Most macroeconomists believe that the real interest rate is what equilibrates AD and AS. If the real interest rate is too high, AD will be too low, and lower than AS, so there's excess supply of output. And the fall in AD will cause an equivalent fall in output. That's the orthodox Neo-Wicksellian position.
I want to consider a thought-experiment: assume the economy is initially in equilibrium at a 3% real interest rate. (The natural rate of interest is 3%). The Pro-Usury Party then gains power. The new PUP government passes a law that says all loans must be indexed for inflation, and the real rate of interest must be a minimum of 10%. The governor of the central bank pleads for an exemption from the law, on the grounds that money is barren. The PUP allows that central bank liabilities, like currency, can pay 0% interest. The central bank is also granted the power of compulsory purchase of bonds, if it wishes to do an open market operation.
What happens? In particular, will AD fall and will output fall with it? Will the PUP cause a recession?
Here's my answer: no.
There is a market for output, where goods exchange for money. And a market for bonds (loanable funds), where bonds exchange for money. The bond market will not clear. There will be an excess demand for bonds, since the binding pro-usury law keeps the rate of return on bonds above equilibrium and the price of bonds below equilibrium. Savers will want to buy more bonds than dissavers and investors want to sell. The market for output will continue to clear. The price of output may have to adjust up or down, or the central bank may have to adjust the money supply up or down, to make it clear. But it can still clear, given the right level of the real money supply. (That's why I stuck in that proviso about the central bank being able to buy bonds by compulsory purchase if it want to increase the money supply.)
With the bond market failing to clear, desired investment by anyone who needs to borrow to finance that investment will be lower, since the rate of interest is higher. But desired investment by people who would otherwise lend their savings will be higher. Since they cannot buy as many bonds as they want, given the excess demand for bonds. Overall investment will probably be lower, since the savers don't have good direct investment opportunities themselves. That means there will be slower growth in the capital stock, plus malinvestment, so the long run growth rate will be reduced.
But I don't see an excess supply of goods. I don't see a recession. The demand for money may increase a bit, since some people will be unable to buy bonds and will choose to hold money instead rather than consume or do direct investment. So the central bank may need to increase the money supply to match the increased money demand. (Or a lower price level may do the same job of increasing the real money supply).
People will want to buy bonds. But they can't. So they will buy goods instead. Despite the real rate of interest being above the natural rate, there will be no recession.
An excess supply of output is not the flip-side of the excess demand for bonds. An excess supply of output is the flip-side of an excess demand for money. We do not exchange output for bonds; we exchange output for money.
In my view, high real interest rates themselves do not cause a recession. The tight monetary policy, of which high real interest rates may or may not be a symptom, is what causes a recession. This thought-experiment is designed to elucidate this point, by seperating monetary policy from interest rate policy.
"The central bank is also granted the power of compulsory purchase of bonds."
Doesn't your thesis depend critically on that assumption?
Doesn't it fail without it?
Posted by: anon | January 29, 2011 at 08:31 AM
It seems to me that, in this model, the marginal interest rate -- the zero-risk opportunity cost of the marginal unit of capital -- in this model is actually zero, not 10%. So instead of asking why the PUP doesn't cause a recession, you should be asking why it doesn't cause a boom.
Posted by: Andy Harless | January 29, 2011 at 09:11 AM
To answer my own question, here's why it doesn't cause a boom, and why it doesn't refute the neo-Wicksellian model: the central bank provides just enough cash that the marginal value of its liquidity services equals the natural interest rate. So the "true" interest rate -- which is paid in the form of liquidity services -- is exactly the natural interest rate. The only difference from the standard neo-Wicksellian model is that there is no market in which we observe this interest rate explicitly, because there is no interest-bearing asset for which people can freely exchange money and thereby reveal exactly how much they value holding the marginal unit. But people are doing the same calculations: "How much is it worth to me to hold another unit of money?" There is a shadow interest rate that the central bank controls, and if it wished to do so (and I'm guessing it would wish to do so), it could conduct monetary policy by making an estimate of that shadow interest rate and targeting that estimate in the same way that actual central banks target actual interest rates.
Posted by: Andy Harless | January 29, 2011 at 09:35 AM
HA! The Knutster PWNS!!!
Posted by: Andy Harless | January 29, 2011 at 09:42 AM
anon: maybe, maybe not. It's not obvious to me that the demand for money will increase. It might, it might not. Assume it does. If M were fixed, a fall in P could (in principle) increase M/P until it equaled desired M/P. Or the central bank could do a one-time helicopter increase in M.
Andy: which marginal interest rate then matters to ensure that AD=AS. The interest rate on money, or the interest rate on bonds?
Posted by: Nick Rowe | January 29, 2011 at 09:57 AM
I'm especially hoping Bill Woolsey will chime in on this one, especially to respond to Andy's comment, and to respond to my question about which interest rate matters: money or bonds.
Posted by: Nick Rowe | January 29, 2011 at 10:06 AM
Nick:
You need to add another condition to the example. Households save by buying bonds or accumulating money. Firms fund investment by selling bonds.
If you allow households to purchase capital goods and go into business, how about equity finance? All that really happens is that the bond market more or less disappears if you create a sky high floor.
Andy:
I am not sure I understand the shadow interest rate at all, but I think Nick wants the marginal value product of capital (or however we want to describe this) to be 10%. The firms can't fund the investment because of the low quantity of bonds demanded at a 10% interest rate.
If the quantity of investment demanded at 10% requires an interest rate greater than zero to generate an equal quantity of saving supplied, then it works like Nick said. People would like to save alot at the going price controlled market interest rate, but they actually save and amount equal to the quantity of investment demanded by the firms. How much they would save at an interest rate of zero is not relevant. The shadow interest rate would be the value of future consumption relative to present consumption.
If, on the other hand, the interest rate where quantity of saving supplied matches the quantity of investment demanded is less than zero, then after the households have purchased all the high yield bonds provided by the firms, there will still be excess saving. They will accumulate currency.
I think there are zero bound problems. It seems to me that compulsory purchases of bonds would just leave all bonds in the hands of the central bank, an excess demand for money, and saving greater than investment.
Posted by: Bill Woolsey | January 29, 2011 at 10:14 AM
Are you talking about a *real* rate of 10%?
If MPK is fixed to be too high, then you are requiring an excessively high return. Firms will need to shed workers and capital to climb back up the MPK curve.
Irrespective of savings and investment demands, or what the central bank does, there will be unemployment and an output gap.
But if you are talking about a nominal rate of 10%, with a real (i.e. achievable) rate equal to, say, 5%, then you have to hope for inflation, but I'm not sure what the driver of the inflation will be. The central bank can only purchase bonds, and there will be very few bonds sold for 10%, if the achievable return is 5%. If there aren't any bonds, or very few bonds, then the CB can't do anything.
If you have fiscal policy -- the government deficit spends, say effectively sending a check to every firm that doubles its earnings from 5% to 10%, or it sends a check to every worker, subsidizing the employer's cost of hiring, then things might be OK, if the policy was assumed to be permanent, or long lasting enough to not discourage any of the 5% borrowers from committing to pay the 10% rate.
Posted by: RSJ | January 29, 2011 at 01:33 PM
Andy is right, if the CB can still set the price of its interventions, it has full freedom.
Nick: you've confused the coupon with the price. Just because PUP set the rules on the coupon, they have set the rules on price in the secondary market--particularly if they haven't set the rules on the trades the cb makes. See for instance any real bond market or Islamic finance.
You cannot simply decree an exception here without price controls on all equity.
Posted by: Jon | January 29, 2011 at 01:49 PM
What happens is academic. If the purpose is to retrofit a non-self-adjusting system so that it will "self-adjust" with a little help from its friends in the central bank what happens is that the exercise will eventually eat its own tail. If the purpose is to "transform poverty into plenty," the instruments for doing so must continually be re-assessed for their ongoing appropriateness. The appropriate question is a human question about what constitutes "the good life" and not a technical question about how one can make the economic system "self-adjusting".
Posted by: Sandwichman | January 29, 2011 at 02:29 PM
Bill is right, provided that the marginal value of liquidity services is zero.
Specifically, the interest rate on money is the one that matters to ensure that AD=AS.
Take the simple case where direct investment is impossible. In that case (assuming no accelerator effect) the quantity of investment is determined by the interest rate floor. The quantity of consumption has to adjust to make the total of investment and consumption equal AS. The quantity of desired consumption (conditional on income and on how much firms borrow at 10%) will be determined by the marginal return on money, which people will set equal to the marginal rate of substitution between current and future consumption.
Suppose the marginal value of liquidity services is zero. In that case, the interest rate that matters for AD is zero. Whether there will be a recession depends on whether a zero marginal return is sufficiently low to discourage people from saving more than the quantity of investment.
Another question is whether there will be an inflationary boom. Suppose that, before the pro-usury law goes into effect, people are holding certain stocks of money and everything is in equilibrium at a 3% interest rate. Now introduce the new law. If people have positive pre-existing money holdings that they can still spend, they may choose a level of saving that is less than the level of investment that firms choose to do at a 10% interest rate. (They are doing some saving via bonds, and they will choose to continue owning the bonds even if this level of saving is more than they desire, because they have an arbitrage by moving their pre-existing assets from money into bonds.) In that case there will be a boom. To avoid this, the central bank has to restrict the money supply until the marginal value of liquidity services equals the natural interest rate (which is no longer 3%, because of the capital market imperfection, but it's somewhere between 0% and 10%).
I still say, Knut roolz!
Posted by: Andy Harless | January 29, 2011 at 02:54 PM
Nick, the fiscal policy is doing the job in your example. Compulsory purchases by the central bank are just a tax on capital in disguise. Your solution (compulsory purchases) are taken right out of Eggertsson's playbook (http://www.newyorkfed.org/research/staff_reports/sr402.pdf), and they won't persuade anyone that the quasi-monetarist perspective is the correct one.
Compulsory purchases by the central bank create macro equilibrium by raising the natural pre-tax real interest rate until it is equal to the 10% real floor, they also reduce the natural nominal after-tax interest rate to zero until it is equal to the yield on cash.
There is a transition problem. The victory of PUP government is likely to temporarily push the natural nominal after-tax interest rate below zero, this will create a recession.
Long-run macro stability is also questionable in practice. Your argument is equivalent to saying that with the monetary policy stuck at zero, by varying the rate of corporate income tax in the range of 80-120%, the central bank will ensure the macro stability. I think that errors are very likely.
Andy, you have got some nice slogans here.
Posted by: The Money Demand Blog (123) | January 30, 2011 at 05:23 AM
OK. Let's simplify massively:
Assume a representative agent model. Standard Neo-Wicksellian job. Each household has the same consumption/savings preferences, and investment opportunities, so no bonds are actually traded in equilibrium anyway, even before the PUP gains power.
In this case, I would say the PUP raising the interest rate to 10% has no effect whatsoever. There's an excess demand for bonds, but absolutely nothing else happens.
OK Knutsters, what say you?
Posted by: Nick Rowe | January 30, 2011 at 08:43 AM
I'll be back for more after skating!
Posted by: Nick Rowe | January 30, 2011 at 08:47 AM
BTW, yes, any half-awake finance wiz-kid would get around the PUP policy in 5 seconds. Just like with Islamic Finance. But hey, it's a thought-experiment. Assume all the finance guys are true believers in PUP policy. Or, just take the representative agent Neo-Wicksellian model where there aren't any finance guys anyway, because there is no finance in equilibrium. When r goes to 10% everybody wants to be on the buy-side of the loanable funds market, so even the wiz-kids can't do anything, because nobody wants to take the other side of the trade.
Gotcha there Neo-Knutsters?
(Actually, it's a separate question whether Knut himself would go along with this. Maybe, just maybe, he might take my side?)
Damn! Typepad is not posting my comments. It says "your comment has been posted", but nothing appears. Back page, forward page, hit "post" again. Let's see.
Posted by: Nick Rowe | January 30, 2011 at 09:00 AM
I don't see how the new example changes my argument. The relevant interest rate is the one that is actually available, namely the marginal value of holding money (or, when it exists, the interest rate on bonds that you have the opportunity to purchase or sell). Pre-PUP, the marginal value of holding money was 3% (as evidenced by the fact that people were satisfied with their money balances when bonds paid 3%). Post-PUP it is still 3% (as evidenced by the fact that people had been satisfied with the same money balances when they had the opportunity to buy 3% bonds). The question is, "At what interest rate can you lend and borrow?" The answer is, "3%, even if the official interest rate is 10%." You can borrow at 3% by reducing your money balances and foregoing the marginal liquidity services. You can lend at 3% by raising your money balances and taking advantage of the marginal liquidity services.
Whether Knut would have appreciated the liquidity preference issue is another question. I may have to start chanting, "May! Nard! May! Nard! May! Nard!"
Posted by: Andy Harless | January 30, 2011 at 10:24 AM
What was this again with interest rates and investements? Do people really buy less houses and start to build less houses when the interest rates rises to 10%. And do firms really - well, you get the idea.
Just look what happened around 1980, in South America, when Volcker performed his real life eperiment: starving South America to combat USA inflation.
Posted by: Merijn Knibbe | January 30, 2011 at 12:34 PM
Andy:
Representative agent model. Forget investment. Zero savings in equilibrium. Consider two cases:
1. PUP raises r on bonds to 10%, but leaves r on currency at 0% (I.e. the Governor of the Bank wins his argument).
2. PUP raises r on bonds *and currency* to 10%. (i.e. the Governor loses his argument for an exemption).
I would say that 1 will have zero effect on Y. But 2 will cause a big recession.
That means it's the rate of interest on the medium of exchange that matters, not the rate of interest on bonds. It doesn't matter (in this representative agent economy) if r on bonds is fixed above equilibrium. You get an excess demand for bonds, only. It matters a lot if r on the medium of exchange is fixed above equilibrium. You get an excess demand for money, and an excess supply of everything else.
Posted by: Nick Rowe | January 30, 2011 at 03:16 PM
Merijn: "What was this again with interest rates and investements? Do people really buy less houses and start to build less houses when the interest rates rises to 10%."
When *real* (not nominal) interest rates hit 10%? Yes. Absolutely. Remember 1982 (Canada)? That's what happened.
Posted by: Nick Rowe | January 30, 2011 at 03:19 PM
"That means it's the rate of interest on the medium of exchange that matters"
No. It means that it's the marginal rate of interest that matters. If it happens to be the bond market that is not clearing, then marginal financing has to be done with money, so the rate of interest on money is what matters.
Suppose it were the other way around, that the money market didn't clear but the bond market did. For example, suppose, for some technological reason, access to the banking system had to be temporarily suspended, but businesses could still borrow money by issuing paper to suppliers. The interest rate that would matter would be the interest rate on the paper businesses were issuing, not the interest rate that banks were paying on money.
At the extreme, imagine an economy with perfectly frictionless trade credit. In that case, there is no medium of exchange, but their can still be recessions if the interest rate goes to zero, because money will still be a store of value.
Posted by: Andy Harless | January 30, 2011 at 04:25 PM
I don't see how the pro usury policy is enforced.
We do know how little difference some extreme sounding anti-usury policies can make. for example Sharia. Maximum allowed interest rate 0. Perfectly consistent with say a 30 year fixed interest mortgage. House costs X. Buyer has a down payment y and wants a 30 year mortgage with an interest rate of say 5% nominal. Hmm bank adds up interest and principal repayments corresponding to the mortgage with no discounting (discounting is sinful) gets Z > X-y. Raises price on house to y + Z and sells it to buyer for y making a 30 year zero interest loan of Z. No interest paid. No difference. No problem.
Well except in the UK there is a tax on sale of houses which had to be paid twice. So they changed the law so as not to inconvience those who wanted to obey both Her Majesty's laws and Allah's. This legal reform actually happened.
Just the same way, a minimum mortgage interest rate of 10% wouldn't matter. 10% real would be like whatever you want real (but it would have to be real).
Same for corporations. If there is excess demand for bonds, then the corporation will pay its suppliers (and workers I bet) partly in bonds. In exchange for lower prices from suppliers and lower wages.
The aims of the PUP would be frustrated, because they don't know what the authentic price that would be paid in cash is.
On a different issue, I don't see why you discuss the money supply. The analysis seems to be a long run analysis and not the effects on the day after a surprise introduction of the law.
In standard Keynesian and New Keynesian models the long run restrictions on the money supply can't affect unemployment. So why are you so careful to make sure monetary policy insn't hampered ?
Posted by: Robert Waldmann | January 30, 2011 at 04:36 PM
Robert: The PUP policy *can't* be enforced, in practice. They will get around it. But just *assume*. It's only a thought-experiment. It doesn't have to be realistic.
"On a different issue, I don't see why you discuss the money supply. The analysis seems to be a long run analysis and not the effects on the day after a surprise introduction of the law.
In standard Keynesian and New Keynesian models the long run restrictions on the money supply can't affect unemployment. So why are you so careful to make sure monetary policy insn't hampered ?"
Just to avoid getting into arguments about whether price flexibility is sufficient to get to full-employment in NK models. But notice that if we have a representative agent model, I don't even need monetary policy or price flexibility. The PUP policy will have no effect on the outcome, even if both M and P are constant.
Andy: "Suppose it were the other way around, that the money market didn't clear but the bond market did."
There IS no "money market". "Money market" is an oxymoron (or maybe a tautology?). In a monetary exchange economy, *every* market is a money market.
Posted by: Nick Rowe | January 31, 2011 at 07:10 AM
This is frustrating. What I have here is a devastating critique of the Neo-Wicksellian paradigm. And people aren't getting it. Boo hoo!
I really wish I had started out with the representative agent version of the model. It's so much simpler. No investment. Consumption loans only. In equilibrium no bonds are traded. Now pass a law that sets the real interest rate at 10%. Absolutely nothing happens, as long as the rate of interest on the medium of exchange can stay at 0% (or whatever it was before). There's an excess demand for bonds in terms of money. But there is no excess supply of goods in terms of money. There is no fall in AD. There is no recession. Nothing happens. The whole Neo-Wicksellian paradigm is totally wrong. It's incoherent.
Do I need to write a second, way more inflammatory, post?
Posted by: Nick Rowe | January 31, 2011 at 07:27 AM
If I think of Australia, where almost all mortgages are variable rate - then my view of the monetary policy transmission there is, short term interest rates go up, long term rates (including mortgages) follow, reducing after tax income, demand falls. Interest rates aren't affecting the demand for investment goods - they effect primarily disposable income.
Posted by: reason | January 31, 2011 at 09:03 AM
reason: income effects don't really exist:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/12/income-effects.html
Posted by: Nick Rowe | January 31, 2011 at 09:06 AM
A ha! - so you are counting on the increase in Bond Holders income, making up for the decrease in Mortgage holders income. But this comes back to what we talked about a few days ago - you see my guess is that older and richer bond holders don't respond to an increase in income as stronger as younger mortgage holders do to a fall in income. And I reckon the statistics back me up, no matter what Milton Friedman thought 40 years ago.
Posted by: reason | January 31, 2011 at 11:39 AM
oops
should read ... "as strongLY as"... - and of course the vice-versa case is similar. Non-linear multipliers.
Posted by: reason | January 31, 2011 at 11:41 AM
Nick, something which has bothered me in the past with the Swedish/Austrian approach is something Scott likes to remind us of: the interest rate is a price.
When I parse the austrians closely enough, they are not saying that it's the interest rate movements that have the effect but rather the expansion and contraction of credit. It just happens that when the banking system expands credit unnaturally fast, the market rate of interest falls and the real rate falls, and when the banking system contracts real rates rise.
So whats doing the work in their story? The interest rate or the changes in credit supply vs credit demand. You seem to suggest its the quantities that are doing the work in transmitting the effects.
Posted by: Jon | January 31, 2011 at 12:45 PM
Jon: I don't think it's a prices vs quantities issue. It's more of a *which* price (or *which* quantity) issue.
Suppose there's bonds and currency. Currency is a medium of exchange, and bonds aren't.
I am arguing (this is very rough, and not to be taken too literally):
High interest rate on bonds will not cause a recession; a high interest rate on currency could cause a recession.
Or,
Low quantity of bonds cannot cause a recession; a low quantity of money could cause a recession.
Basically: it's money, not bonds. The Neo-Wicksellian fallacy is to say that it is the rate of return on bonds (and the supply and demand for bonds) that equilibrates AD and AS. I say no, it's money.
Posted by: Nick Rowe | January 31, 2011 at 01:01 PM
"Money market" is an oxymoron (or maybe a tautology?). In a monetary exchange economy, *every* market is a money market.
It seems to me your whole argument is a tautology (and more obviously so when it's put forward as a representative agent model without investment). If you insist on there being a medium of exchange involved in every transaction, then any value related to transactions that might actually take place is going to be meaningful only in terms of the medium of exchange. You can't have a meaningful interest rate that isn't an interest rate on money or on something you can buy with money. Therefore, subject to your implicit assumption that every transaction involves the medium of exchange any statement that a neo-Wicksellian makes about interest rates is also implicitly a statement about money.
In my last comment, I tried to get rid of your implicit assumption by looking at an economy where purchases are made with trade credit -- essentially intertemporal barter (using the unit of account and the interest rate to define the terms of trade and leaving one side of the barter to be determined later) -- instead of a medium of exchange. In such an economy (assuming sufficient price rigidity) the interest rate will be determined (in the short run) by central bank policy. If the central bank sets the interest rate too high, there will be a recession, even though there is no medium of exchange.
Posted by: Andy Harless | January 31, 2011 at 01:20 PM
Andy: (thanks for commenting BTW, and for coming to grips with my argument):
If there is trade credit, *and if offsetting trade credits can be cancelled in some sort of multi-party clearing house (A owes B$100, B owes C $100, and C owes A $100, and the clearing house wipes out the lot)* then I would say that trade credit *is* a medium of exchange. It's not just postponing payment, it *is* payment. Trade credit is money, and then the interest rate on trade credit is the interest rate earned by holding money.
I want to be *explicit* about my assumption that every transaction involves the medium of exchange (call it currency). I assume: you buy goods for currency and sell goods for currency. You buy bonds with currency and sell bonds with currency.
There's the interest rate on bonds, and the interest rate on currency. When Neo-Wicksellians say that "the" rate of interest needs to adjust to equilibrate AD and AS they *think* they are talking about the interest rate on bonds, not currency. But I think they would only make sense if they were talking about the interest rate on currency instead (for a given supply and demand for currency).
Posted by: Nick Rowe | January 31, 2011 at 02:00 PM
The interest rate that matters is the marginal rate of substitution between current and future consumption, which is the interest rate on all freely tradable assets. In a free market, it's easier to think of it as the interest rate on bonds. (It's also easier to observe, because you don't have to estimate the marginal value of liquidity services.) In your example, bonds aren't freely tradable. The MOE is freely tradable by definition, so it always works, but I don't think that gives it conceptual priority.
Posted by: Andy Harless | January 31, 2011 at 08:25 PM
Which is the only interest-rate in a Wicksellian model. There cannot be two. Andy seems to see a set of substitutable claims on the future--bonds or money. Money is supplied elastically and the CB makes up the difference; picking an inflation rate r such that -r, the real return on money equals the natural rate.
Nick's narrative doesn't really explain how the PUP law interacts with nominal contracts. Suppose, you cannot contract for a future payment? That's not enough either though. You basically need a world without capital goods and without storage.
Otherwise whether you have money or not, the real-rate will keep reemerging.
Posted by: Jon | January 31, 2011 at 11:55 PM
Jon:
Think of a price floor for air fares.
The gap between the demand curve
and the supply curve at that price is often described as waste.
It is like rent seeking. Firms will spend money on adversiting, or maybe enhance quality, more roomy seats, better meals, etc.
OK. And so, that sellers offer price for the quantity demanded given the floor "shows up"
I think you are trying to say that the equivalent is the true interest rate.
Well, I guess.
Posted by: Bill Woolsey | February 03, 2011 at 07:44 AM