A modest proposal for paying negative interest on currency (or something)

Willem Buiter considers various ways to make interest rates negative. The problem is how to pay negative interest rates on currency. His most interesting proposal is to separate the unit of account from the currency. The dollar would remain the unit of account (at least, he hopes it will). But he would replace dollar notes and coins with a new currency, the "rallod", which would depreciate against the dollar. If it depreciated at (say) 5% per year, it would allow nominal interest rates on dollars to go down as low as minus 5%, which should be sufficient to get us out of the recession.

It seems a very complicated proposal, when you read it. It certainly fails my "Grandmother" test. (Grandma could never understand Britain's new decimal currency, and the switch from shillings and old pennies to new pennies, so refused to deal in anything smaller than the pound, which stayed the same.)

So I decided it needed to be simplified.

My first idea, rather than changing the currency, was to change the other units of measurement. After all, we buy apples in kilograms, milk in litres, labour by the hour, etc., so why not just redefine the kilogram, litre, and hour so that they got 5% smaller per year? So even if the price of apples in dollars per kilogram stayed the same, at the end of the year we would get 5% fewer apples per unit of currency. Exactly the same as if the "rallod", were worth 5% less in terms of dollars.

But after careful consideration, and long consultations with colleagues in other departments, I decided that my idea would not make for good relations between economists and other scientists, who seem rather attached to their existing units of measurement. The engineers in particular seemed to suffer from "unit illusion", and couldn't adjust to a world of inflation. So I abandoned my first idea, and returned to thinking about Willem's proposal.

How to simplify Willem's proposal for the "rallod"?

Well, the name "rallod" has to go for starters. It's ugly, and sounds far too radical. The common convention for currency reforms is to introduce a new currency, and call it the "New...." whatever the old currency was. The "New Franc" replaced the old Franc in France. So let's just call it the "New Dollar". And to be doubly sure there is no misunderstanding, we can refer to the old dollar as the "Old Dollar".

And it would be a major hassle and expense to call in all the old dollar coins and bills and print up new dollars to replace them. It would be much simpler for the government and central bank to just make a declaration: "Henceforth, all existing dollar notes and coins are now declared to be New Dollar notes and coins!". One problem solved!

Now, since the New Dollar will be depreciating at 5% per year against the old dollar, and all existing contracts are denominated in old dollars, we will have to remind people that the number of New Dollar notes they will need to pay to fulfill any existing contract (if they choose to pay with currency) will need to be increasing at the rate of 5% per year. That would apply to all debt contracts, wage contracts, price contracts, etc. And the courts would of course enforce that interpretation of the contracts. And there is nothing unjust about doing this, of course, because the New Dollars won't be worth as much as the old dollars.

Alternatively, and perhaps it would be simpler, we could just rewrite all the old contracts, add a 5% per year premium to any price, wage, or rate of interest, so that payment could be specified in New Dollars, if that's what people wanted to pay in.

Actually, now that I come to think of it, do we really need the ugly neologism "New Dollar"? Since we wanted to avoid the cost and hassle of printing new notes and coins, the notes and coins still say "dollar" rather than the "New Dollar". So, to avoid confusion, let's keep the word "Dollar" for the New Dollar. And use the words "Old Dollar" (or "Classic Dollar"?) for any contract written before the New Dollar...I mean the Dollar...was introduced.

Actually, now I think of it some more, isn't there only one thing we need to do? Just pass a law saying that all contracts written in dollars before a certain date must have an additional 5% per year premium written into them?

So a 4% mortgage would now become a 9% mortgage, by law. A contract specifying a 3% wage increase per year would now specify 8%, by law. And so on. And just to anchor inflation expectations to the new currency regime, to help people adjust, the central bank should announce that the inflation target will be raised from 2% to 7%, and that whatever inflation rate people had previously expected, in old dollars, should now be revised upwards by 5%.

So, to recap: raise all existing nominal interest rates by 5% per year; raise all existing wage and price contracts by 5% per year; raise the inflation target by 5% per year; and tell people to expect inflation to be higher by 5% per year. New interest rates and new wage and price contracts can be set wherever they need be, of course. That's my simplified version of Willem's proposal.

Why wouldn't my version work? Only the names are different.

Comments

[reason]

You know my view on this. If nominal interest rates have to go negative we have a more serious problem and have to look to another solution. My solution - print money and spread it around (the problem is that level of debt to base is unsupportable - we need more base money and the wider it is spread the better). Then we need higher interest rates to encourage savings and discourage borrowing (and avoid inflation).

[reason]

Nick,
by the way - your version sounds a lot like what Steve Keen has been saying about debt forgiveness.

[Adam P]

pehaps all we need to do is seperate unit of account from medium of exchange.

Here's a quote (my own) from our discussion on the other thread.

"We all know that the aggregate endowment next period can't increase, if a helicopter gives us all more money and I spend my share but you save yours then YOU will succeed in consuming more next period at my expense! You'll get a greater share of tomorrow's endowment at the expense of giving me a larger share of today's endowment. That's exactly the deal you would like BUT it's not the deal I want. We both want more tomorrow in return for less today. Hence, we both save the helicopter money just to keep up with each other."

this was said in explaining why more money doesn't satisfy our demand for more real savings. It gets held as unit of account. Give me money, government certified medium of exchange, but have it expire in 1 year and I'd be crazy not to spend it!!

We all just get Federal Reserve debit card that with $10,000 balance but it expires in 1 year.

[Adam P]

Of course, you have to forbid buying government bonds with the debit card.

[Adam P]

I just realized why the debit card thing won't work. Vendors would never accept money that expires. Forget it.

[JKH]

My solution:

a) Adopt Adam P’s proposal for a $ 10,000 FR debit card term credit. It’s as sensible as any I’ve seen. (Adam, it’s not the money that is used from such a card that expires, but the money that remains unused.)

b) Change the name of the renowned but venomous FT columnist to Retiub Melliw, and starting printing his columns backwards. That’s sensible too.

c) Nick, publish your grandmother’s findings in full. That’s probably the most sensible.

[Patrick]

If I'm a household with no or little debt I can see how this might work. But say I'm a household with a debt problem. I'm not spending because I am paying down debts. Along comes Nick and gives me 5% raise, makes everything 5% more expensive and jacks-up the interest on my debts 5%. It's a wash. My behaviour stays the same.

Instead, say Nick comes along with his wage & price controls in reverse, but also writes down the principle on my debts by 5%. Now I've got some breathing room and I'll go buy the kids those new shoes they need, or take the family out for dinner etc ...

I like the debit card idea. The progressive in me wants to make the amount on the card inversely proportional to income - poor people get more money than rich people. It would be even more effective if you could say the money on the debit card must be spent; it isn't redeemable for cash. I'd even make it so that the unused balance decays linearly over time instead of just expiring.

[Phil]

Is this just a clever way of saying, let's print more money to create an extra 5 points of inflation, and adjust previous contracts to take that into account? Sorry if this is a dumb question, I'm a non-economist.

[reason]

Adam P
The problem with your solution (which I thought was a good idea when I first heard it as well) is that money is fungable. People will spend the money on the card, but save other money.

[reason]

But Adam P,
I think you are on the right track, money as a medium of exchange is in conflict with money as a store of value.

P.S. This is even clearer with a gold standard where people think gold is something real. If people start to hoard it, then there is less in circulation and so you get deflation. This makes hoarding it more attractive and so people hoard more. Then people with debts (denominated in gold of course) are in big shit! With gold of course there is no lets create more and spread it around recourse (making hoarding less attractive because of inflation) for the monetary authorities to get out of this trap.

[Nick Rowe]

There is a way to make Adam's proposal work: you just randomise it. Pick a number from 00 to 99 out of a hat every month, and all notes with a serial number with those last 2 digits are *declared* invalid.

Of course, this raises the question of how exactly government "declarations", if they are just "cheap talk", are supposed to change the equilibrium. This is why I hate the phrase "fiat money". In the Latin bible, God in Genesis says "Fiat lux!", and there was light. But governments do not have God's power. I prefer a sort of Menger/Mises story of the origins of *fiduciary* money out of an original commodity or convertible money. Paper money rests on faith, not government fiat. The opposing, "fiat money" story, is at its most extreme in Knapp's (?) State Theory of Money. I hear echoes of that story in the Fiscal Theory of the price level.

What if people ignored government declarations on what is and is not money, and what it's worth? They sometimes do.

This problem underlies my and Willem's proposals.

When Cambodia wanted to introduce a new paper money ex nihilo, after the Khmer Rouge had totally destroyed the old one, they first made the new paper money convertible into rice.

Phil: yes, basically. But there are also two underlying issues: the role of government declarations or "fiat"; the indeterminacy of inflation rates under interest rate control.

reason: but forgiveness of debt is roughly equivalent to lowering interest rates on existing contracts. My proposal would raise them. I don't think Steve Keen would like it.

Only old Brits may remember the controversy and confusion caused by Prime Minister Harold Wilson's speech when, following a devaluation of the pound, he tried to explain to "the man in the street" that it would not mean "the pound in your pocket" is worth less. http://en.wikipedia.org/wiki/Harold_Wilson#First_term_as_Prime_Minister_.281964.E2.80.931970.29

[Phillip Huggan]

Changing the units of measure to test a monetary innovation?!

Another problem is I don't think Central Banks do their job well enough to handle this kind of power. There is a political risk of raising interest rates too high. People get unemployed and your poll ## go down. But it isn't clear there is that same immediate downside risk by going too low. I can imagine this policy exacerbating bubbles the more I think of it. A generation after the Chicago School is purged from University textbooks, maybe. It has occurred to me economics is about keeping Neocons out of power rather than about helping society progress efficiently. Double-edged swords are a no-no for now IMO.

[Alan]

I hope the "unit illusion" crack was just a troll.

[Adam P]

reason, just dropped a response to you on the deLong blog.

[Nick Rowe]

Alan: Not really a troll. I doubt many engineers read this blog anyway. And if I had to put any one group in charge of the political economy (other than economists of course), engineers would be at the top of my list.

It was very tongue in cheek, of course, even silly, but did have a serious purpose. I wanted to draw the parallel between changing units of measurement, separating the unit of account from one familiar concrete medium of exchange (currency), and inflation. The engineers' "unit illusion" crack was my way of doing this. Practical people (engineers rather than theoretical physicists, ordinary shoppers rather than theoretical economists) don't like us messing with their system of measurement.

[himaginary]

Maybe it’s useful to think from Fisher equation’s point of view.
http://en.wikipedia.org/wiki/Fisher_hypothesis
From Fisher equation, nominal interest rate equals real interest rate plus inflation rate. In your proposal, you're treating inflation rate as endogeneous in Fisher equation model. However, as inflation rate is actually exogeneous in the model, raising nominal interest rate is likely to just end up in raising real interest rate at the same amount.

And even if you assume inflation rate is endogeneous and it goes up at the same amount of the hike, real interest rate would remain unchanged, so the problem remains unsolved. It is real interest rate we want to pull down at the end of the day.

[Don the libertarian Democrat]

I don't see the problem:

http://blogs.ft.com/economistsforum/2008/11/the-case-for-negative-interest-rates-now/#more-259

"These include the periodic stamping (for a small fee) of banknotes (without the stamp they would not be valid). The idea is that by imposing a running tax on banknote hoarding, nominal risk-free rates could fall to negative levels.

In the age of the information technology revolution, surely the authorities could devise a simple and practical method of effective taxation of banknote hoards?

There are two cues to a practical method of taxing notes. The first comes from what happened during US financial crises in the 19th century.

Banks under stress of cash drains (depositors withdrawing funds) suspended temporarily the 1:1 link between cash and deposits, so their notes sold at varying discounts. The second comes from the launch of the euro; a conversion of old banknotes into new.

These cues lead to the solution.

The relevant government would announce that existing banknotes were to be converted into new notes at a fixed date, say three years from now, at a discount (for example 100 old dollar banknotes would be converted into 90 new).

In the interim, 1:1 conversion of banknotes into deposits would be suspended. Instead, a crawling peg would be established. At the start, the exchange rate between deposits and banknotes would be virtually 1:1. At the end it would be 0.9 banknotes/deposit.

As the discount grew, retailers would quote different prices for cash or cheque/card settlement. And as to the note switch-over costs, the “experiment” of Europe’s economic and monetary union demonstrates the feasibility.

The looming conversion would provide an essential degree of freedom for monetary policy. In terms of our illustrative arithmetic, the risk-free interest rate could fall to a negative 3.33 per cent a year without triggering cash withdrawals from the banking system.

Is the exercise worth it?

The main reason for believing it is stems from an appreciation of how the bursting of a global credit bubble influences the equilibrium level of risk-free interest rates relative to risky rates of return and in absolute terms.

Most of us would agree that the bursting process ushers in a period during which soberly-measured risk premiums increase sharply.

This means that the risk-free rate must plunge to be consistent with an average overall cost of capital which reflects the new glut of savings.

So, in terms of our illustrative arithmetic, it is plausible that the neutral risk-free nominal rate of interest in the US and Europe, especially taking account of a likely near-term drop of the price level, is significantly negative.

The central bank and government, by devising a system in which such negativity can express itself, can give a big fillip to the recovery process."

If it's one thing the govt can do, it's tax.

[reason]

Nick,
sorry I didn't read the last bit about the debts - which sort of undoes the rest. You clearly don't believe in Fisher.

[Nick Rowe]

reason: I do believe in the Fisher equation. But as you say, how you interpret it all depends on whether you treat as endogenous or exogenous. And whether you are talking about new loans or old loans (are we talking about the expected real rate of interest ex ante, or the realised real rate of interest ex post?). I was sort of playing with those interpretations at the end. If the central bank can change expected inflation just by declarations, we get a very different interpretation of what is happening to real interest rates than if expected inflation were exogenously fixed (in the short run). We are on the same page.

Hi Don!

There is nothing you can do by taxing money or introducing a new currency that could not be done by changing the expected rate of inflation.

Fiat/fiduciary/paper money is just a symbol, like language. (Structuralism?). Under interest rate control, there is nothing to anchor the meaning of words/money (the price level), except Lewis-conventions (inertia of actual and expected inflation). If we go to bed in one equilibrium and while we slept a magic wand doubles all the prices, and expected prices, we would wake up next morning in the same (real) equilibrium. Just the same as if a magic wand changed the word "cat" to "dog" and vice versa, in our minds and books.

When we are in a Lewis/Schelling conventional/coordinated equilibrium, we stay there, even though there are multiple equilibria. But a symbolically important "declaration", even if it is mere "cheap talk" (not backed by real actions that alter the payoff matrix) can change the focal point and switch us to a different (nominal) equilibrium ("nominal" = where all the prices/inflation rates/meanings of words are different).

The Brits drive on the left. No individual driver has an incentive to deviate from the left equilibrium unless he expects all others to deviate. The government could switch the equilibrium to driving on the right by declaring that Britain is now part of Canada (change the currency). Or it could do it by declaring that Brits will now all drive on the right (announce a new rate of inflation). Either would work, if and only if it is believed to work.

Do dictionaries determine the proper meanings of words, or do the meanings of words determine the proper dictionary?

[Nick Rowe]

Don, continued: That's why I was playing with the meanings of words dollar, new dollar, old dollar, etc.

On way to change the equilibrium that is not merely a declaration (cheap talk) is for the government to do something real (get the army trucks to start driving on the right). By pegging the time path of the price of some real good, by buying and selling gold, or whatever, and abandoning interest rate control (which just swaps money for future money) we are no longer in a Schelling/Lewis coordinated game. There is only one equilibrium.

[Nick Rowe]

Inflation is a tax on currency. And we don't have to go to any hassle of collecting it. Just print more money. Inflation means a negative (real) interest rate on currency.

[Nick Rowe]

I should probably do a sensible version of this slightly silly post sometime, making all this stuff explicit, rather than implicit.

[reason]

Nick
I think your answer addressed to me, was actually to himaginary.

Adam P
I replied to your reply at Brads place.

[reason]

I meant by the way not Fisher's equation, but Fishers concept of debt deflation. If you keep the real value of nominal debts the same despite deflation, you are not avoiding the problem. The debt-deflation spiral comes from the fact that deflation increases the real value of debt principals (someone borrowed 50 cattle and has to pay back 60). Debt forgiveness was designed to stop the real burden of debt from growing unsupportable.

[reason]

And P.S. Adam P, glad you made the right interpretation of my post at Brad's - I find your posts usually very good, so I guessed you must have expressed yourself badly.

[Nick Rowe]

reason and himaginary: yes, sorry, I got muddled, and therefore muddled between the Fisher equation and Fisher debt-deflation too.

Reason: yes, inflation will be different from a tax on currency in terms of the Fisherian debt-deflation effects. Some sort of debt foregiveness (or changing the nominal interest rates on existing debt) would be needed to make them equivalent.

[Don the libertarian Democrat]

Nick,

I think that I agree with you, which is why I said the following on Buiter's blog:

"I'm less concerned on the method used, than on the use of the concept"

However, being a dunce, I'm attracted to views that I ( mis ) understand. QE and Stamping are just clear positions to me, and work towards the desired goal by clear incentives and/or disincentives. Stamping is a disincentive to buy short term bonds, as I envision it. QE is a solution to Debt-Deflation. I suppose they come to the same thing, or, as you seem to say, could come to the same thing. I avoid theory at all costs nowadays.

[Nick Rowe]

Don: "I avoid theory at all costs nowadays." ??!!!

Don, by hanging around this and Willem's blog, and thinking and commenting, and running your own blog, with thoughtful commentary, I would say you are doing a really bad job of avoiding theory ;-)

[Don the libertarian Democrat]

Nick,

I can't tell you how hard I laughed when I read your comment. Too true. Too true. As Wittgenstein said:

Nothing is so difficult as not deceiving oneself.

Take care,

Don

[Scott Sumner]

Nick, For some reason I am totally confused. Is this like a monetary reform? I.e., like swapping 100 old pesos for one new one? (Except obviously in the direction of inflation rather than deflation.) Is the goal to affect any real variables? And if so, how?

Or is it all a joke that went over my head?

[Nick Rowe]

Scott: it's a joke, sort of, or perhaps semi-satire. And I'm confused too. My writing it that way was supposed to reflect my own puzzling over the paradoxes. Some day I should try to write a sensible version.

The whole thing is like the 100;1 currency reform. Only that's a currency reform in levels. Proposals to introduce a new currency, that would depreciate against the existing unit of account, is like a currency reform in rates of change.

The serious questions are:

What can be accomplished by introducing a new currency that could not be accomplished by raising the expected rate of inflation?

Why do "names" (nominal variables) matter?

How can a government "declaration" divorce the medium of exchange from the medium of account, while a government "declaration" cannot change the expected rate of inflation?

How is the rate of inflation determinate under interest rate control?

My post was a (probably failed) attempt to explore those issues by writing a tongue in cheek satire.

My most sensible answer is in my replies to Don the Libertarian Democrat above. But Don is a philosopher, so I couched my explanation in terms of the philosophy of language.

Very short version: Under interest rate control, the inflation rate is like the equilibrium in a coordination game. Does currency reform shift the equilibrium by changing the focal point? Why cannot an announced increase in the rate of inflation do the same thing?

[Phillip Huggan]

"The serious questions are: Why do "names" (nominal variables) matter?"

I don't know, Frank.
A problem with introducing a depreciating currency is that you are also shifting the portflio of consumption matters. Basically it is the equivalent to issuing a whole whack of TV commercials. We have an $80B looming annual health-care deficit (also our deficit thx to tax cuts). We are 56th of 57 countries in environmental performance (kicking Saudi ass, yay). 20th among OECD countries in delivering daycare. No federal low-income housing strategy since Trudeau.
I haven't thought deeply about it but I thinking encouraging Canadians to spend money quickly will reinforce short-term, often negative quality-of-living rates of consumption. If you could just make the money dependant upon freeing up hospital beds or something, sure. But depreciating currency might just be spent on beer and popcorn. Getting out of a recession could be stimulated in a way that increases long-term growth rates, but I think (correct me if wrong) forcing quick spending doesn't mimic existing capital allocation; I think long-term projects (like retrofitting buildings with copper to be staph and flu proof) require long-term planning, planning that might not be possible with a time limit. But not sure. Anyway, Iceland is an obvious guinea pig for any economic innovations like changing the metric system or changing glasses prescriptions to rip people off when buying milk.

[Patrick]

Nick: "it's a joke, sort of, or perhaps semi-satire ... ome day I should try to write a sensible version"

Well, outlandish ideas can sometimes come in handy. Had things gone a little differently in Sept/Oct and unemployment was now 30%, we might be seriously thinking about devaluing the kilogram or legislating inflation.

[dan robles]

The term Rallod was first used by the Ingenesist Project several years ago and documented in a press release in October 2008.

The rallod is simply dollar spelled backwards. The rallod is a hypothetical currency that represents future productivity backed by current innovation. I derived the term from 'dollar' which for all practical purposes is a currency that represents future productivity backed by a current promise of future productivity (debt).

The underlying assumption is that the only way to increase productivity (create more dollars) is to innovate (spend rallods) on better, faster, more efficient means of doing things, i.e., increased productivity.

Since 'innovation' and 'debt' are a mirror image of the other, and if the logic holds, then their proxies, 'rallod' and 'dollar' are also a mirror image of each other, hence the term.

There are similarities in our definitions because obviously there is a deficit of human productivity caused by various constraints and failures in all of the traditional factors of production. Negative productivity growth is negative dollar growth and therefor negative interest. Negative interest at least keeps the old game in play.

Since both the rallod (as I have defined) and the dollar are proxies for human productivity, they are of the same species. Since the value of currency is largely a social agreement, the convertibility between the two should not violate the psychology of the markets too much (passing the grandmother test). In fact, the notion that the scorecard is based on real productivity instead of speculative productivity should be very encouraging for all (again, grandma would be pleased).

Our "Next Economic Paradigm" from the Ingenesist Project specifies an economy based on rallod currency.

Please feel free to use the term as defined here. Let me know what you think, find, or develop from this discussion. Let me know if you have any questions.

Dan Robles
The Ingenesist Project

[Kien Choong]

Dear Nick - a practical way to achieve negative interest rate on cash is simply for the government to impose a modest tax on cash holdings - say 2%pa. The tax could be imposed on financial institutions. The revenues raised from this tax could be used to offer a rebate on non-cash assets held by financial institutions - e.g., long-term corporate bonds. This would be a subsidy for holding risky assets and would facilitate the circulation of money into the real economy.

Willem Buiter's proposal to tax currency by issuing currency with an expiry date seems unnecessarily complicated. The government can easily achieve the same effect by imposing a straight forward tax on cash holdings (as suggested above).

There is no real need (I think) to tax cash holdings held by every individual in the country. Just a tax on cash holdings of financial institutions seems sufficient. However, it would be open to the government to also impose a tax on bank deposits. Again, the revenue raised could be used to provide a tax rebate on investments in (say) equity.

The government could enact legislation to impose the tax (and rebate), but delegate the power to determine the amount of the tax (and rebate) to the central bank. The central bank would then have a tool to implement negative interest rates on cash holdings, and encourage financial institutions to invest the cash in risky assets, thereby reducing the yield on risky assets. By altering the amount of the tax on cash holdings (e.g., from 1% pa to 1.1% pa), the central bank would be able to alter the negative interest rate from time to time (e.g., once every quarter).

What do you think?

Best,

Kien