More on helicopter money, prompted by Simon Wren Lewis’ post, and exchanges with Eric Lonnergan and others in the comments appended to my last post on this.
First, responding to Simon. He says that I argue “that if the central bank assumes money is irredeemable, and starts printing a lot of it, people may stop wanting to use it. If they do that, it will no longer be seen as wealth.” Simon remarks: “This is real angels and pins stuff that can come from taking microfoundations too seriously.”
Most people won’t click through to my post, and if you read that paragraph, you would not get why I wrote the post, or what I think about helicopter drops and why. The message is: ‘Tony, taking a too-literal reading of the wrong microfounded model, decides helicopter drops won’t work, so don’t take him seriously.’
I’m against helicopter drops. That bit is faithfully got over. And I worry about them using three lines of reasoning.
1) OLG and ‘new monetarist’ models of money, in which money is not redeemable. But in which we can show that some monetary policies generate money that has value, and some [like excessive money creation] destroy it.
2) casual historical empiricism that money’s value has been destroyed by excessive money financing.
3) I look at Buiter’s proposal to take the standard non-microfounded model of money [which assumes money has liquidity services, and that at the end of infinite time is redeemed] and to remove the redeemability assumption. And decide that it doesn’t get us closer to deciding what theory says would happen if there were helicopter drops. Why not? Because this begs the question about why we should assume money has those liquidity services.
Whether 1) and 3) should be part of the calculation to helicopter money in or not depends on whether you want to use theory to answer the question.
Simon goes on to say: “Just ask yourself what you would do if you received a cheque in the post from the central bank.”
This urges us not to worry about trying to find a theory to guide us about whether money will be felt as wealth or not. It just is, isn’t it, by introspection. Well, no. Sure, I can get the answer from myself that I will go ‘wtf is going on with policy now, are we really this far up sh!ts creek? Whatever, I had better try to spend it’. But figuring out what happens next, in general equilibrium, whether people would feel more wealthy or not, and whether this amounts to a worthwhile policy or not, is much harder. Which is partly why theory and empirics might be useful.
Simon then says: “As Nick Rowe points out in this post, we can cut through all this by noting the link between money creation and inflation targets. The money required to sustain an inflation target will not be redeemed, so it can be regarded as wealth.”
If this is intended as a point of theory, (I’m not sure whether it is or not) then it’s not right. And if it isn’t intended as theory, then redeemability is irrelevant. To explain.
Intended as theory: if we think theory is useful, we have two choices of model. We might decide to consider only overlapping generations or new monetarist models. If we do, then the thinking they do to the effect that money is not redeemable is pointless: in these models it isn’t, period. The question is how different monetary policies (eg different amounts of helicopter drops) affect the value of money. Some preserve it, and some, we know, destroy it.
Alternatively, we might consider models like the one Buiter studies, and doctors, these being models that assume liquidity value of money, and either assume redeemability, or don’t.
But neither does the Simon-Nick thought process make sense for this model.
Which assumption we make, whether you go with Buiter or not, is a once and for all decision about what the public sector budget constraint looks like out into infinite time, period. Given this choice about the appropriate way to view the constraint placed on governments, one then figures out optimal monetary and fiscal policy (and what that means for the sequence of money and bonds and taxes that appear in the sequence of budget constraints each period). Or one says ‘what happens if the government attempted to follow such and such a monetary and fiscal policy, say involving an inflation target of y per cent?’, and then works out the same sequence. In trying to work this out, one might find that certain combinations of candidate inflation targets and fiscal policies are not feasible. To repeat, as a mathematical matter, in describing a working artificial economy, we don’t say ‘we can dispense with the redeemability assumption because all money created in pursuit of a given inflation target won’t be redeemed’. In fact, in the redeemability version of the model, we never get to the end of time, and so money might (depending on choices/outcomes for m) never actually be redeemed!
If Simon and Nick’s remarks are not points of theory, then redeemability is not something to concern ourselves with. Instead, we would forget about it and turn to empirics to figure out what would happen following a helicopter drop.
If you are still with this post, further points are made in the exchanges around the notion that helicopter drops in pursuit of an inflation target need not generate hyperinflation and the corresponding destruction of money’s value.
Perhaps! But perhaps not. For example, the inflation target in the UK is a very fragile thing. The Treasury can change it whenever it wants by writing a letter. It did already once, to change definition (or was it to create a bit more inflation through the back door?). It could do it again. (After all some, like Krugman, Blanchard, Summers (and me) have been urging that this is a good idea, in case the natural rate will be very low for a while. ) It can even take back control over monetary policy for short periods if it wants to. Or for longer periods with a vote in Parliament. The entire BoE act could be repealed by a simple majority. So I don’t view the inflation target as a cast iron protection against helicopter drops undermining monetary and fiscal policy. There’s a good reason why monetary financing is outlawed by the Treaty of Rome. Allowing yourself tightly regulated helicopter drops is not time-consistent. Once government gets a taste for it, how could it resist not helping itself to more?
Simon, Eric and others might be right. It could work safely. But what is the point of taking the risk, when there is ample room for more fiscal stimulus of a conventional sort that leaves our monetary and fiscal affairs intact? Or for more unconventional operations exchanging reserves or short-dated gilts for private sector assets? If we were at 150% debt/GDP, or over-extended as far as QE or credit easing was concerned, I might be prepared to countenance it. But now, this debate seems frankly, well, academic!
Tony: I was intending to make a point of theory.
Assume that all central bank money will be redeemed for goods in 2084, at a price of $100 per basket of goods. The central bank will then start afresh, with a new money. So the price level will be 100 in 2084.
Now suppose the central bank doubles the base money supply, by helicopter, in 2014. And at the same time announces that in 2084 it will redeem all the money at a price of $200 per basket of goods. So the price level will be 200 in 2084. And the central bank’s real liabilities (and the government budget constraint) are unchanged.
But at the previously-existing price level, the real value of the stock of base money is doubled. Net wealth has increased. The 2014 price level would need to double to reduce net wealth to its original level.
(Or you could point to the substitution effect, of 1% (percentage point) higher expected inflation over the next 70 years.)
If I give you an asset, that is worth $2 to you, but only worth $1 to me, then our net wealth increases by $1.
“The entire BoE act could be repealed by a simple majority. So I don’t view the inflation target as a cast iron protection against helicopter drops undermining monetary and fiscal policy.”
But isnt that also the case under current policy tools? We need greater accountability so the monetary authority abides by its targets regardless of what the policy tools employed by the central bank are.
“Once government gets a taste for it, how could it resist not helping itself to more?”
Your right here. I think monetary and fiscal cooperation may only be a good idea strictly on a temporary basis. Helicopter drops should be done independantly by the central bank though issuance of emoney. Here’s my brief proposal: cmamonetary.org
Pingback: Links for 10-24-14 | The Penn Ave Post
but what is it about how you think the real world works, that makes you against helicopter drops?
The potential benefits are alleviating an economic slump that is causing big problems all over the place. The potential costs are what? What are you worried about? Forget models. If you are really opposed to real helicopter drops then what do you think might really happen? And what makes you think the risk of that bad outcome outweighs the potential benefits?
The EU is facing deflation, the EU has a debt problem. Those two birds could be killed with one stone (the debt problem, only indirectly – I’d also favour ECB purchases of debt to be rolled over until certain targets – say NGDP growth – met, in addition to sending cheques to citizens, to get at the debt problem more directly). I am tempted to say that your adherence to thinking through a very limited set of highly unsatisfactory models would, if you were in a position of power, stop you from taking the only real course of action we have to get out of the mess we are in, which would be a very large black mark against the use of mainstream economics in policy making. I know this begs the question, but I do not know it’s wrong.
sorry, I’d like to withdraw that comment which was written before properly reading your post. You are worried about hyper inflation and governments that won’t stop once they’ve started. Plus you think we still have other options.
Points No.1 & 2 in the article above are irrelevant in that it’s obvious that “excessive money financing” leads to excess inflation. But we are not concerned with excessive money printing. We’re concerned with an attempt to print enough to give us full employment, but not so much as to give us excess inflation.
The only circumstance I can envisage in which excess inflation could kick in BEFORE full employment is reached would be if the population was excessively Ricardian: i.e. where firms and households took a keen interest in the money supply figures and got neurotic in the event of it rising a smidgeon. The reality is that a large majority of small firms and households take NO INTEREST WHATEVER in the money supply figures.
There has of course been SOME INTEREST in those figures over the last two or three years in that QE increased the money supply, and a few ignoramuses predicted excess inflation as a result. But the actual effect was muted in the extreme.
As Stiglitz put it “Ricardian equivalence is taught in every graduate school in the country. It is also sheer nonsense.”
But the fact that something has no empirical basis never stopped economists treating it as true. Ricardian equivalence keeps thousands of economists employed with their models in ivory towers the World over.
Next, Tony Yates wants “fiscal stimulus of a conventional sort” instead of money printing. “Fiscal stimulus of the conventional sort” involves printing government debt (i.e. a promises to pay the creditor base money at some point). What’s the difference between that and printing base money? Not much.
Tony Yates is also concerned that “Once government gets a taste for” money printing, it will print excessive amounts. Well the solution to that is to put money printing decisions in the hands of an independent committee of economists (as suggested by Positive Money). Oh wait a moment: that decision ALREADY IS in the hands of such a committee. In the UK that’s the BoE MPC. I.e. the MPC can print money in that it can do QE.
What’s the difference between money and bonds: “not much”. Cool. That’s one thing I can stop worrying about then.
I’m delighted to have reduced your “worries”. Just to expand on my point and reduce your worries even further, both base money and debt are assets as viewed by the private sector. Plus they are both liabilities (of a sort) of the public sector). Plus there is no effective difference between £X of base money and £X of debt due to mature in a week’s time: the latter is simply a promise by the state to pay the debt bolder £X in a week’s time. So there are no effective differences there. In contrast the difference is obviously more significant in debt with ten years to maturity
So in as far as a complicated point can be compressed into a few words without loss of accuracy, my point that there is not much difference between base money and public debt is approximately correct.
As I wrote on SW-L’s post, I think it is just too strong to say that Ricardian equivalence is false. It is just weak. As is, I think, your dismissal of the risk of inflation arising from QE. As far as I am concerned, when QE increases the base money supply several fold, you have to be an ignoramus not to raise your assessment of the risk of high inflation. And, while ordinary investors are unlikely to go into the details of central bank balance sheets and follow money supply statistics to produce a numeric expectation of inflation, I think they instinctively know that at the moment central banks are just making up policy as they go along to forestall an asset price correction, so investors bid up real assets, like houses, commodities, farmland etc.
Academic economists could do themselves a favour by being less dogmatic and more eclectic. But then, that probably comes from an insistence on working with mathematical models and trying to keep them simple.
On REQ: there’s lots of micro evidence that agents are credit-constrained, or suffer from myopia. Plus, time series evidence of identified fiscal shocks shows that these have pronounced effects, implying a breach of REq. But, note, for this debate about helicopter drops, we don’t need that REQ fails anyway. That is an important side issue about the effects of all policies, h drop or otherwise.
But Tony, you are trying to uphold the all or nothing approach. I know for a fact that Ricardian equivalence works to some degree, because as a regular visitor to Japan, I have met elderly people who hang onto their savings in anticipation of tax increases. I imagine it is a weak effect, but I don’t think you can just leave it out of a mathematical model, if you insist on working that way.
For what it is worth, I am against helicopter drops because I would expect them to be a source of instability. If you just evenly or randomly gifted people money, I would expect them to begin to spend it in anticipation of higher prices, whereupon an inflation-targeting central bank would soon have to tighten monetary policy. Individual recipients could make their own choice whether to spend the money and maybe avoid a small step change in prices, or, more Ricardian, save it in anticipation of enjoying a brief rise in short-term interest rates as the central bank tightened (obviously, the situation is somewhat complicated by the fact that central banks typically target inflation rather than a price level). However, if the government then refused to stand behind the central bank fiscally if necessary, it would precipitate a constitutional crisis.
The debate about helicopter drops, and indeed all extensions of monetary easing is a waste of time. Monetary policy has done, in fact had done by about 2009, all it can do without making matters worse, and the sooner that monetary policy experts get down to discussing how to unwind inflated asset prices, normalise monetary policy, and get on with structural reform, the better.
Hyperinflation is like an avalanche. Forces can build up for a long time and then things move fast.
I thought OLG models (I presume you talking about Wallace type models here) tended to be about monetary assets, rather than money, in that they deal with demand for a store of value rather than the need for something to facilitate payments. In which case, do they say anything different about helicopter money to what they say about fiscal stimulus funded with debt issuance? Or is there some assumption that debt issuance will necessarily be temporary and money issuance not?
Can you say a bit more?
Answer to final q is yes, because, provided we rule out default, bonds always valued, but money not always valued.
Taking Wallace’s model, I don’t think his results depend in any way on his money being the medium of exchange. You could just as easily imagine that the good was the medium of exchange and people were buying and selling units of money purely as a store of value. You would still get the same results, wouldn’t you? Wallace seems to say as much in his conclusion.
So in terms of this analysis, at least, there would seem to be little difference between monetary units that are perpetual (money) and monetary units that have a set redemption date (bonds), but that are redeemed by delivery of perpetual units. Both serve equivalently as a store of value.
I’m not sure I understand your reply there. How can bonds be valued when money isn’t if bonds are only ever a claim for money?
Which paper are you talking about? One can assume that bonds are redeemed directly by real consumption goods, for example. In fact, this is a very widely used assumption.
“The Overlapping Generations Model of Fiat Money”
Click to access cp49.pdf
When you say redeemed directly by consumption goods, do you mean goods are purely the settlement medium or do you also mean that the redemption value is fixed in terms of goods? In other words are these bonds that pay 100 apples or bonds that pay $100 worth of apples?
The first clearly doesn’t depend on money having a value, but the latter does, doesn’t it? I presume when we are talking about government debt, we’re talking about bonds redeemable for a value determined in money units.
The former. We could study nominal bonds, but haven’t so far mentioned it.
I’m rather surprised. It would seem to make quite a bit of difference to this question and I would have thought it made more sense to study nominal bonds, as that’s what governments tend to issue.
I don’t think it does, not to the basic issue of how one should think of h drops and money. It does to issues about bond pricing, optimal monetary policy generally, default, of course. Anway, forgive me if I move on now.
helpless, helpless, helpless
what are we to do
thats of course is the commonly engendered feeling when discussing economics today
from the supply side: we wont produce enough if we give people money, everything will collapse
if we end up with supply side issues it will only be because we have destroyed our environment by pollution, allowed our infrastructure to crumble or destroyed our wealth with weapons of war
we clearly have enough for everyone
lets just figure out an intelligent way to include everyone
“helicopter” money is simply a way to redistribute wealth to counteract the greed of the rich
if we could directly tax the wealth of rich to counteract the constant redistribution of wealth towards the rich
or if we could increase wages and increase employment to take care of everyone
then we wouldnt need “helicopter money”
helicopter money is a work around, not the ideal
As for governments getting a taste for it, doesn’t the 70s answer that? While anything can happen, it doesn’t mean it is at all likely, and nothing goes on forever. Yet we have become accustomed to failing with all those tools.
Pingback: WEEKEND READS: Catch Up On The Best Economics And Finance Links Of The Week (DIA, SPY, SPX, QQQ) | #TheTrend
Pingback: Jordan Meyerowitz | WEEKEND READS: Catch Up On The Best Economics And Finance Links Of The Week (DIA, SPY, SPX, QQQ)
Pingback: Por que não helicopter money? « Desvio Colossal
Pingback: A short reply to Tony Yates (on helicopters) | philosophyofmoney
Pingback: The Constant Reserves Multiplier | philosophyofmoney
Pingback: Helicopter money: setting the tale straight | Bank Underground
Pingback: The Constant Reserves Multiplier - Philosophy of Money