policy will have winners and losers. I think this troubles people, especially when they claim to be for fairness or ‘the little guy.’
And likewise, if you accept that in a gold standard, or a free market in money standard, the winners are the ones who get the new money first (relative to those that don’t ever touch it), and you are an advocate of the gold standard or a free market in money, then you have to accept that your policy will have winners and losers as well!
It’s just like the Keynesian or MM insistence that their inflation schemes do not cause business cycles… To admit that, or any culpability whatsoever, would admit their very life’s work is a sham. I’m still waiting for anyone of them to tell me what would change their mind? But that’s like asking a theist what it would take for him to lose their faith in God. And nothing short of having the deity show up in person and tell them that he doesn’t exist will do. Same with Keynesians and MMs. But I get it, though. It’s tough to let go of your long cherished beliefs. Damn tough.
The only way out of this hypocrisy is to admit that whatever monetary order one advocates, the Cantillon effect will take place. The more important question is whether individuals are physically forced to be relatively benefited or relatively impoverished because of it. If the state enforced a gold standard say, because they collected taxes in gold, then to the extent an individual is not a gold miner, or does not receive the new money relatively early on, then they would be subject to the same effect, although it would almost certainly be far less pronounced as compared to printed money.
13. December 2012 at 14:33
MF, who is advocating a gold standard?
13. December 2012 at 15:30
Razer:
Many people are. I was just giving examples of alternatives to fiat money. My point is that the Cantillon effect takes place in all monetary orders, no matter what they are.
14. December 2012 at 10:54
The Fed “buying” anything IS inflation. Lots of it.
14. December 2012 at 13:59
What happens in the future when the Fed wants to get the $4 trillion off of its balance sheet? Will that cause deflation?
15. December 2012 at 07:26
John:
You are taking the Rothbardian reduction version of Cantillon effects.
Most Austrian economists say that money in fact does not appear in proportion to existing money balances, and so they impact relative prices.
Rothbard said that even if money appeared in proportion to existing balances, then there would still be Cantillon effects because different people would spend the money first.
Why do you identify Rothbard’s version with Cantillon effects?
You should think about that.
15. December 2012 at 07:54
Scott:
You sometimes revert to your monetarist roots. Monetary policy is an “exogenous” changes in growth rate of the quantity of money.
Given expectations of future nominal GDP, a growing quantity of money generated by growing bond purchases raises the demand for the bonds and result in lower yields.
But the more rapid money growth will result in more rapid inflation, more rapid nominal GDP growth, anticipations of which will result in reduced demands for bonds at any given nominal interest rate, and so lower bond prices and higher nominal yields.
I think it is conceivable that even the shortest term bonds could fall in price immediately, but given this thought experiment, the lower short term interest rates for a while, higher long term interest rates, immediately, and higher short term interest rates later, seems plausible.
But suppose that a new growth rate of the quantity of money is not the regime. Suppose the regime is to create whatever amount of money and purchase whatever amount of long term bonds is needed to peg their price? Or, more plausibly, the same policy is followed for short term bonds?
The result of that policy, if persistent, is explosive. But I don’t think the explosion would be instant.
Anyway, suppose the experiment is create the amount of money needed to keep short term interest rates at a certain level subject to the constraint that the inflation rate will remain 2%.
If the constraint is credible, then there is no impact on expected inflation and so no tendency for nominal interest rates to rise.
But we still have the other impact, and in fact, there is no “create a certain quantity of money and buy a certain amount of bonds,” but rather create a quantity of money and buy a quantity of bonds such that the price of bonds and their yield is where the Fed wants it.
Now, we can imagine that the Fed would have to buy all of the bonds, but the more realistic scenario is that they buy some, and the expectation is that they will have to reverse course later if the resulting increase in the quantity of money is inconsistent with the inflation target.
Now, the more recent policy seems to be allowing for higher inflation. And also, to the degree spending on output rises, it could (and I think should) result in more output growth. This loosening of the inflation constraint a bit should result in higher real and nominal interest rates.
It is possible that this offsets the direct effect of the purchases of bonds.
Interestingly, suppose the Fed changed the regime, and didn’t buy any more securities? We will create enough money to get unemployment down to 6.5% as long as inflation doesn’t rise above 2.5%. We make no promises about buying any securities or about target interest rates.
Would actual output and inflation rise by less? Would interest rates rise more? Independent of the clarification (change) in the goal targets, what is the impact of the promises to keep interest rates low or the actual purchases of securities?
4. April 2013 at 19:30
[…] ==> But that’s nothing. Here’s the post I was looking for, from Dec. 12, 2012, titled “Take that, Cantillon effect fans”: […]
4. April 2013 at 23:36
[…] On 12/12/12 Scott wrote: “[O]nce again Fed stimulus raises interest rates and lowers bond […]
24. March 2017 at 04:19
[…] “there’s no success like failure http://www.themoneyillusion.com/?p=18115#comments […]