Monday, June 25, 2012
Why can't the Treasury borrow directly from the Fed
Hat tip to Scott Fullwiler.
Marriner Eccles, Chairman of the Board of Governors of the Federal Reserve System 1947:
There was a feeling that this [Fed overdrafts to the Treasury's General Account] left the door wide open to the Government to borrow directly from the Federal Reserve bank all that was necessary to finance the Government deficit, and that took off any restraint toward getting a balanced budget. Of course, in my opinion, that really had no relationship to budgetary deficits, for the reason that it is the Congress which decides on the deficits or the surpluses, and not the Treasury. If Congress appropriates more money than Congress levies taxes to pay, then, there is naturally a deficit, and the Treasury is obligated to borrow. The fact that they cannot go directly to the Federal Reserve bank to borrow does not mean that they cannot go indirectly to the Federal Reserve bank, for the very reason that there is no limit to the amount that the Federal Reserve System can buy in the market. That is the way the war was financed.
Therefore, if the Treasury has to finance a heavy deficit, the Reserve System creates the condition in the money market to enable the borrowing to be done, so that, in effect, the Reserve System indirectly finances the Treasury through the money market, and that is how the interest rates were stabilized as they were during the war, and as they will have to continue to be in the future. So it is an illusion to think that to eliminate or to restrict the direct borrowing privilege reduces the amount of deficit financing. Or that the market controls the interest rate. Neither is true.
*****
It boggles the mind to see yet another reminder of how thoroughly this was understood 60 years ago. It's not feasible to imagine that later Fed staff, economists and financiers didn't know this. Most had to consciously choose to not believe it, and therefore to not teach the truth to later students.
Four score years ago, our forefathers set up a workable fiat currency system, with much trivia modified unchanged from the shambles of the failed gold-std. It is our job to see that monetary policy of the people, by the people and for the people does not vanish from the face of the earth?
Subscribe to:
Post Comments (Atom)
208 comments:
1 – 200 of 208 Newer› Newest»Countries can print money. Wow. What an amazing insight. Thanks MMT!
The Greatest Generation wasn't very good at passing on their ideas and knowledge to their kids.
FDO15, that's the Fed chair 60 years ago. Don't you even read the posts before commenting?
You are an intelligent person when you want to be. Why let your emotions make your brain go dark?
Tom, the Fed chair knew 60 years ago that the government could print money and woulnd't default due to lack of money. So what?
@ FDO15
"Tom, the Fed chair knew 60 years ago that the government could print money and woulnd't default due to lack of money. So what?"
So please go discuss that with the Peterson Foundation, Romney, Obama, McConnell, & on & on; maybe the CATO Inst & the Koch bros too
@ FDO15
So what? The present Fed chair is a supposed expert in what happened 60 years ago, and he apparently doesn't get it. Nor do the people writing the budget in the House.
You guys are delusional. I am pretty sure they get it. They just don't care about your work because their definition of insolvency is inflation.
Be logical about this all. Even if the world started using the MMT framework the right wingers would simply change their rhetoric to specify their position regarding inflation as the form of default. It wouldn't really change a damn thing. Your whole point here is meaningless.
Gotta love the MMT haters that scream and yell about how we explain "fantasy land" operations then a few days later try to argue out of the other side of their mouths that "countries can print money" is self-evident. Don't bother responding to this because I won't read it and don't care what any MMT hater has to say to me. The bigger you get, the more haters there are, and you prove this every time you hate.
FDO15, get a grip, man. You are losing it. You are capable of saying something constructive in criticism. Why do you persist in making a fool of yourself with your ranting? No control? As I said, get a grip. We can take actual criticism. Rants? Meh.
FDO15 has a point about the right wingers just moving the goalposts to support their ideology,
I'm only speaking from personal experience, but I'm finding that when you attempt to talk about this with people with right wing views, at first they won't accept the notion that a government that issues its own currency isn't revenue constrained, when you show them the many papers, blogs and other sources of information that supports MMT they then start to accept that the government can't run out of money...but then they quickly switch their point of argument to printing and inflation, and of course Weimar Germany and Zimbabwe make an appearance. I've tried to explain the differences in circumstances, like with Weimar having to repay debts in a foreign currency, I've shown them Bill Mitchell's excellent blog about what happened in Zimbabwe, but they won't listen, they flat out refuse to accept that there won't be hyperinflation if governments don't slash spending.
I personally believe the mainstream media is to blame, most people get their facts from TV, everyday on the news its the same people being given airtime to misinform people, but these guys are treated as an authority, so it's years and years of propaganda standing in the way.
So while FDO15 may have history for ranting, he does make a valid point.
How to lose an argument gracefully in 5 steps.
Being graceful under pressure is an art, and being able to put forth a gracious attitude even when someone has clearly gained the upper hand in an argument is to be admired and respected, even by the argument's winner.
1. Be objective. Once you remove your emotional attachment to your argument and look critically at what your opponent is saying, it may be that you begin to see things his way. Honestly re-evaluating your position regularly will help you to realize this
2. Say something that suggests you know you might be wrong or that your opponent might be winning. Examples: "You know, I think you might be right." "Hmm, you've really given me something to think about." "That's a point I hadn't considered before."
3. Concede the point. Tell your opponent, "Well, I'm sold," or "I admit your points are very well-taken, and I'm now persuaded to think it over some more. I do see your points, and I think I need to re-examine my opinion now."
4. Don't belabor the issue. It's even harder to win an argument gracefully than to lose one. Your opponent may enjoy gloating or doing some other sort of superiority dance after you concede. If the person arguing with you tries to push the win, just stay calm and smile. Tell them you'll talk to them again later if they want. Most people will leave it at that.
5. Don't throw a tantrum or resort to name calling.
Anonymous,
FD015 does bring up a point which is of interest to anyone wishing to implement MMT but it does nothing to negate the principles and value of MMT.
The right wing will say and do anything to prevent Government spending on social programs. Like we need FD015's brilliance to tell us that. If there was a world war tommorow the right wing would be all over MMT like a rash on a babies bum.
Fair point Andrew, I've just read the McCulley thread and see he's been active in that one as well, I've been reading this blog for quite some time and tend to skim over posts that are obviously more interested in insulting than interacting, I tend to not even bother looking at the names of those posts, so I wasn't aware he had a history on here.
This is interesting from Eccles:
"Nothing constructive would be accomplished by the proviso that the Reserve System must purchase Government securities exclusively in the open market. About all that such a ban means is that in making such purchases a commission has to be paid to Government bond dealers."
Ha! And if the current crop of moron PDs wants to give up the golden goose, I'm sure we can get an MMT group together to operate this ENTIRE system for Treasury much more efficiently and with much less acrimony than with these current moron Dealers. .... and STILL rake in a TON of USD balances...
rsp
"FDO15 has a point about the right wingers just moving the goalposts to support their ideology........
So while FDO15 may have history for ranting, he does make a valid point"
And in an earlier comment section he was excoriating various MMTers for "playing loose" with language and reminding them that "words have meanings". He was saying this from his almighty high horse, admomishing them that if they ever wanted to really be taken seriously in the econ world (as if they arent getting serious attention right now) they needed to stop going back and changing their story and writing posts about "general vs specific cases".
Now.... he blithely adds to a discussion about SOLVENCY that........ "THEIR definition of insolvency is inflation" Okay then lets "redefine" solvency now FD.
What a total DICK!
FDO15, the MMT argument is actually that countries can 'print money' without causing inflation.
That's quite different to just saying countries can 'print money', which as you say is not much of a revelation.
Before I encountered MMT I tended to casually and unthinkingly accept the general vague notion that 'printing money' necessarily leads to runaway inflation.
I subsequently realised that other economists also make the same point as MMTers, though perhaps less clearly or forcefully. I found out that even Ludwig Von Mises agreed with MMT on this (!) when he argued that increases in the supply of money are not inflationary if they are offet by an increase in the demand for money. It's odd to see a quasi-MMT statement like that sitting in plain sight right at the heart of austrian school writings. It makes you realise that generally accepted "common sense" if often just ignorant superstition.
MMTers bring this issue to the forefront, whilst debt-hysteria maniacs refuse to even consider it. Instead they blindly accept the 'common sense' error that 'printing money' is necessarily, always, inevitably inflationary - as if this were a truth handed down to them by God himself (Niall Ferguson's deranged rantings are a good example of this).
*is often just ignorant superstition.
Having said that, some of the historic periods in which the government has 'printed' a lot of money have indeed coincided with high inflation, due to: a) very rapid increases in and extraordinarily high levels of government spending (second world war for example), b) abnormal supply constraints (war, again), and c)inability to collect taxes (i.e. during the civil war I believe).
So printing and inflation are closely connected in many people's minds for historical reasons beyond just Weimar and Zimbabwe.
There is also an additional potential problem that the belief in the inflationary effects of 'printing' may potentially become self-fulfilling to an extent. i.e. people are more likely to 'dump' the currency if they fear that 'printing' will lead to inflation, thereby helping to bring about the anticipated inflation by themselves. Artsie non-linearity.
The experience with QE should have helped demonstrate that 'printing money' is not necessarily inflationary in itself. Even those who believe that the rise in oil and commodity prices was directly due to QE have to acknowledge that at worst inflation went up a bit and then came down again.
However, the argument is now that 'under normal circumstances' - i.e. not a liquidity trap or balance sheet recession, the expanded monetary base would be inflationary.
It seems to me that this argument is really that low or zero interest rates could potentially be inflationary under 'normal' circumstances, which I'm not certain I completely disagree with. Anyone?
I subsequently realised that other economists also make the same point as MMTers, though perhaps less clearly or forcefully. I found out that even Ludwig Von Mises agreed with MMT on this (!) when he argued that increases in the supply of money are not inflationary if they are offet by an increase in the demand for money.
I'm pretty sure that all economists would agree with this. (E.g., think of the generic money demand function from macro 101: M/P = f(Y,i)).
The problem is that the demand for money is endogenous, i.e., the govt can "print" money but it can't print demand.
If we could get to the point where the public no longer feared the possibility of default on the public debt, and was only concerned about the possibility that excessive deficits might result in accelerating demand-pull inflation, that would be progress. I would be happy to move onto that debate.
Another interesting debate we could have is whether we need a system in which the Treasury borrows at all - from the Fed or otherwise. The Fed could possibly manage interest rates by adjusting interest on reserves, a possibility Scott has explored in his research, and Treasury deficits could be financed by direct overdrafts on government accounts at the Fed.
But one battle at a time.
Some have argued that the Vietnam war also contributed to the 1970s high inflation.
It would appear that maxing out resources and productive capacity to sytematically destroy resources and productive capacity is actually be quite inflationary.
Probably should try to avoid doing it in future.
"…the govt can "print" money but it can't print demand…"
The government can "print demand" if there is a significant cohort that would spend if they had the dollars.
Most people don't have everything they need or want.
vimothy, you are right that the government can't print demand. But it can generate demand. Suppose the economy is operating at less than full employment, and the government orders a fleet of 100,000 automobiles. Then assuming consumer demand for automobiles is unchanged, the economy needs to manufacture 100,000 additional automobiles. It needs to finance this expanded production, and needs more money to do it. And at the point of purchase, the seller demands the money price of the automobile from the government. So the government can both create additional money, and generate the demand output that occasions the demand for the additional money.
I'm not saying automobiles would be the most socially useful thing the government could purchase. I would thing solar power plants, electrical grids, tunnels, bridges, schools and educational services would be more useful.
and generate the demand output that occasions the demand for the additional money.
Meant to write:
and generate the demand for expanded output that occasions the demand for the additional money.
"The problem is that the demand for money is endogenous, i.e., the govt can "print" money but it can't print demand."
That's why government deficit spending or "printing" of money should match, or offset "savings desires".
Standard models tend to assume full employment, which seems to lead some to the conclusion that 'printing' is inflationary.
Actually maintaining full employment by offsetting savings desires (or demand leakages) is a crucial part of providing the conditions under which overall productive capacity can grow sustainably. A steady growth in productive capacity makes further non-inflationary increases in demand possible, in a virtuous cycle. That's my simplified understanding anyway.
Problems can potentially arise if "savings desires" substantially reduce suddenly for some reason.
@Dan
The government could buy the cars and build a reef with them.
"it’s the demand leakages that create the ‘need’ for govt deficits"
the other point of course is that the assumption that borrowing, or bond issuance, is necessarily less inflationary than 'printing', is generally incorrect. It depends on a misunderstanding of what money is and how the monetary system works, even though on the face of it it seems like a reasonable assumption.
Like I said though, there's a chance that psychological factors based on incorrect assumptions and misunderstandings may have the potential to become self fulfilling. Which is why indirect 'printing' through QE may be better than direct 'printing' by the Treasury.
Most people don't rush to withdraw all their money from the banks because of QE. They see it as a slightly puzzling internal operation by the central bank which doesn't seem to change much. If the government spent by printing money directly however, people might be more inclined to get visions of Mugabe going mental with the printing press, and panic might break out perhaps (?).
"The problem is that the demand for money is endogenous, i.e., the govt can "print" money but it can't print demand"
And Ive noticed almost everyone complaining about how overstuffed their savings accounts are too!!
Four score years ago, our forefathers set up a workable fiat currency system, with much trivia modified unchanged
Our "forefathers"? The FOUNDING fathers allowed the existing system of gold and silver to continue to take place.
from the shambles of the failed gold-std.
The gold standard didn't "fail". It was systematically weakened and abandoned because it put too much of a constraint on the growth of the state.
It is our job to see that monetary policy of the people, by the people and for the people does not vanish from the face of the earth?
Fiat money is not a money for the people. Gold is money for the people, since gold protects people's income and money from the ravages of state/banker inflation.
Fiat money is money of the banking/political elite, by the banking/political elite, and for the banking/political elite.
"The people" were robbed. First in 1933 when their gold was stolen. Then again in 1971 when the rest of the world's gold was stolen.
Who is this historical revisionist "Roger Erikson"? He's utterly clueless.
Peter,
People used to complain that gold backed money served the interests of bankers over everyone else, remember?
Remember "You shall not press down upon the brow of labor this crown of thorns; you shall not crucify mankind upon a cross of gold"?
Remember the Silverites, and the Greenbackers?
Talk about historical revisionism.
And don't forget that the American colonies first managed to gain financial independence from the British bankers by creating their own independent American FIAT currencies, known as colonial scrip..
MMT policies, including reform of the banking sector, are designed precisely to rein in the excesses of the banking sector and allow government to act more efficiently in the interests of "the people".
The basic aim is full employment AND price stability.
"You guys are delusional. I am pretty sure they get it. They just don't care about your work because their definition of insolvency is inflation.
Be logical about this all. Even if the world started using the MMT framework the right wingers would simply change their rhetoric to specify their position regarding inflation as the form of default."
Dan points out in reply that it would be progress to move the debate from 'solvency" to "inflation." I very much agree.
The concern with "solvency" is something the neoliberals introduced during the 1980s and 90s. We never used to hear it in the post-war period until then. From 1945-1970 of so the debate was about UE vs. "inflation." The UE hawks usually won that debate, which is one of the reasons why the Ds were the majority party.
But, then, we had the cost-push oil/Fed inflation during Carter's time. Carter tried to contain that by making mighty efforts to balance the budget, while spreading the idea that Government deficit spending was a key to controlling it. During those years, very few mainstream economists would point out that it wasn't deficit spending causing the inflation, but the oil cartel, the Fed, and Labor, all of whom interacted to drive costs up.
So, at that point, the Ds began to become the Party of fiscal responsibility and balanced budgets. They associated deficit spending with inflation, and were now very afraid of inflation because of Carter's electoral fate. They joined with some classical Republicans like Pete Peterson and adopted the pre-Reagan Republican stand on balanced budgets, a stand which was reinforced by their need to critique Reagan on budget grounds.
Knowing that the UE vs. inflation argument was a losing one for the inflation hawk side, especially since the fear of inflation subsided in the face of the 80s and Reagan's big deficits, but still afraid of inflation themselves, and needing an issue against him, they stopped making the UE vs inflation argument, and began to argue from the neoliberal position as outlined by people like Alice Rivlin and Robert Reichsauer (sp?) at Brookings (also both CBO Directors in the early years of CBO). That is, they started talking about UE vs. the debt/deficit/possible insolvency problem, and prioritized that problem and the idea of fiscal responsibility as their reason for seeking balanced budgets and for adopting a defensive posture on expanding the safety net and discretionary spending.
During the 1980s in Congress and the 1990s with Clinton, they won that argument and convinced the public that lowering the debt and minimizing insolvency risk was more important than minimizing UE, and they largely bought into the NAIRU to support their argument.
So, here we are. Those who care about UE, lost the UE vs. inflation argument because of the traumatic cost-push inflation that had nothing to do with deficit spending. When the inflation broke in the 80s and Reagan proved that Federal deficits could co-exist with low inflation, but still afraid of inflation themselves, and needing an issue, they started doing "fiscal responsibility," and buried the UE vs. unemployment debate using fiscal sustainability/responsibility "we are running out of money" fairy tales.
We need the UE vs inflation debate back however, because we can that win that one with MMT, and that's why exposing the insolvency myth is essential for us. FDO is right that exposing it may mean that the right will just say, "oh, we meant insolvency in the sense of inflation causing currency devaluation." But if they say that, then as Dan says that is progress, because we can win that argument, especially since now with MMT informing what we do, we can produce FE w/price stability, and prove that we can have both.
Peter,
Remember MMT largely focuses on government deficit spending.
Increasing this so that it is the correct size to offset private savings desires is not necessarily about 'growing the state'. It's about taking away less money from the population in taxes relative to what is spent.
Wouldn't you like lower taxes Peter?
Letsgetitdone,
Do you think the Vietnam war may have also played a part in the 1970s inflation?
The government could buy the cars and build a reef with them.
True, Paul, and even that might have some kind of beneficial multiplier effect. But it's a waste of resources. Much better to pay people to produce something from the resources that adds value to our society - preferably over a long period of time.
Lots of responses and I probably don't have time to answer them all, sorry, so I’ll try to hit what seem like the main points.
Why do economists think that printing money is inflationary? One reason is that that’s what their models tell them (though the idea is an old one), but also because that’s what the empirical evidence suggests.
A nice way to get at the problem deductively is this: if printing money doesn’t cause inflation, then the Fed can have infinite purchasing power, since the Fed can print money. However, it’s clear that the Fed cannot possibly have infinite purchasing power, so printing money must cause inflation.
Dan’s comment is interesting. I think it can be usefully separated into two different questions.
The first one asks, if demand for money is a function of output (and interest rates, say, as in my simply model above), then can the government can control the demand for money via control of output?
Unfortunately, it’s hard to see how this would work. What is the mechanism that the government is meant to exploit?
What's more, if the government really can control the level of output directly, then it doesn’t need to concern itself with stabilisation policy. In fact, we can put away the textbooks and consider economics substantively solved and of academic interest from here on in.
The second question asks, okay, but why can’t we have the government add to demand and use freshly printed money to do it?
Well, in some sense, that is how monetary policy works, although there is a lag between the government spending money and the central bank printing it, so that only the money printing contributes to demand in the current period.
The main reason that governments generally don’t go around funding their expenditure by printing money is that the demand for (government) money is just not quantitatively significant. People don’t want to hold it in large enough quantities for it to be something that really makes a difference.
Note to all. Yes, there are always problems. Problems are the price we pay for living.
General defense against all lunacy is a more informed electorate. Even better defense is an electorate informed well enough to actually practice exercising operations based on what it knows.
Everything is obvious once made obvious. No problem with that.
Only rational thing to do, always, is to make a suggestion.
So where do we go next?
Vimothy,
" it’s clear that the Fed cannot possibly have infinite purchasing power, so printing money must cause inflation."
At some point, printing money causes inflation, of course. Because at some point spending causes inflation. No one's saying you can print infinite money and it will always have the same value. Odd argument, what does it prove?
"the demand for (government) money is just not quantitatively significant."
15 trillion (dollars) seems like quite a quantitatively significant number to me.
That's basically just government printed money which is earning interest.
The government could cut out the bond issuance process, deficit spend by printing money, and still pay interest on reserves (by printing more money) if it wanted to.
The interest rate on reserves would only go to zero in the case of spending-by-printing if the government decided that it didn't want to pay interest. That's sort of the situation right now. People are still saving that money though, and that's where 'demand leakage' is going to.
Have you read this paper by Scott Fullwiler?:
http://www.cfeps.org/pubs/wp-pdf/WP38-Fullwiler.pdf
He explains how IOR could replace bond issuance. They're two ways of doing basically the same thing.
*I should say that's partly where the demand leakage is going to.
"…Well, in some sense, that is how monetary policy works, although there is a lag between the government spending money and the central bank printing it…"
What is the transfer mechanism that turns this into spending?
y,
Yes, I've read that paper by Scott Fullwiler.
Let me ask you a question: if issuing bonds and paying interest on reserves do the same thing, then what's the point?
At some point, printing money causes inflation, of course. Because at some point spending causes inflation.
Why does spending cause inflation?
No one's saying you can print infinite money and it will always have the same value. Odd argument, what does it prove?
If you make nominal expenditure arbitrarily large by printing money, the price level will have to rise.
15 trillion (dollars) seems like quite a quantitatively significant number to me.
$15 tn is not the size of money demand, but the size of the national debt. Your argument is that you can replace this quantity of bonds, bills and other debt instruments with notes and coin and bank reserves, i.e., with Fed liabilities, and there will be no effect on the price level or its growth rate.
What is the transfer mechanism that turns this into spending?
Are you asking why monetary policy should affect aggregate demand?
"Are you asking why monetary policy should affect aggregate demand?"
Yes, especially now.
From the point of view of money, it's about equilibrium conditions for an asset portfolio composed of cash and "other stuff" with a non-zero nominal yield.
From the point of view of interest rates (bit outside the conversation so far), it's about finding channels though which interest rates can have an effect on demand (e.g., wealth, intertemporal substitution, cost of capital, income, etc).
"if issuing bonds and paying interest on reserves do the same thing, then what's the point?"
I don't know. I'm hoping you can tell me.
"Why does spending cause inflation?"
I guess its a supply and demand thing. The point is, whether the demand comes from 'printed' money or 'borrowed' money doesn't change things, does it? Or maybe it does. Please tell me why it should.
"If you make nominal expenditure arbitrarily large by printing money, the price level will have to rise."
That's the point, no one is recommending making nominal expenditure "arbitrarily large".
"Your argument is that you can replace this quantity of bonds, bills and other debt instruments with notes and coin and bank reserves, i.e., with Fed liabilities, and there will be no effect on the price level or its growth rate."
If you can explain to me why it would then that might not be my argument anymore.
If the interest on those reserves was zero then perhaps it would be inflationary depending on the circumstances. I'm not sure about the inflationary impact of low interest rates as yet.
I don't see why the amount of cash or coin in circulation would necessarily increase. Would the fact that people's savings are now in the form of interest-bearing reserves rather than interest-bearing bonds suddenly make them want to withdraw more (non-interest bearing) cash, or suddenly spend more? Maybe it would. Could you please explain why?
Would it suddenly make them *able* to spend more? I don't think so, but maybe I'm wrong. Bonds are supposed to be pretty easy to turn into reserves or cash.
Would it make it easier for banks to lend? No.
"…From the point of view of interest rates…"
So, why haven't Fed policies worked to stimulate effective demand so far?
Some have argued that the Vietnam war also contributed to the 1970s high inflation.
Nah, that's completely insane. Inflation was relatively low - and the economy was absolutely booming - when we were really spending our faces off in Vietnam. It spiked with the oil shock, but afterwards when we had relatively persistent inflation (which was, it should be noted, steadily declining) we were basically done with Vietnam and spending very little.
"if issuing bonds and paying interest on reserves do the same thing, then what's the point?"
I meant what's the point of issuing bonds, though one could ask the same thing with respect to IOR too. The point being that issuing bonds *is*, for all intents and purposes, 'printing money'.
Fullwiler also argues that IOR could result in smaller interest payments overall, and mentions ways in which govt bonds' role as a benchmark asset in the market could be replaced.
What did you disagree with in his paper? Have you discussed it with him?
y:
People used to complain that gold backed money served the interests of bankers over everyone else, remember?
Yeah, all three of them.
I wasn't talking about verbal complaints. I was talking about who gold and fait ACTUALLY serve. Complaints notwithstanding.
Remember "You shall not press down upon the brow of labor this crown of thorns; you shall not crucify mankind upon a cross of gold"?
I am not a Christian.
Remember the Silverites, and the Greenbackers?
Yes.
Talk about historical revisionism.
I know, that's what Roger engaged in.
And don't forget that the American colonies first managed to gain financial independence from the British bankers by creating their own independent American FIAT currencies, known as colonial scrip.
They already had independence with gold and silver.
MMT policies, including reform of the banking sector, are designed precisely to rein in the excesses of the banking sector and allow government to act more efficiently in the interests of "the people".
It's the exact opposite. There is no restraint implied in MMT, other than hyperinflation.
The basic aim is full employment AND price stability.
You can't get that from MMT accounting tautologies.
Prices should not be stable. Prices should reflect the relative marginal utility individuals place on goods and money. If that means gradually falling prices of consumer goods, and wages and capital, then so be it.
"You can't get that from MMT accounting tautologies."
It would help if you bothered to say why we can't.
I'll have a double,
My two cents:
Wars involve excessive increases in demand accompanied by constraints or bottlenecks in supply.
For the period of the war, huge quantities of money are spent employing vast quantities of people to do nothing genuinely productive: build missiles that then get blown up, kill other people or get killed, etc. The economy becomes geared to supporting these activities and gradually destroying its own labour force, without developing the economy's real productive capacity beyond that which is needed to support the war effort.
Food and oil tends to become more scarce too as it's all being used to do these pointless things (pointless in the case of Vietnam).
Then when the war's over you get a further surge in demand in other underdeveloped markets as people come home.
Regarding the popular conception of inflation and money printing. most people think that the money comes from govt. They don't realize that most of the money supply is created through private lending. So they draw the erroneous conclusion that govt money printing, i.e., Fed increasing the amount of resevers, is adding dollars to the economy.
I was just listening to an NPR news report on Operation Twist, where the Fed sells short term bills and buys the same amount longer out on the yield curve. The announcer said that unlike QE 1 and QE 2, Operation Twist "does not put new dollars into the economy." (me: rolls eyes).
That's a nice sleight of hand isn't it?
Government 'printing' of money is always inflationary whereas bank 'printing' of money isn't. All you have to do is pretend that banks don't create money. And most people will believe you.
Y: ""it’s the demand leakages that create the ‘need’ for govt deficits""
Non-govt saving overall results in demand leakage that reduces contracts production and employment to reduce supply accumulating as unplanned inventory, and this results in under-performance and idle resources economy-wide unless offset by a corresponding got defiict.
Buy gold, Peter. Load up on gold. Listen to Glenn Beck. Listen to Glenn Beck. Study Rothbard. Study Rothbard. But don't bother us with this nonsense.
"it’s the demand leakages that create the ‘need’ for govt deficits"
Maybe we should stress more often how the paradox of thrift is tightly embedded within MMT thought.
"And don't forget that the American colonies first managed to gain financial independence from the British bankers by creating their own independent American FIAT currencies, known as colonial scrip.."
And there are good historical arguments supporting the view that the fundamental cause of the Revolutionary Was was the economic conditions resulting from the Crown making this illegal in the colonies.
vim "Why do economists think that printing money is inflationary? One reason is that that’s what their models tell them (though the idea is an old one), but also because that’s what the empirical evidence suggests. ***A nice way to get at the problem deductively is this: if printing money doesn’t cause inflation, then the Fed can have infinite purchasing power, since the Fed can print money. However, it’s clear that the Fed cannot possibly have infinite purchasing power, so printing money must cause inflation."
The money has to get from the Fed spreadsheet as reserves into deposits that are spent to increase prices. The Fed has zero power to do that itself, although reducing the price is an incentive. Not working now, BTW. Moreover, given increasing deposit-creation due to rising demand for credit, as LLR the Fed cannot directly control the quantity of the base. It has to use monetary policy by changing price.
Economists don't seem to understand operations well enough to know where to be looking for causes (transmission mechanism) and are stuck with expectations and confidence — subjective psychological factors that are difficult to measure.
vim, "The first one asks, if demand for money is a function of output (and interest rates, say, as in my simply model above), then can the government can control the demand for money via control of output?"
Better rethink your simple model. See my comment above.
y,
The way I see it, as a point of logic, if issuing bonds and paying IOR are equivalent then we ought to be indifferent to choosing between them.
With that in mind, it seems a bit strange to be campaigning for something you think doesn't matter.
Leaving IOR to one side, then,
I guess its a supply and demand thing. The point is, whether the demand comes from 'printed' money or 'borrowed' money doesn't change things, does it?
The point is, it’s a bit misleading to say that “spending causes inflation”. People are always spending, so why should inflation be 3% this year and not -3%, or 300%? You need to relate spending to supply somehow, and then think about what sort of thing would cause spending to rise or fall against the level expected by the supply side of the economy.
That's the point, no one is recommending making nominal expenditure "arbitrarily large".
All we’re doing here is putting bounds on the problem. It’s clear that if the govt prints money and spends it on output, at some point prises have to rise. Now the only question is where that point actually is.
If you can explain to me why it would then that might not be my argument anymore.
Because the private sector has as much base money as it wants at this particular point in time. If you increase it by 1500%, unless something magical happens, people will have a lot more cash than they actually want to hold. So something else will have to happen to bring supply and demand in that market back into equilibrium.
I don't see why the amount of cash or coin in circulation would necessarily increase. Would the fact that people's savings are now in the form of interest-bearing reserves rather than interest-bearing bonds suddenly make them want to withdraw more (non-interest bearing) cash, or suddenly spend more? Maybe it would. Could you please explain why?
If there is no material advantage to be had from IOR, the only way to gain from money financed govt spending would be for the public to hold the national debt as notes and coin. But that’s only going to happen if the rest of the world turns into Russian Mafioso overnight.
IOR on reserves accrue only to banks, while bonds can be held by ordinary investors, mutual funds, pension funds, etc. That's one difference.
@ y
There are arguments that Nixon shut the gold window to gain policy space to prosecute the war, and that the 70's inflation resulted at least in part from the inflationary effects of war spending, as you point out. There was also strong labor bargaining power at the time, which resulted in effective labor resistance of falling real wages due to price rises. Nixon was forced to consider wage-price controls, which are a typical wartime tool to control inflation.
See Wikipedia — Nixon Shock for a historical summary of what happened in 1971.
"it seems a bit strange to be campaigning for something you think doesn't matter"
Sorry I probably wasn't clear. If you accept the argument that bond issuance and IOR are practically the same in their effect (leaving aside possible psychological factors I mentioned above) then bond issuance is in effect 'printing' - so in essence keep doing the spending and don't worry about the debt, just think about which interest rate is best and which level of spending relative to taxation is best for full employment and growth in the economy with the level of inflation you want.
Fullwiler argues IOR should result in less outlays through interest overall so maybe give that a go if that's the goal.
Also, if you're going to pursue the zero interest rate policy favoured by MMT economists then it probably makes more sense from an operational perspective to just issue money instead of repeatedly selling bonds to the dealers that are then bought by the central bank. Also reduces the dealer's 'cut' if that's what you want to achieve.
Whether a permanent ZIRP would be inflationary, when combined with other MMT policies including 'alternative' forms of monetary policy is another matter I'm not entirely sure about as yet.
I'll respond to your other points a bit later. Thanks.
IOR on reserves accrue only to banks, while bonds can be held by ordinary investors, mutual funds, pension funds, etc. That's one difference.
That is indeed a difference.
So, one reason why we might not be indifferent between IOR and bonds is that under IOR we're essentially paying the banking system to hold the national debt.
But note that we're still paying someone to hold it.
Anyway, Tom,
The money has to get from the Fed spreadsheet as reserves into deposits that are spent to increase prices. The Fed has zero power to do that itself, although reducing the price is an incentive.
It's just a thought experiment. It means we can "start from the things which are more knowable and obvious to us and proceed towards those which are clearer and more knowable by nature".
If you like, imagine the Fed going out and actually spending the money. Or imagine that they pass it to Treasury. It's not really important.
BTW, if you're committed to the idea that there's no way to get printed money out into the economy, then this whole conversation is moot.
It's hard for the Fed to get the "printed" money into the real economy because they are restricted by law in what they can buy.
"BTW, if you're committed to the idea that there's no way to get printed money out into the economy, then this whole conversation is moot."
Fiscal. Period.
If you accept the argument that bond issuance and IOR are practically the same in their effect... then bond issuance is in effect 'printing'
The argument is that paying interest on reserves is the same as paying interest on any other debt instrument.
That doesn't mean that printing money is the same as issuing debt. Money is zero-yield--that's what makes it attractive to policy makers, that's where seigniorage arises.
If you argument is that you'd rather pay interest on reserves to interest on bonds, why? What's the advantage?
keep doing the spending and don't worry about the debt, just think about which interest rate is best
But then you have a system where the interest rate on the entire national debt is the Fed's policy instrument. Is this really a good idea? What's the rational?
Also, if you're going to pursue the zero interest rate policy favoured by MMT economists
Why would you want to hold interest rates at 0% indefinitely?
vim, The way I see it, as a point of logic, if issuing bonds and paying IOR are equivalent then we ought to be indifferent to choosing between them.
Operationally equivalent as reserve drains but not equivalent in all effects. If paying IOR reduces interest payments, then it would be preferable unless an argument can be mounted that there is greater public interest is served by higher interest payments. Moreover, it can be argued that the difference in interest payments constitutes a subsidy for holders of govts. Subsidies are economically inefficient.
"If you like, imagine the Fed going out and actually spending the money. Or imagine that they pass it to Treasury. It's not really important."
Not authorized by the FRA 0f 1913 as amended, as far as I can see. Sure, Congress could authorize the Fed to do fiscal policy as its agent, I guess, but it hasn't and I doubt very much it ever will.
The Fed did come arguably close to spending money in the real economy during the GFC via programs like Commercial Paper Lending Facility. However, this spending was balanced by the creation of liability with strict collateral requirements, so no NFA creation occurred. Mostly all repaid now.
Tom,
As I said, if you think that it's impossible in principle for the government to print and spend money, then the whole question of money financed government spending is moot.
The gov't as in consolidated Treasury and Fed can certainty "print" and spend ... just hard for the Fed to do it alone under present arrangements. Taken together they can.
y, yes I do just a bit, but I think that was under control by the time the oil crisis started.
Right, but the institutional arrangements of the US monetary system are more or less irrelevant.
We are (or we seemed to be, when I suggested it) interested in whether, in principle, printing money causes inflation. So just make the necessary adjustments, and then re-run the thought experiment.
Printing money causes inflation in the same sense as eating causes obesity ... both can have bad results if done too much. Not doing at all also has negative consequences.
As I said, if you think that it's impossible in principle for the government to print and spend money, then the whole question of money financed government spending is moot.
Of course, it is possible for governments to issue currency and disburse it into the economy either to transfer private resources to public use or for transfer payments like SS, food stamps, etc. This is accomplished through the appropriations process by which fiscal policy is created. The executive branch just administers this through its agencies, who all carry out the will of the people's representatives as expressed in the various bills that are passed into law.
As Warren says, the correct operational meaning of "printing money" means deficits expenditure that increases non-govt NFA.
Ken "The gov't as in consolidated Treasury and Fed can certainty "print" and spend ... just hard for the Fed to do it alone under present arrangements. Taken together they can."
But pretty much only as directed by Congress through appropriations.
As Warren has said, the operational meaning of govt "printing money" is fiscal though deficit expenditure that adds NFA.
The rest of money creation is net zero and is leverage, i.e., as a corresponding liability in the private sector. That's the real "funny money."
vim "Right, but the institutional arrangements of the US monetary system are more or less irrelevant."
That's the difference between PKE and the mainstream. PKE denies that.
Tom,
Of course, it is possible for governments to issue currency and disburse it into the economy either to transfer private resources to public use or for transfer payments like SS, food stamps, etc. This is accomplished through the appropriations process by which fiscal policy is created. The executive branch just administers this through its agencies, who all carry out the will of the people's representatives as expressed in the various bills that are passed into law.
Sure. But these are practical matters. We also interested in matters of principle.
As Warren says, the correct operational meaning of "printing money" means deficits expenditure that increases non-govt NFA.
It's hard to see how that could be the correct meaning, in any sense.
“Printing money” is just a synonym for the central bank issuing base money. It doesn’t have anything to do with deficit spending.
Deficit spending -> budgetary authority issues liabilities
Printing money -> monetary authority issues liabilities
That's the difference between PKE and the mainstream. PKE denies that.
You're taking that statement out of context and then reading it as part of some kind of point of principle.
But all I'm saying there is that particular arrangements aren't important to the thought experiment, not that they aren't important in general.
"Deficit spending -> budgetary authority issues liabilities"
Correction:
Deficit spending -> budgetary authority issues assets
vim, "“Printing money” is just a synonym for the central bank issuing base money."
And in the MMT view, essentially irrelevant to anything, other than insofar as it has an effect on the cb's target rate. So either set rate to zero or non-zero. If non-zero then bond to drain excess reserves to hit rate, or else pay IOR equal to or greater than the target rate.
What's the point of running thought-experiments that are essentially irrelevant to the real world?
It's a bit comical for a biologists to watch economists & accountants discuss complex issues that quite obviously spill outside the narrow approach of either discipline.
Here's a bio view of what you call inflation.
Aggregates compete via "auto-catalysis." [just look it up]
There are two avenues for auto-catalysis.
1) tune a given aggregate to increasingly more dense engineering
(cost of coordination scales rapidly)
2) enlarge the aggregate, to unlock more avenues for indirection
(luckily the cost of coordination scales much less rapidly)
So, 2 is still the predominant reality seen in all species. Even then, however, there's a scalable cost to organizing on a larger scale. Luckily, the return on coordination always dwarfs the cost of that coordination.
What you guys call inflation, a biologist would generally call the cost of coordinating larger aggregates. Sure, that's a cost, but it's always outraced by the faster-growing return on coordination.
In everyday terms, going from 100mil to 200mil population means a growing currency supply, and a given currency unit is unlikely to buy the same unit amount of some commodity (say milk).
The solution is to adjust both incomes & practices so that all system members (population) always have enough income to take care of essential needs PLUS engage in local experimentation & innovation.
By most definitions of inflation, a system can't grow without inflation. So what? Who says inflation has to be held constant? We want to grow net capabilities if not net population. We need appeal to no higher authority.
Until the consequences of larger populations outweigh the benefits, we'll continue with both population growth and currently defined definitions of inflation.
After that point, we'll have to transition back to growth through more sophisticated engineering - i.e., recursive tuning of economic & cultural processes in a stable/shrinking population.
We ain't there yet. Lots of other options still to be explored.
Are you subtly messing with me here, Tom? :-)
It's not irrelevant to the real world.
If printing money doesn't cause inflation, then government can fund its expenditure by issuing money.
"If printing money doesn't cause inflation, then government can fund its expenditure by issuing money."
Yes, that is exactly the MMT point.
"So, one reason why we might not be indifferent between IOR and bonds is that under IOR we're essentially paying the banking system to hold the national debt.
But note that we're still paying someone to hold it."
Don't forget we can pay off the national debt with the proceeds from a $60 T coin, then neither the banks nor foreign nations, nor private sector entities will be holding. Why not do that. makes things simple.
As for keeping over night interest rates at zero, why not? What public purpose is served by having it any higher?
"IOR on reserves accrue only to banks, while bonds can be held by ordinary investors, mutual funds, pension funds, etc. That's one difference."
No they don't. Banks have to pass the interest they recieve on reserves on to their depositors if they want to keep them. They keep a part of the overall interest as profit.
If you buy a bond 'direct' from the goverment, you still go through a bank, and they skim a bit off the top there too. Because buying a bond from the government still involves your bank transferring reserves to the treasury (when it comes to settlement), and then receiving reserves from the treasury in payment (when it comes to settlement).
"Don't forget we can pay off the national debt with the proceeds from a $60 T coin, then neither the banks nor foreign nations, nor private sector entities will be holding. Why not do that. makes things simple."
You don't need a big coin to pay off the national debt. The Fed caould buy as much of it as it can get its hands on.
The coin idea is just about paying off the debt that the Treasury owes to the Fed (in the trillions at present) - which is pretty meaningless really given that the Fed hands profits over to the Treasury and its all just internal government accounting between different government departments.
"If printing money doesn't cause inflation, then government can fund its expenditure by issuing money."
The MMT point, as far as I can tell, is that bond issuance effectively IS 'money printing', except that the money in question (bond) accrues interest, over a specified period of time.
If you just 'print' money (reserves) and then pay interest on them its basically the same thing.
Instead of thinking about bonds as 'bonds' it might helpt to think of them instead as time deposits or saving accounts held at the Fed (which can be switched to demand deposits at very short notice).
Get it?
Why are we paying anyone to hold the national debt as non-govt saving of accumulated non-govt NFA when it is operationally unnecessary under the currency monetary regime. It's a left-over from the convertible fixed rate days. First, the interest payments, being operationally unnecessary, constitute a subsidy, and secondly, the issuance of tsys to obtain reserves just confuses the operational reality in the national mindset by obscuring what's actually going on operationally, that is, a reserve drain instead of financing.
As Mike Norman puts it, govt bonds are basically dollars which accrue interest for a while.
In their physical form Treasuries are essentially high-denomination Federal Reserve Notes with coupons attached.
Treasuries are reserves that accure interest.
*accrue
"Why are we paying anyone to hold the national debt as non-govt saving of accumulated non-govt NFA when it is operationally unnecessary under the currency monetary regime."
To encourage sociopathic malfeasants.
And MMR? Remember, the gov't can legally seize a bank's assets, but not the other way 'round.
C'mon. Who's your Daddy?
The difference between money and government bonds is that bonds earn a nominal return, whereas money does not but has particular liquidity characteristics that people value.
Governments earn seigniorage from money, which is the spread between what they pay on money (nothing, in nominal terms) and what the CB earns on its assets.
If you pay interest on money, it looks a lot like a bond (that is continuously rolled over), and it loses its chief attraction for the government in terms of funding.
Vimothy (and others),
Forgive me if this has been covered as I’ve probably missed a couple posts – up to 90 something now…
“Printing money” does not cause inflation. Yes, you can observe cases of excessive inflation and hyper-inflation and yes they do have governments printing money. However this is correlation and not causation. How do I know this? Well as Marriner Eccles points out, a government that is fully sovereign in its fiat currency must be funded directly or indirectly by its central bank. That’s just how it works. Therefore ALL deficit spending is just “money printing” and yet excessive and hyperinflation are the exceptions and not the norm in the world where nearly 200 or so fiat currencies are used. Mainstream economists miss this because they have a poor understanding of monetary and banking operations.
IOR… “If you pay interest on money, it looks a lot like a bond (that is continuously rolled over), and it loses its chief attraction for the government in terms of funding.” For the most part what’s the difference for a fully sovereign issuer of the currency to “print up” enough money to fund its spending and “printing up” enough money to fund its spending plus interest? The only time this might be a problem is if the amount of interest “printed up” and spent plus government spending and private sector spending exceeded the economy’s ability to meet total spending (causing inflation).
The reality is that paying IOR is a monetary policy tool, nothing more. It’s probably best to start with why a central bank exists. The primary reason/mission of any central bank is to ensure the safe functioning of the payment system that underpins any fractional reserve banking system. Illiquidity can destroy any fractional reserve bank regardless of how well run it is. The CB’s job is to ensure that there are always sufficient reserves available to the banking system to clear and settle payments. How does it operationally go about doing this? It’s not feasible to call every bank and ask them if they have enough reserves (and some might outright lie to prevent the regulators from showing up). It’s tried directly controlling reserves (Volker years) and failed miserably. The only reliable way to monitor the reserve demand is to set a target interest rate and defend it. If you’re a fully sovereign fiat issuing government spends by creating reserves then reserves pile up in the banking system. Once there are more reserves in the banking system than needed to clear and settle payments arbitrage ensues in the overnight interbank lending market and the CB eventually looses control of its target rate – UNLESS those excess reserves are drained by issuing bonds OR the CB pays IOR equal to its target rate OR you set the target rate equal to zero which is where excess reserves will push the rate to without one of the other two options being implemented.
"The difference between money and government bonds is that bonds earn a nominal return, whereas money does not but has particular liquidity characteristics that people value."
Okay, sounds like a reasonable definition. However, 2 things:
1) reserves actually earn interest at present. I think reserves are 'money'.
2) bonds are highly liquid. To the point that they can be changed into reserves, cash or deposits at the drop of a hat, more or less.
"Governments earn seigniorage from money, which is the spread between what they pay on money (nothing, in nominal terms) and what the CB earns on its assets."
I don't really understand why the central bank needs to be particularly concerned about gaining earnings from its assets, given that it can create money at will? For a government which issues its own money, revenue from assets and taxes is logically just a means of regulating demand within the economy, and also perhaps supporting certain markets (in the case of assets).
In the case of the US, the 'government' is divided into different parts, including the central bank and treasury. There are reasons for these divisions but to get a broader view of the monetary system in relation to the government it can help to think of them as being functionally consolidated.
Another point - in the US the Fed hands over all profits (after expenses inc. dividends on member bank capital) to the Treasury.
So for example (simplified):
1. The Treasury borrows some money from bank and issues a bond.
2. The Treasury then spends this money on a new bridge.
3. The Fed then buys the bond from the bank with newly created money (let's say).
4. The Treasury then borrows thath money from the bank and issues another bond.
5. The Treasury then uses that money to pay the money it owes on the previous bond to the Fed.
6. The Fed then hands that money back to the Treasury.
7. The Treasury then uses that money to pay back the money it borrowed from the bank.
Through this convoluted process, money has effectively been 'printed into existence' and spent on a new bridge.
"If you pay interest on money, it looks a lot like a bond (that is continuously rolled over), and it loses its chief attraction for the government in terms of funding."
Again, a government which issues its own currency doesn't actually need "funding" does it? SO then what is the role of government borrowing and taxation?
Borrowing serves to 'drain reserves' from the banking system and maintain a positive interest rate. Taxation serves to regulate demand in the economy and creates a basic 'need' for the 'government money'.
Spending by issuing money directly, or else maintaining a zero (or near zero) interest rate on government debt, may be beneficial for the government as higher interest rates increase expenditures and so can possibly add to inflationary pressures over time. Interest on government debt is basically also a kind of unearned subsidy to the private sector which benefits those with a lot of money the most.
Comments appreciated. Cheers.
I'm not an economist btw as you can probably tell, so please forgive my simplifications.
Another point I didn't mention is that the central bank controls the short term interest rate directly, and this translates to yields on government bonds. The central bank does also have the option of determining bond yields directly if it wants to.
"Taxation serves to regulate demand in the economy and creates a basic 'need' for the 'government money'. "
In my view a much more important consequence of taxation and spending is the economic activity created.
Without this re-distribution money would merely accumulate at the top and become inert…
…because every process within the non-government (that succeeds) is parasitic (in a mathematical sense).
Adam,
In general, the CB holds as assets some portion of the national debt against its liabilities in the form of base money.
If the CB fully funds net government expenditure then it must hold the national debt in its entirety and issue an equivalent amount of money.
It's very easy to see that this does not describe the US, where the national debt is an order of magnitude greater in size than the Fed's balance sheet.
So the argument here is about whether the government could, if it chose to, fund all of its expenditure in this way and not just a tenth or fifth or whatever it happens to be at the moment.
The reason funding via the CB is attractive is that the nominal interest rate on money is zero. (Essentially, the govt can borrow for free in nominal terms). Unfortunately, though, this means that people economise on their holdings of it. IIRC, notes and coin in the US is under $1tn and pre-crisis reserves were minuscule (about $10 billion).
That leaves paying people to hold money--but that's basically what a bond does, so it seems a bit redundant.
I don't think the functioning of the payment system is relevant. We're interested in funding opportunities for the government.
“I don't think the functioning of the payment system is relevant.” Correct, but if the CB wants a target interest rate greater than zero and there are reserves in excess of what is needed to clear and settle payments then the CB either needs to get the Treasury to drain those reserves by issuing debt or it must pay IOR. It’s a policy tool for monetary policy. It’s irrelevant to fiscal policy beyond the need to “print up” additional moneys to cover the interest costs.
“If the CB fully funds net government expenditure then it must hold the national debt in its entirety and issue an equivalent amount of money.” Not true. What is the difference, beyond the order of transactions between these two scenarios:
A) US Treasury gets the FED to print up reserves and add them to its account. The Treasury then spends those reserves which adds deposits to private sector bank accounts and reserves to bank reserve accounts. The Treasury then sells bonds to private individuals lowering deposits and reserve balances.
B) US Treasury sells bonds to private sector banks draining reserves from bank reserve accounts. This causes upward pressure on the overnight interbank lending rate which causes the FED to add reserves to the reserve system in order to defend its target rate.
In both cases the FED created reserves to cover the transaction and it did not end up with the bonds on its balance sheet. “B” is the exact process Marriner Eccles is describing when he says “…the Reserve System indirectly finances the Treasury through the money market…”
"Without this re-distribution money would merely accumulate at the top and become inert… because every process within the non-government (that succeeds) is parasitic (in a mathematical sense)."
No it's not that simple.
"No it's not that simple."
Because?
"The reason funding via the CB is attractive is that the nominal interest rate on money is zero. (Essentially, the govt can borrow for free in nominal terms). Unfortunately, though, this means that people economise on their holdings of it. IIRC, notes and coin in the US is under $1tn and pre-crisis reserves were minuscule (about $10 billion)."
Yes, people won't choose to hold money in large quantities if it earns no interest.
But reserves don't leave the central bank so I don't really see what you're point it.
If the government were to (deficit) spend by issuing money directly (simply crediting bank reserve accounts without issuing bonds), and then were to choose not to pay any interest on those reserves, those reserves would simply sit in the bank's reserve accounts earning zero interest.
They wouldn't go anywhere, unless they were withdrawn as cash. But a zero interest rate on reserves is not suddenly going to make people want to withdraw more cash.
"Because?"
Because the world, and 'the market', is dynamic and unpredictable. People create their own assets and their own wealth all the time, and these circulate in complex ways.
I think the world is better because of government redistribution, but I wouldn't say without it the thing would reach some inert parasitical dead end, like some alien face hugger on the face of humanity.
"…People create their own assets and their own wealth all the time…"
My comment was wrt financial assets only. People can't create those.
From an essay by Kenneth E. Boulding…I thought this part was illuminating (h/t to circuit in another thread yesterday):
http://dieoff.org/page160.htm
…"In regard to the energy system there is, unfortunately, no escape from the grim Second Law of Thermodynamics; and if there were no energy inputs into the earth, any evolutionary or developmental process would be impossible."…
Think of the above in terms of the expansion of financial assets re the non-government stock/flow dynamic.
Y,
What I’m trying to say is that as you give reserves or the base bond-like characteristics, they become de facto bonds. David Andolfatto describes it like this:
What is a Fed note? One can think of it as a risk-free claim to a future Fed note. The nominal interest rate on this note is zero.
What is a Treasury note? It is a risk-free claim to a future Fed note. The nominal interest rate on the Treasury note is typically positive (we say, the Treasury note is "discounted").
What happens if the Treasury note is not discounted? It then becomes indistinguishable from a Fed note. Both assets represent risk-free claims to future cash and these claims earn zero interest.
http://andolfatto.blogspot.com/2010/09/what-is-clear-and-not-so-clear-about.html?showComment=1285341872242#c3026924494349817140
As you can see, he’s actually discussing what happens when the nominal interest rate on a T-note goes to zero, rather than what happens when the nominal interest rate on reserves goes to the T-note, but it’s exactly the same principle. They become indistinguishable. Since they become indistinguishable, choosing between them can’t have a material impact on the government’s spending plans and the whole thing is academic.
Note that this doesn’t mean that choosing between money financing and bond financing is immaterial in all states of the world—only when these instruments pay out at the same rate.
You’re right that govt bonds are highly liquid. But are they as liquid as money? No. If they were, then no one would want to hold money. Now, the private sector holds a quantity of govt bonds and a quantity of money. You can think of this a being a single risk-free portfolio for the whole private sector. What determines the relative proportions of the two types of asset in this portfolio? If you make arbitrary changes to the composition of this portfolio, what will happen?
I don't really understand why the central bank needs to be particularly concerned about gaining earnings from its assets, given that it can create money at will?
The revenue that the government earns from the CB’s ability to create money at will is the difference between what the CB pays on its liabilities and what the government pays on its liabilities, times the size of the CB’s asset holdings. The ability to print money just means that the CB can issue a particular type of liability with a particular liquidity profile. It doesn’t mean that the government has unlimited spending power. The amount of revenue that the government can raise in this way is strictly and nontrivially limited by the public’s desire to hold notes and coin.
Through this convoluted process, money has effectively been 'printed into existence' and spent on a new bridge.
Let’s try and go through this process a bit more carefully. Financial assets and liabilities only.
1, The Treasury borrows some money from bank and issues a bond
Private sector assets: up one bond, down one bank deposit.
Public sector assets: up one bank deposit.
Public sector liabilities: up one bond.
2. The Treasury then spends this money on a new bridge.
Public sector assets: down one bank deposit.
Private sector assets: up one bank deposit.
So what’s happened overall? Public sector liabilities have increased to the tune of one bond. Private sector assets have increased to the tune of one bond. No increase in money.
Now you add some stuff with the central bank. This is where indirect financing comes in.
3. The Fed then buys the bond from the bank with newly created money (let's say).
Private sector assets: up reserve balance, down bond.
Public sector assets: up bond; down reserve balance (on CB balance sheet).
Since the Fed remits profits on the bonds it holds to the treasury, the treasury is effectively financing at the zero nominal rate.
Overall we have: public sector liabilities, up one reserve balance; private sector assets, up one reserve balance.
Beyond that point, your story gets a bit confused.
Another point I didn't mention is that the central bank controls the short term interest rate directly, and this translates to yields on government bonds. The central bank does also have the option of determining bond yields directly if it wants to.
The CB controls only the very short end of the nominal yield. It does not control the real rate at which agents in the economy are willing to defer the use of resources.
Paul, ok agreed about the energy input thing. Also agreed that taxation also serves to redistribute spending power within the non-government.
In theory I suppose the central bank could lend the private sector whatever reserves and cash it needed to pursue its activities, including credit expansion by the banks. Then the non-govt overall would become increasingly indebted to the govt over time, though if the govt kept lending regardless then that wouldn't necessarily be a problem overall. In order for the non-government to stop being in debt to the government the government would have to spend (!).
"So what’s happened overall? Public sector liabilities have increased to the tune of one bond. Private sector assets have increased to the tune of one bond. No increase in money."
This is true usually except that this rule has been violated routinely over history, and private sector dollar assets were increased without increasing bonds held by the public.
The government does what it wants/needs to do, regardless of institutional arrangements.
There is no mathematical constraint.
vimothy: "it loses its chief attraction for the government in terms of funding."
out of paradigm. Govt doesn't need to fund itself operationally. Bonds are a reserve drain. Warren says that this was his fundamental insight here
"If you make arbitrary changes to the composition of this portfolio, what will happen?"
No one is talking about making arbitrary changes to anything to begin with.
If the government began spending by just crediting bank reserve accounts (with banks then in turn crediting depositors' accounts) whilst government bonds were still outstanding, what would happen? Is that your question?
I think the interbank interest rate would fall and bond yields would also fall. Am I wrong?
"The ability to print money just means that the CB can issue a particular type of liability with a particular liquidity profile. It doesn’t mean that the government has unlimited spending power."
Agreed, the government doesn't have unlimited spending power, in the sense that beyond a certain point increased government spending will lead to an increasingly high rate of inflation. At some point it has to stop spending or the money will eventually lose all value.
"The amount of revenue that the government can raise in this way is strictly and nontrivially limited by the public’s desire to hold notes and coin."
This makes absolutely no sense to me, so you might have to explain further.
If the government were to spend by simply crediting bank accounts and not bothering with bond issuance, banks would end up with excess reserves and the interbank interest rate would fall, eventually to zero. The yields on any remaining bonds would also fall.
Those excess reserves would simply sit in bank reserve accounts at the central bank. They can't leave
those accounts unless they are withdrawn as cash. If interest rates fall as the result of these excess reserves, this will not necessarily change the public's desire to hold cash (notes) or coins in any way. Whether the public desires to hold these excess reserves as cash strikes me as being irrelevant. The excess reserves will simply sit in bank reserve accounts even if no one wants to hold them as cash.
* "They can't leave
those accounts"
What I should have said is reserves can't leave bank reserve accounts *as a whole*. I.e. they can move from one reserve account to another within the central bank system but can't leave that system UNLESS they are withdrawn as cash, OR if the government (or central bank) removes them by taxing them away, or borrowing them.
"So what’s happened overall? Public sector liabilities have increased to the tune of one bond. Private sector assets have increased to the tune of one bond. No increase in money."
Actually if the government bond is bought by a bank, rather than a depositor (me, for example), then deposits do increase whilst bank reserves decrease.
Govt doesn't need to fund itself
Then why does it issue liabilities? And why do you care what type of liabilities they are?
"Then why does it issue liabilities?"
A powerful banking lobby maybe?
"There is no mathematical constraint"
I might add no real operational constraint either.
Tsy issues tsy securites in amt $x
Fed auctions said tsys
Non-govt now holds
Fed takes proceeds of auction and marks up Tsy reserve acct $x
Tsys disburses amt $x into non-govt
Result, non-govt holds NFA - 2 $x, whereas it held previous held 1 $x.
And reserve quantity remains the same.
So, injection of NFA into non-govt and reserve drain of the add.
Non-govt. has more NFA held as saving in tsy, the add flows through the economy, and the Fed manages its target rate without disruption from the reserve add owing to the reserve drain.
Adam,
if the CB wants a target interest rate greater than zero and there are reserves in excess of what is needed to clear and settle payments then the CB either needs to get the Treasury to drain those reserves by issuing debt or it must pay IOR.
The government can’t choose the supply of money and the nominal interest rate separately. They’re jointly determined.
Not true.
How can it not be true? If the government only borrows from the central bank then the central bank must hold the entire stock of government debt. That seems pretty unambiguous to me.
When Eccles says, “the Reserve System indirectly finances the Treasury through the money market”, he is of course correct but only insofar as the Fed actually indirectly lends to the Treasury. This only a fraction of total the government total financing needs so other sources of financing are also required for the government.
Look, we’re arguing here about whether the government can finance all of its net spending with money. If the government already does this, then hang up your boots, soldier because the battle is already won.
Paul ""No it's not that simple." Because?"
True the deficit flows to saving in tys, which is held as financial assets by those who have enough revenue to store wealth.
But the deficit also flows through the economy, producing non-financial effects.
Moreover, there is some leavening financially due to progressive taxation.
But, yes, the deficit is saved as an increase in some people's financial wealth in the form of tsys due to current institutional arrangements.
"…If the government already does this…"
The government does this when it wants to…and has done so many times in the past.
When cash flows into the non-government without an increase in bonds held by the public, what do you call it?
@Tom
I think you misunderstood my comment.
If the government didn't tax progressively economic activity would be substantially lower than it might be otherwise.
That was the only point I was attempting to make.
Tom Hickey:
Buy gold, Peter. Load up on gold. Listen to Glenn Beck. Listen to Glenn Beck. Study Rothbard. Study Rothbard. But don't bother us with this nonsense.
You haven't shown anything I said to be "nonsense."
I have shown what you said to be nonsense.
You don't like gold? I'll gladly take any you might own off your hands!
"The CB controls only the very short end of the nominal yield. It does not control the real rate at which agents in the economy are willing to defer the use of resources."
The CB can control the long end of the yield if it wants to. That's what it has done to a great extent with QE. 'Normally' however the CB just controls the short end and allows the market to determine the long end. However, there is a limit to how high the long end yield can go, and this is related to the short end determined by the CB. They're "tethered" to each other so to speak. Is this not the case?
Here's Fullwiler again:
"Treasury notes and bonds are essentially fixed-rate, long-term time deposits held by the non-government sector. It is well known that the rates on these instruments are primarily determined by current and expected future levels of short-term interest rates, though varying term and liquidity premiums can also be significant. FOMC Chair Bernanke recently reiterated this point while reciting the implied arbitrage relationship:
All else being equal, if short-term rates are expected to be high on average over the relevant period, then longer-term yields will tend to be high as well. Were that not the case, investors would profit by holding a sequence of short-term securities and declining to hold long-term bonds. . . . Likewise, if future short-term rates are expected to be low on average, then long-term bond yields will tend to be low as well. (Bernanke, 2004)
Of course, since the short-term rates themselves follow the Fed’s target closely, this means that longer-term rates are based upon current and expected actions of the Fed...
Overall, any interest rate paid on the national debt is set by the stance of monetary policy whether it is held as demand deposits (i.e., Fed accounts) of the interest-bearing variety or short-term time deposits (Treasury bills), while the expected stance of monetary policy is significant when longer-term, fixed-rate time deposits (Treasury notes and bonds) are issued. As Bernanke and others recognize, for traders to try and set bid-ask rates on Treasury securities that deviate significantly from these principals, even amid the expectation of growing future deficits, would present an arbitrage opportunity for other traders to exploit."
Scott Fullwiler, "Interest Rates and Fiscal Sustainability" (2006)
http://www.cfeps.org/pubs/wp-pdf/WP53-Fullwiler.pdf
vimothy": "Then why does it issue liabilities? And why do you care what type of liabilities they are?"
Currency is a govt liability of zero maturity. Govt securities are govt liabilities of non-zero maturity. Typically one pays interest the other does not.
Govt issues its currency as tax credits in order to provision itself. It can, if it choose, also extend the use to other non-govt uses as extensively as it chooses.
Since there is no operational need to issue securities as a reserve drain if it chooses to pay IOR or set the rate to zero, why pay interest as a subsidy. What is the public purpose. It potentially adds to effective demand and inflation as full employment approaches and therefore reduces policy space unnecessarily.
Tom, Vimothy,
Yeah I made an error regarding reserves decreasing in my example of a bank buying a bond.
Overall reserves would be unchanged in that case whilst deposits would increase and the bank would have an additional bond too.
“How can it not be true? If the government only borrows from the central bank then the central
bank must hold the entire stock of government debt. That seems pretty unambiguous to me.”
Actually it’s very non-transparent that’s why so many people think the government is funded by bonds. Let walk through some of the accounting. Lets assume the government needs to “borrow” $100 to spend $100…
Step 1: Government sells bonds…
Treasury Assets +$100 Reserves
Treasury Liabilities +$100 Bonds
Primary Deal (Bank) buys the bonds…
Bank Assets -$100 Reserves
Bank Assets +$100 Bonds
Step 2: Central Bank Open Market Operation in response to decline in bank reserves
FED Assets +$100 Bonds
FED Liabilities +$100 Reserves Issued
Banking System
Bank Assets +$100 Reserves
Bank Assets -$100 Bonds
Step 3: Government Spends Money
Treasury Assets -$100 Reserves
Treasury Net Equity -$100 (Spending as payments to non-government without a physical purchase) or Treasury Assets +$100 (Spending on actual things)
Banking System…
Bank Assets +$100 Reserves
Bank Liabilities +$100 Deposits (from government spending)
Step 4: Central Bank Open Market Operation in response to increase in bank reserves
FED Assets -$100 Bonds
FED Liabilities +$100 Reserves Withdrawn
Banking System
Bank Assets -$100 Reserves
Bank Assets +$100 Bonds
In the end the FED’s balance sheet remains unchanged. Banks have added bonds and deposits BUT did not give up reserves. Where did the reserves come from to buy the bonds? The FED via steps 2 and 4!
Nice, Adam1. For future reference, you can use cite Table 3 on page 18 here if you don't feel like typing all that out ( http://www.boeckler.de/pdf/v_2011_10_27_lavoie.pdf ).
Y,
Beyond that point, your story gets a bit confused.
Sorry, that looks a bit blunt and it wasn’t meant to be. This is complicated stuff and it’s easy to get lost. I do all the time. For example, it’s hard to mentally keep income and money separate when you’re thinking about financial flows. This makes it something that can be difficult to explain or discuss. That’s what I was reaching for a bit ineptly then.
If we start at stage 3:
3. The Fed then buys the bond from the bank with newly created money (let's say).
Remember that the Fed doesn’t create “money“ generally. It only creates a particular type of money. Reserves are a tiny, quantitatively insignificant amount, so the money that corresponds to the Treasury bonds held by the Fed is notes and coin held in vaults and in circulation.
4. The Treasury then borrows thath money from the bank and issues another bond.
Does it really borrow that money? No, it does not. The size of the Fed balance sheet does not change when the Treasury borrows—notes and coin held by the public remains what the same level as before the bond issuance. Instead, the public swap the new bond for an existing bank deposit. Bank deposits—not a Fed liability—go down, holdings of Treasury bonds by the private sector go up.
5. The Treasury then uses that money to pay the money it owes on the previous bond to the Fed.
Whatever it spends it on, it’s financing it by selling bonds to the private sector. This must be the case, unless the Fed turns around and buys the newly issued bond. Then the public hold more cash.
6. The Fed then hands that money back to the Treasury.
It hands back the interest to the Treasury. It doesn’t hand back a wad of $100 bills. Now the government has essentially sold two bonds to the central bank, and the private sector has taken on the same amount in cash newly issued by the CB.
This is only a hypothetical though.
In general, the central bank doesn’t issue money in order to finance government expenditure. The CB issues money because people want to hold it and the CB wants to manipulate interest rates.
So one question we could ask is: Could the government fund ALL (and not just some) of its net spending via money issued by the central bank? That’s the sort of thing MMTers are suggesting is possible, not already a practical reality.
Vimothy,
I’m finally seeing where you’re coming from. You keep saying “create money” when you specifically mean expand the monetary base (create base money).
That precludes the recognition that deficit spending by the treasury increases financial assets (money) held by the non-government sector without having to have the non-government sector directly fund said spending – because the FED/Central Bank will always ensure there is sufficient liquidity to complete said bond sale.
The Sum of Vertical Money = Outstanding Treasuries + Reserves + Coin & Paper Currency – Loans from Central Bank
So when you ask, “…Could the government fund ALL (and not just some) of its net spending via money issued by the central bank?”
The MMT response is sure. All you are doing is replacing outstanding treasuries with more reserves. Since reserves are not lent out by banks (horizontal money) in the creation of bank loans; reserves just pile up in the banking system and create downward press on interest rates. This requires some type of reserve drain or IOR if the central bank wishes to have a policy interest rate greater than zero.
Adam,
The only money that they government can be said to create is central bank money, i.e., base money.
That precludes the recognition that deficit spending by the treasury increases financial assets (money)
Deficit spending does not increase financial assets (money). It increases financial assets (bonds). Money is a liability of the Fed. If money increases it must be because the Fed has issued it. The Fed does this independently of Treasury bond sales.
Adam1:
I’m finally seeing where you’re coming from. You keep saying “create money” when you specifically mean expand the monetary base (create base money)."
Right. vimothy is a monetarist. So he rejects the fundamental understanding of MMT. Of course, that doesn't mean he isn't a nice guy. ")
Y,
The CB can control the long end of the yield if it wants to.
The CB can try to influence the long end of the yield curve, but there is nothing that allows it to control it directly. The only thing the CB really controls directly is a very short term nominal interest rate.
There is a theory of the yield curve known as the expectations hypothesis. This hypothesis says that long rates average expected future short rates, plus some risk premium.
That’s only a theory, though. In practice there are a lot of well known problems with it. E.g., under the EH the yield curve should be flat on average but tends to slope upwards.
If you think back to QE2, interest rates actually rose.
Vimothy:
"The CB can try to influence the long end of the yield curve, but there is nothing that allows it to control it directly. The only thing the CB really controls directly is a very short term nominal interest rate."
The cb can set the price of, say, the 10 yr and signal they it is will to take any quantity to maintain its target. To do this it loses control of the overnight rate unless it pays IOR or sets the rate to zero. No problem doing it operationally though.
Vimothy:
"The amount of revenue that the government can raise in this way is strictly and nontrivially limited by the public’s desire to hold notes and coin."
This makes absolutely no sense to me, so you might have to explain further.
If the government were to spend by simply crediting bank accounts and not bothering with bond issuance, banks would end up with excess reserves and the interbank interest rate would fall, eventually to zero. The yields on any remaining bonds would also fall.
Those excess reserves would simply sit in bank reserve accounts at the central bank. They can't leave
those reserve accounts (overall - i.e.they can move between reserve accounts), unless they are withdrawn as cash, or if the government (inc. central bank here) removes them by taxing them away, or borrowing them.
If interest rates fall as the result of these excess reserves, this will not necessarily change the public's desire to hold cash (notes) or coins in any way. Whether the public desires to hold these excess reserves as cash strikes me as being irrelevant. The excess reserves will simply sit in bank reserve accounts even if no one wants to hold them as cash.
"unless the Fed turns around and buys the newly issued bond. Then the public hold more cash."
Why, if the central bank bought a bond directly from the Treasury would the public end up holding more cash? I really don't get what you're trying to say here.
“Right. vimothy is a monetarist. So he rejects the fundamental understanding of MMT. Of course, that doesn't mean he isn't a nice guy. ")”
Right, hasn’t blown his top yet and has been very cordial (unlike some other occasion dissenters around here).
Ha! Thanks guys.
I don't think of myself as a monetarist, by the way, but I think I know what you mean.
Can you briefly explain the cash thing please?
Y,
The central bank can’t choose the policy rate and the supply of electronic reserves separately. So that rules out just parking reserves in the banking system.
If the central bank pays interest on reserves, there’s no difference in principle to paying interest on treasury liabilities. So there’s no gain, which rules that out.
That only leaves changes to the stock of cash.
"The central bank can’t choose the policy rate and the supply of electronic reserves separately. So that rules out just parking reserves in the banking system."
Ok, in my understanding, if the banks end up with 'excess' reserves as the result of government deficit spending through direct 'money creation' (i.e. crediting accounts without borrowing/issuing bonds),
the interbank interest rate will fall, eventually to zero (as the banking system as a whole ends up with more reserves than it 'wants').
These excess reserves just sit in the banking system and that's it - unless of course they are withdrawn as cash, taxed or borrowed away by government. The interbank rate will stay at zero, unless enough cash is withdrawn by the public to cause the interest rate to rise.
So clearly "parking reserves" in the banking system is not 'ruled out' if the central bank is happy for the interbank interest rate to be zero.
Regarding the difference between IOR and treasury liabilities, I'll just quote Fullwiler again as it's quicker:
"with IBRBs and no securities issued, the interest rate is the rate paid on IBRBs; where short-term securities are issued, as above these rates are set via arbitrage with the Fed’s target; as longer maturities are issued, again as above these rates are set largely via arbitrage with the expected path of the Fed’s target... the rates on various types of Treasury debt are set by the current and expected paths of monetary policy and according to liquidity premia on fixed-rate debt of increasing maturity. Since long-term rates are normally higher than short-term rates, total interest on the national debt would be significantly reduced if IBRBs eventually replaced Treasuries. Those—like the Treasury—fearful that IBRBs would reduce seigniorage income neglect that this would be far outweighed by the reduction in total interest paid on a national debt increasingly held as IBRBs. Indeed, there is no inherent reason for Treasury liabilities to exist across the entire term structure except as support operations for longer-term
rates".
So, given that there is a difference (and a possible gain), the IOR option is not ruled out either.
The questions then are:
(a) is there a good reason not to keep the interest rate at zero?
(b) is there a good reason for the treasury to issue liabilities to support longer-term rates?
According to the MMT economists the answer to both (a) and (b) is generally 'no'.
Why do you think they might be wrong?
vimothy, "if the central bank pays interest on reserves, there’s no difference in principle to paying interest on treasury liabilities. So there’s no gain, which rules that out."
According to Scott Fullwiler, interest payed out under IOR will generally be less.
But Warren recommends setting the rate to zero with 3 mo max T-bills issued for short term parking, in which case interest payments are minimal.
"If the banks end up with 'excess' reserves as the result of government deficit spending through direct 'money creation'... the interbank interest rate will fall"
If you did this, nominal interest rates would collapse to zero.
But say you don't care. What would happen?
In the best case scenario, a stable long-run equilibrium is deflation, which forces you to pay a positive real rate on govt borrowing.
Alternatively, the worst case, unstable outcome would be hyperinflation.
y: "The questions then are:
(a) is there a good reason not to keep the interest rate at zero?
(b) is there a good reason for the treasury to issue liabilities to support longer-term rates?
According to the MMT economists the answer to both (a) and (b) is generally 'no'.
Why do you think they might be wrong?"
Good one.
Actually, this has been debated before, at Warren's, IIRC.
Some people fear that a zero rate and no bonds would be inflationary. MMT says no, because 1) interest payments increase non-govt deposit account and potentially contribute to excessive effective demand relative to expansion capacity of output at full employment, and 2), tsys are so highly liquid that they effective serve as money in the sense they do not inhibit consumption desire, as might less money-like saving vehicles.
"The central bank can’t choose the policy rate and the supply of electronic reserves separately."
Actually the central bank can only choose the policy rate. It must then meet any level of demand at that policy rate. Any attempt to literally control the supply of (as you call it) money would require the central bank to close its discount window. If the central bank closes its discount window long enough it is tantamount to causing a banking crisis which would be in violation of probably every central bank charter (with he possible exception of the craziness going on in Euro world).
vimothy: "If you did this, nominal interest rates would collapse to zero.
But say you don't care. What would happen?
In the best case scenario, a stable long-run equilibrium is deflation, which forces you to pay a positive real rate on govt borrowing.
Alternatively, the worst case, unstable outcome would be hyperinflation."
Needless to say, the MMT economists disagree. What they would say instead is that the benchmarks (imposed by a command system) have been removed and the market would be free to set rates based on credit demand, the 5 C's, and competition among market participants.
What wrong with ending a command system run by a board made up of unelected and unaccountable bureaucrats with no skin in the game and connection with some of the market participants and not others. Sounds like a giant leap forward for both capitalism and democracy to me. What's not to like about that.
Of course, if Congress decided it can do a better job of setting rates than the market it can take that economic and political chance — with full accountability to voters.
Tom,
IOR means that banks hold positive reserve balances which are remunerated at the a target rate. That rate is a "floor" for the overnight interest rate. This means that OMO no longer work. In order to change the overnight rate, the Fed would need to change the rate at which it pays out on reserves, i.e., move the floor.
So now the nominal rate on the national debt will be whatever the target rate for overnight lending happens to be. If this is low, the nominal rate on the national debt will be low. If it is high, then the rate on the debt will be high.
Actually the central bank can only choose the policy rate. It must then meet any level of demand at that policy rate.
But that's what I'm saying. If the central bank chooses an interest rate, then that implies a particular stock of money. If it choose a stock of money, then it would implies a particular interest rate. To pick an interest rate is to pick a money stock and vice versa. It can't separate these two things. They have to be taken together.
"But that's what I'm saying. If the central bank chooses an interest rate, then that implies a particular stock of money."
I don't that's right. Or at least it's definitely not a "particular" stock of money. The point is made most clear if you imagine a system of IOR. You can have many different amounts of excess reserves and still have the same interest rate.
"If it choose a stock of money, then it would implies a particular interest rate."
I don't think that is right either. Or at least definitely not a "particular" interest rate. If the banking system is not supplied the reserves required to meet settlement needs, then the interest rate will continue to rise until the discount window is accessed. So at the end of the day, it's de facto interest rate targeting.
Vimothy: "In order to change the overnight rate, the Fed would need to change the rate at which it pays out on reserves, i.e., move the floor."
Yes.Unless the rate is set to zero.
"So now the nominal rate on the national debt will be whatever the target rate for overnight lending happens to be."
There is no operational reason to issue tsy securities, so what national debt? Existing debt will be redeemed at par as it matures, and yields will be set by the market unless the Fed decides to target the yield curve.
wh10,
I'm taking the base case to be the pre-crisis Fed.
Under IOR, the two are no longer jointly determined. Hence the irrelevance of OMO.
v: "But that's what I'm saying. If the central bank chooses an interest rate, then that implies a particular stock of money. If it choose a stock of money, then it would implies a particular interest rate. To pick an interest rate is to pick a money stock and vice versa. It can't separate these two things. They have to be taken together."
The Fed chooses a target rate (price) and uses quantity of reserves as a buffer to hit it, unless it pays IOR or set the rate to zero, in which case quantity of reserve in excess of what is needed for settlement is irrelevant.
Vimothy - okay.
What are your thoughts on the second part of my comment (which doesn't need to invoke IOR)?
Tom,
What does "operational" mean here?
Financing the entire stock of government debt at the shortest possible maturity is either a good idea or a bad idea. Of course, the government can always do whatever it likes if it doesn't care about the consequences.
Vimothy - we also have empirical evidence from the Fed in the 80s (I believe) that trying to target a monetary aggregate did not yield a "particular" interest rate but led to very significant interest rate volatility, effectively requiring the Fed to re-intervene as per the old norm. I think it's been called 'dirty interest rate targeting.'
Vimothy,
Maybe it's terminology, but "stock" implies a fixed amount at a specific time period. You cant pre specify a "stock" of reserves at a specified interest rate. What you can do is specify the interest rate and delivery what ever amount is required to meet that level of reserves. The central bank doesn't know the level of demand of reserves at the time it sets its interest rate (although yes it can "estimate" or guess) all it can do is deliver an infinite supply at that rate or change its rate but it can't specify the rate and supply.
That said, on paper, you could specify a supply and let the rate float but that leaves the central bank open to insufficient liquidity and a banking crisis which is why the central banks can't choose this option in the real world.
wh10
If the banking system is not supplied the reserves required to meet settlement needs, then the interest rate will continue to rise until the discount window is accessed.
Okay, so why is the interest rate rising? It's rising because the interest rate is not independent of the stock of money. In equilibrium, the interest rate equalises the marginal return on bonds versus the marginal return on money. If you choose the interest rate, then you have to let the supply float to find equilibrium. You can't choose the interest rate and then choose a money stock independently.
we also have empirical evidence from the Fed in the 80s (I believe) that trying to target a monetary aggregate did not yield a "particular" interest rate but led to very significant interest rate volatility
What I'm saying here is not in disagreement with the main MMTers, as far as I'm aware. If the Fed targets a money supply, then it loses control of the IR. If it targets an IR, then it loses control of M.
Vimothy, you are changing your tune a bit. Your claim was that you could pick a stock of money and that would equilibriate with a *particular* interest rate- but that's clearly not true empirically in the real world or theoretically. I presented the empirical case. Picking a stock of money will either lead to the interest rate winding up at the base rate or the discount rate, and those are matters of picking an interest rate. Ie, de facto interest rate targeting. Correct me if I am wrong there.
(Also, we're talking about the interbank market. Demand for reserves is by and large a function of settlement needs and meeting reserve requirements.)
Vimothy, well you didn't say that here:
""If it choose a stock of money, then it would implies a particular interest rate.""
That statement, to me, was implying a clear link between a stock of money and a *particular* interest rate.
Furthermore, even if you pick a stock of money, you lose control of interest rates only in a certain sense. The CB still can set a floor rate and a ceiling rate - de facto interest rate targeting.
but yeah Vim we're pretty much on the same page
What I'm saying is the same.
What you are presenting as empirical evidence is what we would expect to see.
Adam1 said it better than I did btw
Think of the interest rate as a function of money and money as a function of the interest rate.
As a policy maker you can take one of these out of the system and set it exogenously. But you can't do this for both.
The point is you can either target an interest rate or money, but not both. This is quite a famous result. See: http://adam.vwl.uni-mannheim.de/fileadmin/user_upload/adam/Makro_B/Poole_1969_Monetary_Instrument_.pdf
Vimothy: "what does "operational" mean here?
There is nothing in the monetary system to prevent it. As you say the constraints are consequential, basically price stability and fx stability.
Financing the entire stock of government debt at the shortest possible maturity is either a good idea or a bad idea. Of course, the government can always do whatever it likes if it doesn't care about the consequences.
Right, the argument is over the consequences. Opponent argue it will be inflationary and MMT economists argue that it won't be.
vimothy: "If the banking system is not supplied the reserves required to meet settlement needs, then the interest rate will continue to rise until the discount window is accessed.
Okay, so why is the interest rate rising?"
The interbank market for borrowing reserves overmight. Because banks needing reserves will bid up the overnight rate to get reserves from banks holding excess reserves in order to beat the discount rate. In the case of large sums every tick counts.
vimothy: "If the Fed targets a money supply, then it loses control of the IR. If it targets an IR, then it loses control of M.'
This is the monetarist mistake. It identifies M (money supply (stock really) with the size of the monetary base — reserves and currency.
The relevant money supply affecting price change is the money available for spending (effective demand). M1, M2 M3, and MZM (money available for spending) are greater than M0 and MB. Inaddition, tys can be converted to deposits immediately as can lines of credit.
Taking money stock to be the monetary base is just not in accord with reality.
vimothy:
"Think of the interest rate as a function of money and money as a function of the interest rate."
The interest rate is the overnight rate on bank reserves. The stock of bank reserves is not the stock of "money." It is the stock of bank reserves.
The reason that monetarists are confused is that they don't carefully define "money" and stick to it.
There are many stocks of "money." M0, MB, M1, M2, M3 (no longer reported officially) and MZM. KNowing the effects of each is crucial.
viomothy: "The point is you can either target an interest rate or money, but not both."
you can set price of reserve borrowing in the overnight market and let quantity float. You cannot set the quantity, however, and run an LLR. If you want to set quantity of reserves and let the price of reserves float, then you have to forgo the LLR and be willing to let the overnight rate go to both zero or through the roof, and also risk banks going down due to liquidity issues.
Tom,
Electronic bank reserves at the Fed are part of the stock of base money. Base (or narrow or reserve) money is the only asset I’ve referred to as money, and I’ve done so consistently throughout the thread.
The reason I’ve done this is that the base is an ultimate liability of the sovereign and so a source of financing for it. You guys were suggesting that this source could support the whole national debt. If we start calling every asset money, then saying, e.g., “money financed deficits are inflationary” becomes meaningless, or at least, very confusing.
The Fed expands and contracts the base in order to actively manipulate the interest rate in the overnight market, or passively in response to changes in demand for base money.
It doesn’t set this interest rate directly. Instead it tries to supply a particular quantity of reserves consistent with a target rate. This is not an exact science though, and the actual rate fluctuates around the target.
In order to make sense of this process you need a model that can explain why changes in the stock of reserves should matter.
"You guys were suggesting that this source could support the whole national debt"
If the government were to finance all of its spending by creating new money when it spends (and all prior treasury liabilities were to be retired), then there wouldn't be a national debt as such. Taxation would "destroy" money and spending would create it.
"In order to make sense of this process you need a model that can explain why changes in the stock of reserves should matter."
Let's say there's a given quantity of reserves in the system for a given quantity of deposits.
Banks then create additional deposits when they make loans.
The 'loan contract' is the bank's asset and the deposit is the bank's liability, whilst the deposit is the borrower's asset and the loan contract is their liability (debt). In this case both assets and liabilities are created on the spot, so to speak.
In a system with a minimum reserve requirement, the increase in deposits created by the loan will generate a need for additional reserves. If the quantity of reserves is the same as before, this additional need will put upward pressure on the interbank interest rate. So in order for the interbank rate to remain the same, additional reserves will have to be added to the system somehow.
Say these additional reserves are added by the central bank via an OMO. The resulting quantity of reserves in the system is now larger than before, but the interest rate remains the same as before.
So there is a relationship between the quantity of reserves and the interest rate, but not a simple one in which a given quantity of reserves always entails a specific interest rate. Rather, the interest rate is determined (chosen) exogenously by the central bank whilst the quantity of reserves needed is determined endogenously by the banking system when it makes loans.
The central bank will set the price of reserves (the interest rate) and allow the quantity of reserves in the system to float, by supplying any quantity of reserves needed to maintain that interest rate.
----
If the banking system ends up with more reserves than it needs for its reserve requirements and for settlement purposes, the interbank interest rate will begin to fall - unless the central bank withdraws enough reserves from the system to stop the rate from falling, or else pays interest on reserves at its target rate. Assuming the need for reserves does not increase, if the central bank does neither of these things then at some point the interest rate will hit zero.
That's my understanding (no doubt very simplified).
I suppose the questions then are,
1. Does a lower interest rate necessarily lead to more deposits being created (through additional loans)?
2. Does a higher interest rate lead to less deposits being created?
3. If a lower interest rate does lead to more deposits being created, does this necessarily result in inflation?
4. If the interbank rate is zero, does this lead to hyperinflation?
Vimothy: "In order to make sense of this process you need a model that can explain why changes in the stock of reserves should matter."
We have an explanation. Reserves are a buffer stock that expands and contracts through Fed monetary ops that manage the overnight rate when the Fed doesn't pay IOR or set the rate to zero. Thus, the quantity of reserves floats relative to price and need for reserves for settlement. The Fed automatically provides liquidity and then manages the amount of rb to maintain the target rate using OMO, tys issue being a reserve drain that also contributes to rb management for hitting the target. The quantity of reserves is only relevant when the Fed targets a rate without either paying IOR or setting the rate to zero.
That's not really what I'd call a model, Tom.
You need to be able to explain in general terms the relationship between quantities of reserves and the ON rate.
Y,
If the government were to finance all of its spending by creating new money when it spends (and all prior treasury liabilities were to be retired), then there wouldn't be a national debt as such.
That seems like quite a semantic argument. There would still be a stock of sovereign liabilities held by the public. This stock of liabilities would differ from the current stock in average maturity and nominal interest rate, but these are differences in degree. If you want to call this stock something other than the national debt then that’s fine by me but it’s not a substantive issue.
Let's say there's a given quantity of reserves in the system for a given quantity of deposits.
Banks then create additional deposits when they make loans.
…
So in order for the interbank rate to remain the same, additional reserves will have to be added to the system somehow.
Say these additional reserves are added by the central bank via an OMO. The resulting quantity of reserves in the system is now larger than before, but the interest rate remains the same as before.
So what we have is:
Banks have a particular demand schedule for reserves.
The Fed supplies a particular quantity.
The price of reserves is the ON rate, at the intersection of those two schedules.
If demand shifts, the price will change, unless the Fed also shifts the supply.
Yes, though if the Fed pays interest on reserves then that relationship between quantity of reserves and the ON rate doesn't apply.
So, the government could spend by simply crediting reserve accounts, and then pay interest on those reserves to stop the ON rate from falling to zero (assuming the treasury and fed were combined).
I'm interested in finding out whether you're right to say that the outcome of a permanent zero interest rate policy would be either deflation or hyperinflation.
"You need to be able to explain in general terms the relationship between quantities of reserves and the ON rate."
Should a cb wish to set a target rate and defend it rather than pay IOR or set the rate to zero, then the cb has to manage rb quantity in the interbank market by provide enough reserves to clear while keeping within the corridor range.
If rb quantity becomes too large, then offers to lend fall below the floor of the corridor, and if too little, then the bidding goes above the ceiling of the corridor.
So the cb adjusts quantity to keep the overnight market rate within the corridor. Quantity is adjusted on the large scale by issuance of govts and Tsy-cb coordination, and the cb fine-tunes rb quantity with OMO using repo and reverse repo.
This is not a hypothesis. It is a general description of how a corridor system operates in practice.
y: "So, the government could spend by simply crediting reserve accounts, and then pay interest on those reserves to stop the ON rate from falling to zero (assuming the treasury and fed were combined)."
No need for cb and Tsy to be combined? Not sure what you are thinking of here.
This is not a hypothesis. It is a… description.
Right. You’ve described something but you haven’t explained it. There’s no theory here.
In general, my observation is that you guys don’t really seem to ever talk about theory. It’s just operations this and operations that. That’s why you can suggest that the government should do things that are prima facie crazy, like permanently setting rates to zero / money financing govt spending, without any fuss. In an operational sense, these things are certainly possible, so why not?
vimothy, to get to a hypothesis involving a causal nexus one must first get the description of the context correctly and precisely define terms, etc. so that the hypothesis can be checked against the facts.
MMT draws no hypotheses involving causal inferences from the general description about. It does conclude that monetarism is wrong because they don't describe the operations correctly and construct hypotheses that involve causal inferences based on false premises.
This explains why monetarism is a failed theory and approach, as events go to show.
Tom,
I don't really know anything about monetarism. But I think that it's certainly not the case that the handful of working MMT academics are better at describing or more knowledgeable about monetary operations than the world's central banking community.
Being familiar with operations is actually the most trivial aspect of monetary economics. By themselves, operations mean nothing, explain nothing, and imply nothing. You need theory--the true pons asinorum of the discipline--to make sense of operations and to make use of them.
vimothy: "But I think that it's certainly not the case that the handful of working MMT academics are better at describing or more knowledgeable about monetary operations than the world's central banking community."
Argument from authority. One of the informal fallacies.
And the last time I checked the Fed's intro book for the public featured the discredited money multiplier. A paper published by BIS economists set the record straight on the money multiplier. So I don't take the Fed as an authority on this, when their literature has it wrong.
Being familiar with operations is actually the most trivial aspect of monetary economics.
Yes, but if you get the so-called trivial stuff wrong, it is important.
By themselves, operations mean nothing, explain nothing, and imply nothing. You need theory--the true pons asinorum of the discipline--to make sense of operations and to make use of them.
The description shows that the cb can set price by letting the quantity float and how this is actually done in practice.
The price of reserves, changes in the price, and expectations about amount and direction of future change, as well as plausible rate of change or reversal of policy have causal implications in that the the rate (price) set by the cb is THE benchmark rate that influences virtually all other rates and yields in the currency zone, as well as being a factor influencing the fx rate and current/capital accounts.
Quantity is constantly shifts under LLR, and no causal implications flow from it.
Argument from authority. One of the informal fallacies.
An appeal to authority need not be fallacious. It might be entirely appropriate. And in fact, you frequently make appeals to authority on this blog.
And the last time I checked the Fed's intro book for the public featured the discredited money multiplier. A paper published by BIS economists set the record straight on the money multiplier. So I don't take the Fed as an authority on this, when their literature has it wrong.
I don't see what this has to do with anything. We were talking about operations.
Let's take the US as an example. Is it your contention that MMT economists better understand Fed operations than the Fed itself?
Before you answer, you might like to consider where MMT economists got their knowledge of Fed operations in the first place.
Yes, but if you get the so-called trivial stuff wrong, it is important.
It certainly might be in some contexts.
The description shows that the cb can set price by letting the quantity float and how this is actually done in practice.
You need a model of the market--at the very least--to make sense of it. A description is not enough. Why is there a relationship between the quantity of reserves and the interest rate? What defines the (partial) equilibrium in this market?
Without this sort of knowledge, being able to carry out this operation is not something that is any use to the authorities.
vimothy: Before you answer, you might like to consider where MMT economists got their knowledge of Fed operations in the first place.
As I understand it from MTT economists, from Fed SOP manuals, talking to people at Fed and Tsy actually involved in day-today ops, and looking at accounting data published by Fed and Tsy.
IN contrast, most economists just make stuff up, and the senior level at the Fed and Tsy aren't involved in actual ops.
vimothy, :You need a model of the market--at the very least--to make sense of it. A description is not enough. Why is there a relationship between the quantity of reserves and the interest rate? What defines the (partial) equilibrium in this market?
Again, operational. It is bids (demand) and offers (supply) in the interbank market for rb by banks in order to get reserves for settlement and to meet RR without incuring the penalty rate.
Very straight forward. And observable in terms of market action as it takes place among banks in the IB market for rb. Thus the cb can step in with OMO for fine-tuning at any time the market is open.
Which is actually want happens, although the necessity for OMO is reduced by the banks' realization that the cb has the power to do this. So they just fall in line. (Fullwiler)
It's one of the most straightforward matters in monetary economics.
Unlike hypotheses about the causal effect of the interest on S and I.
Vimothy,
"Why is there a relationship between the quantity of reserves and the interest rate? What defines the (partial) equilibrium in this market?"
There isn't necessarily a relationship between the quantity of reserves and the interest rate.
If the banks have 'excess' reserves and the central bank pays interest on reserves, then IOR rate will be the interest rate regardless of how many excess reserves the banks have.
I think you are perhaps assuming that there is necessarily a relationship between the quantity of reserves and the interest rate because you (appear) to take Tsy issuance as a given, as 'normal'.
If you take Tsy issuance out of the picture and replace Tsys with 'excess' reserves, then the interest rate will be that paid by the central bank on those reserves, regardless of how many 'excess' reserves they have.
Is this wrong?
That's correct, y.
And if the cb sets the rates to zero quantity doesn't matter.
Banks lend based on demand by credit worthy customers, and under LLR the cb makes rb available as needed.
The quantity of rb does not affect lending propensity or capacity under LLR.
The price of rb may influence demand for credit by affecting the spread the bank charges, but what the influence of rb price may be is influenced by other conditions wrt specific cases.
There is no hard and fast relationship that can be determined by a function relating a dependent variable to an independent variable in any empirical research I know of, e.g., Δquantity of rb to either Δprice level or Δcredit extension.
The partial equilibrium model of supply and demand comes from basic micro. It’s not “operational”. You don’t observe supply and demand. It’s theory.
Supply and demand explain the relationship between price and quantity in the reserve market. That’s why floors and corridors are necessary. Too much supply and the equilibrium price collapses to zero. Too little supply and the equilibrium price explodes.
Vimothy,
I believe this the model you're referring to:
http://krugman.blogs.nytimes.com/2012/04/02/a-teachable-money-moment/
In the above model, if we have a system in which the government spends by simply crediting reserve accounts without issuing bonds (i.e. it finances its deficit spending with new money creation), then the demand curve 'A' will shift to the right when reserves are added (by spending, or by Fed lending), and will shift to the left when reserves are subtracted (by taxes or loan repayments to the Fed).
The interest rate (r*) will be whatever the Fed decides, regardless of the quantity of reserves (B*).
If it decides to pay interest on reserves, then that will be the interest rate (r*). Changes to B* will not effect (r*), given that changes in B* entail a shift of A.
As such we can see that it's quite possible for the Fed to maintain its current practice of altering the interest rate in response to changing conditions, even if the government finances all of its deficit spending through new money creation rather than by borrowing.
So, some questions: Does the interest rate really need to be altered up and down in order to control inflation? Is this really a good and effective policy tool? Does it even work as expected? Are there possibly other, better ways of controlling inflation?
Y,
I was thinking of supply and demand in the reserve market.
This sort of thing:
http://libertystreeteconomics.typepad.com/.a/6a01348793456c970c01630227fb52970d-500wi
To make sense of Fed operations, you need a model like this. That's necessary, but not sufficient--to think about their effect on the wider economy, you're going to need something more.
Sorry, just had a quick look at Krugman's post. Yeah, that's the sort of thing I was thinking of.
The last two paragraphs are a bit contentious, though. I'd take them with a pinch of salt.
As such we can see that it's quite possible for the Fed to maintain its current practice of altering the interest rate in response to changing conditions, even if the government finances all of its deficit spending through new money creation rather than by borrowing.
No one is disputing the ability of the Fed to peg a rate above the equilibrium rate.
If the government finances its deficit by creating reserves, which it then pays the banking system to hold, then it is only financing by money creation in a very narrow, semantic sense.
It is still financing its deficit by issuing interest bearing liabilities. The difference is that these liabilities are at a much shorter average maturity and the nominal interest payments on them move one-to-one with the target rate.
Vimothy, quantity of reserves has zero effect on the "wider economy." Price does.
Understand the operations show that Δrb is an ex-post facto accounting record, which is an effect of "deposits create reserves."
The mainstream understanding fails because they have the causality reversed.
The direction from from demand for credit by creditwothy borrowers > Loans (bank asset, borrower liability) create deposits (bank liability, borrower asset). These are accounting records entered by keystroke on the banks internal ledge. This generates corresponding credit and debt in the Fed's spreadsheet for the bank's reserve account. Owing to LLT, the quantity of reserves is not fixed but endogenously determined by bank credit activity. If the Tsy and cb are using the cb reserve issuance-Tsy security issuance and the cb wants to set a rate target greater than zero and not pay IOR, then the cb uses OMO through repo and reverse repo that involves tsys and reserves.
There is zero theory here. It is description, Where the wider economy is involved, causality comes in and therefore theoretical hypotheses about predictable effects is with respect to price, that is, the interest rate.
The descriptively, change in quantity of reserves is an accounting effect of credit extension — "loans create deposits, and deposits create (generate) reserves" — Warren Mosler. Thus the quantity of reserves is an
accounting residual."
On the other hand, the interest rate is predicted to have certain effects on the wider economy BECAUSE it is the benchmark rate that is used by those who extend credit in the wider economy. What those effects are is controversial, as well as the role of expectations regarding changes in the interest rate at the discretion of the cb.
Changes in quantity of reserves has no causal effect on the wider economy, nor does it have any effect on either the ability of the banks to extend credit or on their actual procedure and decision-making in doing so. The price of reserves affects the spread they charge.
What is so difficult to get about this? It is basic to MMT and it is one of the chief differences between MMT and the monetarist mainstream, where "monetarist" means anyone who imputes non-existent causation or constraint to the quantity of reserves in a cb-based system running LLR.
Tom,
There’s nothing wrong with being interested in operations. They’re not the whole story though.
The Fed always acts in response to events in the broader economy, with the intention of influencing those events in some direction. It thinks about ultimate goals and it thinks about proximate or intermediate goals that might help bring those ultimate goals about. It acts and conceives of itself acting across a (relatively) long time horizon.
If you know some details about a particular instrument, that’s cool, but on their own they can’t explain Fed activities as the Fed itself thinks of them, or as a significant factor influencing the behaviour of the macroeconomy.
Post a Comment