What is "the natural rate of interest"?

"The natural rate of interest" is a theoretical construct. It is a theoretical construct that only has a defined meaning within a certain class of economic models. And even within that class of models, the exact definition may vary from one model to the next.

The class of economic models I am talking about can be called "Natural Rate Models". To see if a model is or is not a NRM, perform the following steps:

1. (Monetary neutrality). Look at all the equations of the model (with one important exception I will talk about later). Look to see if the equation contains any nominal variables. (A nominal variable is one with $ in the units, like the price level, any money price, or nominal stock of money). Double all the nominal variables (or halve, or treble, or whatever), and leave the other variables (the real variables) fixed. Is the equation still true? If your answer is "no", for any equation, the model is not a NRM.

2. (Monetary super-neutrality). Repeat step 1, only this time increase the (percentage) rate of change of all nominal variables by one percentage point (or by minus one, or plus two or whatever). (The rate of inflation, both actual and expected, is an example of the rate of change of a nominal variable. The nominal rate of interest also counts as a rate of change of a nominal variable, since it measures the rate of change over time of how many $ you owe if you borrowed $100.) Hold real variables and the levels of nominal variables constant. Is the equation still true? If your answer is "no", for any equation, the model is not a NRM.

[3. Only if you are being really pernickety: repeat step 2, only this time slipping another time-derivative, so you are looking at rates of change of rates of change of nominal variables.]

4. (Uniqueness). Do the equations of the model specify a unique solution to the real variables? If not, the model is not, strictly, a NRM. (Or, you might prefer to say that it defines multiple natural rates.)

A NRM exhibits "monetary neutrality". An equi-proportionate change in all nominal variables leaves the equilibrium values of all real variables unchanged. (Notice I have said nothing about whether the nominal stock of money is exogenous or endogenous, so I am immune to any critic who disputes the statement that " A doubling in the stock of money causes all nominal variables to double and all real variables to stay the same." by saying that money is endogenous). It also exhibits "monetary super-neutrality". An equi-proportionate change in the rate of change of all nominal variables leaves the equilibrium values of all real variables unchanged.

In a Natural Rate Model, it is useful to call those equilibrium values of the real variables "natural rates (or levels)". There is a natural rate of unemployment (if the model contains an unemployment rate), and a natural rate of (real) interest (if the model contains a rate of interest). There is also a natural rate (or level) of output, employment, real wage, quantity of butter produced, relative price of apples to bananas, etc., for any real variable that exists in the model.

But the precise definitions of those "natural rates" within that model will depend on the particular specifics of the model: on what forces actually determine the equilibrium values of those real variables within that model. If the model incorporates an efficiency wage theory of unemployment, for example, the "natural rate of unemployment" will be determined very differently from a model that incorporated a search-theoretic theory of unemployment.

Similarly, a model in which the equilibrium equilibrium real rate of interest was determined by the marginal product of capital would lead to one "definition" of the natural rate of interest. A model with only consumption loans, where the equilibrium real rate of interest was determined by the rate of time preference, would lead to a very different "definition". But these "definitions" aren't really definitions. They are model-specific statements about what determines the natural rate of interest. And a criticism of a particular model's theory of what determines the natural rate of interest should not be confused with a criticism of the theoretical concept of "the natural rate of interest" itself.

If you have a model that is not a Natural Rate Model, that model does not define a natural rate of unemployment, or natural rate of interest. The concept of a "natural rate" just isn't useful in a model with significant monetary non-neutrality, or non-super-neutrality. Notice how I stuck in the weasel word "significant" into the preceding sentence? That's because few models are strictly natural rate models. For example, if currency does not pay interest, then an increase in the rate of inflation will generally lead to a lower equilibrium value for the real stock of currency. It is conceivable that a higher inflation tax on holding currency, and the lower real stock of currency, could also affect the equilibrium real rate of interest, rate of unemployment, and every other real variable. If these effects were deemed to be empirically important, we could not usefully talk about natural rates of interest, unemployment, or any other real variable. But economists are sensibly schizophrenic on this question. We typically start with a benchmark model that has monetary neutrality and super-neutrality, so that natural rates are well-defined. Then we consider second-order non-super-neutralities as deviations from that benchmark. It helps us keep our heads clear.

Now, remember above, when I said I would talk later about "one important exception" when we were checking the equations for neutrality and super-neutrality in steps 1 and 2? Now's the time to talk about it.

Suppose that all equations satisfied neutrality in step 1. The system of equations would then determine all real variables, but would leave all nominal variables indeterminate. Someone's got to care about the level of nominal variables, otherwise the price level is indeterminate. Any price level would be an equilibrium. Similarly, if all equations satisfy super-neutrality in step 2, the equilibrium rate of inflation would be indeterminate. Any inflation rate would be an equilibrium.

Somebody in the model has to suffer from money-illusion, or inflation-illusion, if the model is to determine the price level, or rate of inflation. Who is that somebody?

The usual answer is that since no other rational agent would want to take on the job, the central bank must be the one. Think of it as a public good problem, if you like. Every individual wants a determinate price level, and determinate rate of inflation, but no individual is willing to suffer the costs of behaving irrationally in order to create that public good. So the central bank (or whoever creates the money) has to do the job.

The way to save a Natural Rate Model from price level or inflation rate indeterminacy is to make the equation describing the central bank's behaviour violate what we were checking for in steps 1 or 2 above.

If the central bank wants to target the price level, it must do something real (or be prepared to let something real happen) if ever the price level rises above or falls below its target. Or if it wants to target the inflation rate, it must do something real (or be prepared to let something real happen) if ever the inflation rate rises above or falls below its target. And if it wants the equilibrium price level or inflation rate to be not only unique, but also a stable equilibrium, it must also make sure that real thing moves (by moving it, or letting it move) in the right direction.

That "real thing" that the central bank moves, or lets move, could in principle be anything whatever. But if you think of monetary policy, and the monetary policy transmission mechanism, in terms of interest rates (something I would question, but not here), then the central bank has to raise the real rate of interest above the natural rate if it wants to reduce aggregate demand and lower the price level or inflation rate. And lower the real interest rate below the natural rate if it wants to increase aggregate demand and raise the price level or inflation rate.

[By the way, this is just Don Patinkin's "Money, Interest, and Prices" with an extra time-derivative thrown in, and with money either exogenous or endogenous.]

Comments

[Tom Hickey]

Bill Mitchell addresses issues surrounding this in his most recent blog post, albeit from somewhat of a different angle.

http://bilbo.economicoutlook.net/blog/?p=6617&cpage=1#comment-2341

[bill woolsey]

Nick:

I just don't look at things this way at all. The natural interest rate is abour models? I think it is about the world.

Superneutrality? I think it is possible that inflation could impact the natural interest rate. It has to do with taxes and expenditures, particularly the inflation tax.

Perhaps I just don't understand, but I see the natural interest rate as being what the interest rate would be in the absense of monetary disequilibrium. I also have some notion of the real interest rate playing a coordinating role in intertemporal production and consumption.

Do prices and inflation clear up imbalances between the quantity of money and the demand to hold it? Do prices and inflation have no impact on the allocation of resources and the productive capacity of the economy? Seem like too different issues to me.

[Nick Rowe]

Bill:

Suppose you believed in an upward sloping LRAS curve. (Or a downward-sloping Long Run Phillips Curve).

I suppose you *could* say that the natural rate of unemployment depended on the price level (or the rate of inflation). But you would be using that concept in a very new way, that contradicts previous usage. I would rather say that the concept would not be useful in that sort of model. Instead you would just say that the long run equilibrium level of unemployment depends on the price level (or inflation rate), and not talk about the "natural" rate of unemployment at all.

I think you could make the same argument about the natural rate of interest. If you have a non-vertical LRAS curve (or LRPC curve), then, in general, the long run equilibrium real rate of interest will depend on the price level (or inflation rate).

Wicksell attempted to define the natural rate in some sort of model-specific way. That was a mistake. His attempts failed. He needed to first speak about neutrality. Friedman introduced the natural rate of unemployment (and re-introduced the natural rate of interest), by talking first about neutrality and super-neutrality. Only after having cleared the ground for the idea of the natural rate, did he go astray and attempt to define the natural rate as "that level that would be ground out if all the frictions blah blah were embedded into a Walrasian blah blah..". An attempted definition that most people have sensibly ignored, having captured the gist of what he was saying by his references to neutrality and super-neutrality.

"Do prices and inflation clear up imbalances between the quantity of money and the demand to hold it? Do prices and inflation have no impact on the allocation of resources and the productive capacity of the economy? Seem like too different issues to me."

Those are two different issues. But if you answer the second question in a particular way, you soon find that "natural rates" are not useful simplifications.

There has always been this tension in mainstream economics. We believe in super-neutrality, and yet think inflation is a bad thing. We can't have both. I think of super-neutrality as a simplified benchmark, and use it as a jumping-off point to talk about how inflation matters, because it creates bad non-super-neutralities.

Tom: thanks. i will take a look.

[Nick Rowe]

Tom: I briefly skimmed Billy's post. Not directly relevant to this one, I think.

[PE]

Natural rate of interest? I thought it had been known since at least Hugh Townsend's 1937 EJ article that there is no such thing. The interest rate is driven by psychological and institutional factors.

[Nick Rowe]

PE: Of course the natural rate of interest is determined (in part) by psychological and institutional factors! The rate of time preference, for example, is a psychological factor. The existence of credit markets, and the laws governing them, for example, is an institutional factor. (And you mean people didn't know that until 1937? What about Irving Fisher for example?)

What's that got to do with its existence or non-existence?

Only if you define the natural rate of interest as (say) determined by technology would that be a problem. But the whole point of my post was to explain why it's wrong to define the natural rate of interest in any model-specific way.

[JP Koning]

Nick, thanks for this useful post. Nominal/real/neutral always gets me.

One question. You say that "every individual wants a determinate price level"... but given a natural rate model, why would people want a determinate price to begin with? If individuals in the model don't suffer from money illusion and nominal factors don't influence real factor, seems that price could rise or fall to any amount and it wouldn't be important to them.

Also, if everyone does want a determinate price level, wouldn't some private actor (rationally pursuing profit rather than irrationally suffering from money illusion) have the incentive to provide this service?

[Too Much Fed]

What is "the natural rate of interest"?

As of right now, I am going to go with MOSTLY it varies based on productivity, whether an economy is demand or supply constrained, and the composition of the fungible money supply (currency vs. currency denominated debt).

[Nick Rowe]

JP: "One question. You say that "every individual wants a determinate price level"... but given a natural rate model, why would people want a determinate price to begin with? If individuals in the model don't suffer from money illusion and nominal factors don't influence real factor, seems that price could rise or fall to any amount and it wouldn't be important to them."

Yep. That's the paradox I was referring to in my response to Bill. If economists think inflation is a bad thing, or that there's some optimal level of inflation, then we can't literally believe in super-neutrality. And we don't.

"Also, if everyone does want a determinate price level, wouldn't some private actor (rationally pursuing profit rather than irrationally suffering from money illusion) have the incentive to provide this service?"

But how could he get people to pay for that service? The only person I can think of who might profit from providing this service is the person who issues the money. If that money has a stable predictable value, people will be more willing to use it.

Too much Fed: The *overall* *level* of productivity (as opposed to the productivity of investment) probably doesn't affect the natural rate. I can't think of any theoretical reason why it should. Plus, empirically, there have been very big increases in productivity over the last century, but little change in real interest rates. But the *growth rate* of overall productivity probably does affect the natural rate.

[Doc merlin]

"Somebody in the model has to suffer from money-illusion, or inflation-illusion, if the model is to determine the price level, or rate of inflation. Who is that somebody?"

We all do. We all have debt denominated in currency value, and marginal taxes denominated in currency, and contracts denominated in currency. There is no illusion. Because of contracts, law, and debt, fixed nominal variables exist. This breaks super-neutrality.
Is that what you meant, when you say, "use it as a jumping-off point to talk about how inflation matters"? Makes sense.

The next statement, unless I am mistaken in what you mean, I disagree with entirely!
"The usual answer is that since no other rational agent would want to take on the job, the central bank must be the one. Think of it as a public good problem, if you like. Every individual wants a determinate price level, and determinate rate of inflation, but no individual is willing to suffer the costs of behaving irrationally in order to create that public good."

I disagree with this on so many levels.
1. The idea that there is a single determinate rate of inflation is completely false. There are many, many rates of inflation. Every single item in the economy has its own rate of inflation. The Fed just aggregates it, massages it, and publishes the statistic. There is no actual single rate of inflation except as a shorthand to make it easier to do macro.
Thinking of inflation rate as a single number, while it makes it easier to do macro, completely obscures huge shifts in the economy. For example: during the 00's housing and commodity prices skyrocketed, but CPI and PPI stayed relatively tame. We had huge inflation in the country, but the price drop in technological goods, obscured this.

2. There is no determinate price level. How someone chooses to buy a good, which good they buy, etc all determine the price level for them. Also, due to price discrimination, different individuals have different prices for the same goods.

[Adam P]

Guys, this statement:

"Somebody in the model has to suffer from money-illusion, or inflation-illusion, if the model is to determine the price level, or rate of inflation."

is patently false anyway so no need to debate it. Please.

[Nick Rowe]

Adam P. ! I'm shocked!

Write down a central bank reaction function, that conforms to the Taylor Principle, for example. I would say that that central bank suffers from "money illusion". That equation does not exhibit "monetary super neutrality". (And a good thing too, I might add.)

Sure, I'm pushing the metaphor when I say that that central bank "suffers from money illusion". But if you think about it abstractly, and think of it as representing the behaviour of some ordinary person or firm, that's exactly what it is.

[Nick Rowe]

Strictly, I meant "inflation illusion", rather than "money illusion" in the above.

Write down a reaction function for a price level targeting central bank, and you get money illusion. It's not HD0 in real variables.

[Nick Rowe]

Doc: "Is that what you meant, when you say, "use it as a jumping-off point to talk about how inflation matters"? Makes sense."

Yes, that's exactly what I meant. And other stuff like that.

On the rest of what you say though, I disagree. When you have a model with monetary neutrality, you get the classical dichotomy. All the real things, like relative prices, are determined by real forces. So if we know one nominal price, and it can be any nominal price, then that, plus all those relative prices, determines all the other nominal prices and all possible price indices.

[bill woolsey]

Nick:

Suppose I claimed that the long run phillips curve had a positive slope. Or, more plausibly, long run aggregate supply is negatively relatived to the inflation rate.

I don't have any problem describing this as claiming that inflation has an adverse impact on the natural rate of unemployment or potential income. (I am not not making any claim one way or another about whether this is true.)

What I see as the key issue is the creationg of an excess supply of money inorder to expand nominal expenditure. Does an increase, more rapid growth, or increasing rates of growth, and on an on, for nominal expenditure result in higher (or more rapid growth) in real demand, and so higher production or employment due to strong sales?

If the price system will adjust in such a way that this doesn't work, then the natural rate approaches are as "right" as they need to be. But to claim that no matter what the monetary institution, that all real values must be the exact same--that seems wrong.

I know I don't believe that rapid deflation can be managed in a way that avoids disruption because of ero interest currency.

[Adam P]

Nick, the initial price level is pinned down by the government budget constraint/valuation equation. The CB reaction function then determines inflation going forward from that initial starting point.

Neither neutrality nor super-neutrality is necessary for the existence of a natural rate (except perhaps in the very (asymptoticaly) long run).

[RebelEconomist]

I think I agree with Bill. A natural rate should be some kind of constant - eg "rate of mercantile profit" - that is a vital (if summary) property of the economy. You (Nick) are defining it according to its implications for mathematical economic models. The ideas of neutrality and superneutrality etc are academic (in both the literal and the derogatory sense) - because central banks do not simply helicopter drop money, but rather withdraw something else from circulation in return (eg debt, gold etc), monetary expansion should not be expected to lead to an equiproportional increase in prices. In other words, after a one-off monetary expansion, there is more of money, less of one item, and the same amount of all other items.

This relates to the point I made earlier on the money, banks, loans, reserves, capital and loan officers post. Any attempt to deduce what the effect of monetary policy is on the economy should take into account the generally restrained supply of the operational item (usually, but not necessarily, an asset). The point I made there was essentially that these restraints mean that it is not necessarily true that setting an interest rate below the "natural rate" implies an indeterminate price level (although I do concede that there is a key difference between setting the price of some abitrary item such as carrots and short term debt with a nominal return).

[Adam P]

Rebel, as I explained on the other thread, this: "setting an interest rate below the "natural rate" implies an indeterminate price level" is NOT what anybody said. Price level determinacy or indeterminacy is a function of the entire CB reaction function, not the value of the real rate at any particular point in time.

The thought experiment on the other thread was an example where the CB tried to maintian a fixed nominal interest rate forever. That is an example of a CB reaction function that implies an indetermintate price level. Other CB reaction functions, where the nominal rate is not fixed for all time, can determine the price level (say one that satisfies the Taylor principle).

[RebelEconomist]

I do not disagree with the idea that anticipated central bank responses matter Adam, but in trying to understand how the system works, a simple step change sustained indefinitely seems a reasonable input to consider. Now if you are saying that a sustained attempt by the central bank to hold the nominal interest rate below the natural rate implies an indeterminate (to be more precise, Nick said "ever-accelerating") price level, then my verdict is "not proven".

[RebelEconomist]

In case others wish to chip in to the debate without going back to the earlier post, I will briefly explain the issue as I see it. The idea is that setting a nominal interest rate below the natural rate stimulates aggregate demand, which given sticky prices will set off an increase in the price level that will reduce the real interest rate even more below the natural rate. This then stimulates aggregate demand further, leading to another round of increase and so on. However, while I can see that the number of rounds may be infinite, I question whether this necessarily means that the price level is indeterminate. Because the sources of aggregate demand - eg investment projects that yield more than the real interest rate - are in restricted supply, the increase in aggregate demand might be small, meaning that the implied price level increase and hence the further decrease in the real interest rate are small, and so on. It seems to me that if each round is less and less, the process might converge, in which case the price level would be determinate. But, as I said, the point may well be of mainly theoretical interest, because even if determinate, a large increase in the price level might be unacceptable (hence the importance of the CB reaction function as Adam says).

[westslope]

That is what I like about the rational expectations-augmented Phillips Curve model. Interest rates are absent.

Now for understanding money markets, bond markets and similar, bring on the interest rates but for those markets, we don't really need a natural rate concept, do we now?

BTW, are Bernanke and colleagues still thinking in 'natural rate' terms these days or do they simply say those things in order to better communicate with the public?

[Adam P]

westlope: this comment:

"That is what I like about the rational expectations-augmented Phillips Curve model. Interest rates are absent."

only communicates the fact that you don't understand the model.

[kevin quinn]

Nick: Friedman surely thought he had a natural rate model despite super-NON-neutrality, since changes in the the rate of growth of the money supply reduce the level of real money balances held, for Friedman. What am I missing?

[westslope]

westlope: this comment:

"That is what I like about the rational expectations-augmented Phillips Curve model. Interest rates are absent."

only communicates the fact that you don't understand the model. -Adam P

Well I was thinking that it means my imagination fills in so-called 'missing details' differently from yours. You can hook the EA-PC model up to a natural rate model if that is what you want to do. Not necessary.

[Nick Rowe]

Trying to think of an intelligent response to all these comments...

In some sense, there's no real point in arguing about definitions, or the meanings of theoretical concepts. Either you find a particular meaning/definition useful, or you don't.

Let me leave one parting shot:

1. If you believe that there exists a (long run) equilibrium rate of (real) interest/unemployment that is, in some sense, (at least approximately) independent of (some aspects of) monetary policy, then it seems to make sense to refer to it as a *natural* rate.

2. If you believe the long run equilibrium always depends on monetary policy, then why not just call it "the long run equilibrium"? Why bother adding the word "natural"? Would you talk about a "natural rate of inflation", for example? To my mind, people don't talk about the "natural rate of inflation", because it's an oxymoron.

Bill: Most economists (unlike you and me) don't believe that "monetary disequilibrium" lasts for longer than a couple of hours at most. "Interest rates adjust almost instantly to get the demand and supply of money into equilibrium; and in any case, the quantity of money is demand-determined". Can't those people talk about the central bank setting an interest rate different from the natural rate?

Adam: "Nick, the initial price level is pinned down by the government budget constraint/valuation equation."

I would say it is pinned down (predetermined) by price-stickiness. Even if you accept the Fiscal theory of the Price Level, a two-for-one "stock split" would double the price level. The unwillingness of the CB/government to do arbitrary stock-splits is an example of a nominal anchor. Otherwise, if the price level suddenly doubled, the CB/government would validate that doubling of the price level by doubling the money supply in a 2-for-1 split. So the price level would be indeterminate. And if the money supply function were HD1 in nominal variables, that is exactly what it would do.

(You could argue that the unwillingness of firms to do arbitrary stock splits is their equivalent form of "money illusion", and provides a nominal anchor to the value of their shares.)

"The CB reaction function then determines inflation going forward from that initial starting point."

Yes, but if that reaction function obeys the Taylor Principle, then adding 1 percentage point to the inflation rate would add more than 1 percentage point to the nominal interest rate, which violates super-neutrality. As indeed it needs to, if the inflation rate is to be determinate.

Rebel: I am trying to imagine a model in which the long run nominal interest rate would adjust to whatever nominal interest rate the central bank set. And I can't, without assuming prices are fixed forever, or that expected inflation lags permanently behind actual inflation. Actually, I probably could fake up such a model, if the long run real equilibrium were not unique. Give me two vertical LRAS curves (or probably 3, because you normally get one unstable equilibrium between two locally stable equilibria), and I could do it. If you allowed me totally free play with all the assumptions. Or maybe an IS curve that bends back in a reverse-S shape, and intersects full employment 3 times. That would do it.

kevin: I don't think you are missing much (as usual). My interpretation is that the effect of money growth and inflation on equilibrium M/P was a relatively "contained" non-super-neutrality, that did not spillover (much) onto other real variables. Miguel Sidrauski did indeed formally construct a model in which inflation reduced M/P but did not affect the real interest rate or savings, investment or anything else (thereby destroying a whole 1960's literature on "Money and Growth", which nobody has ever heard of since).

[Too Much Fed]

Nick said: "But the *growth rate* of overall productivity probably does affect the natural rate."

Care to expand on that one?

[Adam P]

Nick, price level indeterminacy is about multiplicity of equilibria. Yes, in the non-Ricardian regime a two-for-one "stock split" would double the price level but in a way that was determinate, there would be only one equilibrium price level. Even in this case though, you still need the CB reaction function to have a determinate equilibrium. You need both elements.

In a world where the price level was indeterminate it would still exist as some finite number (possibly exploding as we go through time). Price level indeterminacy just means that even if the economy was exactly described by some set of equations and you somehow, godlike, knew all the paramter values and could perfectly observe the values of endogenous variables and exogenous shocks, you STILL couldn't predict a priori what the price level would be, but it would be something.

[Adam P]

Nick, just to be extra clear, what allows the fiscal theory to UNIQUELY determine the price level is the acceptance of a non-Ricardian fiscal regime (I say acceptance not assumption).

[RebelEconomist]

Nick,

Not knowing how you would demonstrate the indeterminacy of prices using an ISLM type model, I find it hard to discuss this issue in your terms, but I can see that the IS curve would be backward bending, since the lower (real) interest rates go, the less investment would increase. In Y, real r space, I guess that interest rate targeting means that the LM curve would be backward sloping (ie prices are increasing when Y rises above potential, and the nominal interest rate is fixed). Not sure where to go from there though! I do think that it is worth considering how such detail is modelled if you want the models to be relevant, because my point relates to QE - ie at least part of the aim of QE is to lower the interest rates on the types of debt purchased by the central bank, which presumably requires some restriction of (debt) supply.

[Too Much Fed]

Any comments about this?

http://krugman.blogs.nytimes.com/2009/12/14/a-new-paradox/

"Gauti Eggertsson is in the process of presenting a new paper on fiscal policy; the paper is here.

http://www.newyorkfed.org/research/staff_reports/sr402.pdf

In his presentation — though not in the paper — he offers great phrase: the “paradox of toil.”

According to his paper, when you’re in the liquidity trap, certain kinds of tax cuts have perverse effects. Cutting taxes on capital income, for example, encourages more saving — which is a bad thing, because we’re suffering from the paradox of thrift. In fact, reduced taxes on capital income actually end up reducing investment.

So what’s the paradox of toil? If you cut taxes on labor income, this expands labor supply — which puts downward pressure on wages and leads to expectations of deflation, which increases the real interest rate, which leads to lower output and employment.

All of this only applies in a situation of zero interest rates, which wouldn’t be interesting except that that’s the situation we’re in.

The general point is that we’re really through the looking glass, in a world in which lots of things have perverse effects — and basing your policy ideas on intuition from “normal” times can lead you very much astray.

PS: Right at the beginning, Gauti made the point that empirical results from periods in which interest rates are not zero tell you little about this situation — which is why most of what Barro and others, including Greg Mankiw, have been saying is besides the point."

I think it is similar to what I was trying to say here.

http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/08/economic-policy-advice-for-the-ndp-part-iv-corporate-income-taxes.html

[Brian Woods]

"Just as the Prices of things are fixed in the altercations of the Market by the quantity of things offered for sale in proportion to the quantity of money offered for them, or, what comes to the same thing, by the proportionate number of Sellers and Buyers, so in the same way the Interest of Money in a State is settled by the proportionate number of Lenders and Borrowers" Richard Cantillon (1725?)*: essai sur la nature au commerce en general.

Quite simple really. Do you need a wordy theory at all?

B Peter

* Ireland's and Europe's Foremost Economist.

[Nick Rowe]

Brian: that's a good, simple, early statement of the loanable funds theory of the rate of interest. But it's not a theory of the *natural* rate of interest. It does not define a rate of interest that is independent of monetary policy.