In the olden days there was a debate between the Old Monetarists and the Old Keynesians.
The Old Keynesians said that the Old Monetarists' MV=PY was useless, because (desired) velocity was not constant.
Milton Friedman responded with what philosophers call the "Y tu mama tambien" argument. The Old Keynesian multiplier wasn't constant either.
In 1963 Friedman and David Meiselman wrote a paper comparing the relative stability of velocity and the multiplier. I can't find that paper online, and I can't remember who won that debate. But it really doesn't matter now anyway, because most people have forgotten the Old Keynesian multiplier completely. And even those who haven't forgotten it wouldn't defend its stability. Keynesians spend their time instead on a last ditch defence — arguing that, under some circumstances, the multiplier exceeds zero.
The Old Keynesians have been replaced by New Keynesians. And the Old Monetarists have been replaced by….umm…mostly New Keynesians too. (Us "quasi monetarists" are too few and too lacking in influence to really count). So the old debate is now moot. But velocity isn't totally dead as a concept. At least, it's not quite dead enough that a New Keynesian (like Matt Rognlie) won't take a swing at it.
So now I'm going to respond with "Y tu mama tambien!".
Very very crudely, and over-simplifying massively, New Keynesians replace PY=MV with delta(PY)=(natural rate of interest – actual rate of interest). If the central bank sets the actual rate of interest below (above) the natural rate of interest, then nominal income will rise (fall). (Yes, yes, I know that's not really at all precisely accurate, but I'm throwing you New Keynesians a bone here, and you ought to take it, because if you want to be more precisely accurate you will only make it more complicated and dig yourselves in deeper.)
So. Neither is constant. Which is more stable? Velocity, or the natural rate of interest?
At least velocity never goes negative, which is more than you would say yourselves about your natural rate of interest. And your mother wears army boots!
Actually, I'm not at all sure how one ought define "stability" in this case. It doesn't mean "stability" in the normal sense — in the sense of "tending to return to the original equilibrium following a temporary shock". It means something like "it doesn't move much, except perhaps in a predictable way so we can forecast it so monetary policy can respond appropriately".
But, do we ever see a real live policy-oriented Taylor Rule-like beast which does more than just treat the natural rate as some unobserved and unexplained exogenous constant? At least monetarists talk about the effect of interest rates on velocity, and try to estimate a velocity (or demand for money) function.
I don't have answers to the question "which is more stable: velocity; or the natural rate of interest?". I'm just going to leave it at that.
Y tu natural rate of interest tambien.
Worthwhile Canadian Initiative 
Oh, this. Wondered when we’d get around to it.
Very simply, they are both right. Keynes was right in that there is no “natural” rate of interest, and Hayek and Friends were right that there is.
How? Simple. They were both right under different circumstances.
The economy can be characterized by two frontiers: The Production Possibilities Frontier, which describes the capacity of the economy with given supplies, and the Money Supply Limit (my own concept, call it the Money Supply).
The Money Supply is different from the PPF I am a good Rowian and we live in a monetary economy.
With these two limits, the economy has two states it can occupy: PPF < Money Supply, a capacity limited economy, or PPF > Money Supply, a money-limited economy.
If in the first state, consumption and investment can converge to a definite value to define the natural, equilibrium rate of interest. Such a rate will grow the economy. Savings = Investment and all that, and the process works. This is Hayek’s world.
On the other hand, we can also have a money-limited world, the second state. Under this state the equilibrium point is beyond the money supply frontier. The natural rate cannot be achieved with the present state of the money supply. Under this scenario the Central Bank can set any rate it wants within the MSF but it won’t do any good. This is really what Keynes was getting at with his famous only graph. But this state is the depressed economy state. The world was in this state in the 1930’s and is back in it today. We were out of it for most of the from 1945 until 2008.
This is why Keynesian Economics and the Multiplier seemed so wonky, the state changed. Keynesianism though shows its Depression heritage by relying on a money-limited economy.
Determinant’s Grand Theory says that with a dual-mode economy both are right. It takes a paradigm shift to make a switch. 2008 was that switch.
Only drunks and crank addicts are Keynesians.
I´m keynesian. Why? I don´t know exactly. I like Minsky. But I´m not believer in fiscal policy.
I´m believer in monetary disequilibria. Markets are not sufficient. I´m not drunk neither crank. Bur I like good wine.
Great post. BTW, I’d much prefer a model that explained PY, to one that explains deltaPY. Because if you can explain the former, you can explain the latter, but not vice versa.
Scott, NK models do in fact determine P.
I know it is sometimes mistakenly written on blogs, even this one, that the models determine inflation but don’t determine the level of P. It’s not true.
Isn’t this all mute. The problem is with ‘M’. It isn’t well defined and it isn’t under the direct control of the central bank.
reason: “The problem is with ‘M’. It isn’t well defined and it isn’t under the direct control of the central bank.”
Y tu ‘r’ tambien!
The rate of interest under the (almost) direct control of the Bank of Canada is the overnight rate. But very few people and firms borrow and lend at the overnight rate. The overnight rate is not the same as the rate of interest that affects savings and investment decisions. There’s a whole spectrum of rates of interest. When New Keynesians talk about “the” rate of interest that investment and consumption depend on, how exactly do they define it? And that rate of interest certainly isn’t under the direct control of the Bank of Canada.
Kevin510k
I wonder what colours monetarists wear when they go the football. It sounds you think Keynesians and Monetarists are tribes and you want to wave your colours.
P.S. I don’t think I belong to either tribe, I think distribution matters and aggregate thinking misses the point.
i.e. The how matters as much as the what.
But Nick, you know perfectly well what the answer to that one is – the overnight rate effects money market rates which through arbitrage also effect long rates. But yes there is no single “interest” rate, and there a whole lot of possible policy configurations, and I think things that both mainstream “Keynesians” and mainstream “Monetarists” may think of as equivalent, are not necessarily so.
Isn’t it also true that (not thinking just of NK models) that lots of Keynesians would say that it is the shape of the yield curve that determines what monetary policy is doing, not the comparison of actual policy rates to some hypothetical neutral rate.
“Isn’t it also true that (not thinking just of NK models) that lots of Keynesians would say that it is the shape of the yield curve that determines what monetary policy is doing”
Yes.
“not the comparison of actual policy rates to some hypothetical neutral rate”
No.
We all understand full well that risk premia and expectations of the path of policy matter a lot, perhaps far more than the current policy rate, yet it all matters in the context of comparing risk-adjusted interest rates and the expected path of interest rates with the path of the natural rate. Otherwise nobody could ever say if an interest rate of 20%, 300% or .03% was high or low. Depending on the natural rate 300% can be a low interest rate and .03% can be a high rate.
reason, Adam: reason’s comment made me write a new post. In hindsight, I really should have written just one post, covering both posts. Maybe we should continue this discussion on the new post. If you want to repeat your comments there, go ahead. Just copy and paste.
I’m already there mate.
Adam P, I’m glad to hear the New Keynesian model can explain P (or PY). I had relied on Nick’s assertion that it cannot, but perhaps misunderstood Nick.
Scott: Let’s put it this way. For every equilibrium time-path P(t), from t=the distant past to the distant future, in the NK model, there is another equilibrium time-path kP(t), for any k. But given the actual past history of P(t), there is (more or less) only one equilibrium future time-path of P(t).
Nick, that’s not true. The cashless NK model pins down P every bit as much as the QTM, and in much the same way.
The way it works is that the cashless model still has bonds, nominal bonds.
The QTM determines P by combining a nominal supply of M with a demand for real balances. P adjusts to make the real value of the money supply equal to the real amount demanded.
In the cashless model we have a nominal supply of bonds and the government budget constraint. The budget constraint determines the real value of the stock of nominal bonds, for a given present value of real tax receipts (present and future) the budget constraint is nothing more than the real demand for bonds (an increase in demand for nominal bonds with unchanged real tax receipts relaxes the government budget constraint). P adjusts to make the real value of the supply of nominal bonds equal to real demand.
Adam: I associate that view with the Fiscal Theory of the Price Level, not with New Keynesianism. I agree that FTPL has a theory of the equilibrium price level, in the same way that the QTM does, but in a way that the canonical NK theory (EulerIS + CalvoPC + Taylor Rule) does not. (I think the FTPL is incoherent, but that’s another issue).
Nick, I’d like to see a post on your take of FPTL. Do you take requests? 😛
Also, I don’t see how “Y tu mama tambien” is a valid response to what (you admit) is a valid criticism. What is wrong with saying “Yes, this is a problem with the model”?
Nick, that’s how it’s done uin Woodford’s book.
RSJ: If you search on “FTPL” or “fiscal theory of the price level” within WCI, you get my thoughts on the subject. Like:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/05/john-cochranes-argument-for-fiscal-stimulus.html
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/11/in-praise-of-a-little-bit-of-fiscal-dominance.html
When someone argues that we should reject theory M for theory NK, because M has a flaw, when NK has exactly the same flaw, I think that the “Y tu mama tambien” response is a good one.
In other words, “Y tu mama tambien” is a valid argument iff one’s opponent’s mother does in fact wear army boots to bed.
Adam: OK. But I see that as mixing two separate theories. And it’s a rather unfortunate pairing, because FTPL introduces a non-neutrality of money, so you can’t talk about “the natural rate of interest” any more, since the equilibrium rate of interest depends on the price level. That’s how it makes P determinate.
Nick, you’re wrong. It is not the FTPL, it is still the interest rate target that determines P.
The difference is that Wooodford assumes a Ricardian fiscal regime (where fiscal authorities are assumed unable to print money, only the CB can), thus the government budget constraint actually acts as a constraint and not an equilibrium condition. Thus we still haven’t determined the demand for real bond balances and so we haven’t determined the price level, the CB then controls the demand for real bond holdings and hence the price level via the (conditional) path for short term real rates (that is, via its reaction function).
The fiscal theory of the price level assumes a non-ricardian fiscal regime, thus the path of fiscal primary surpluses is set by the fiscal authorities and if they are short of funds they’ll create more money. In this way the government budget constraint is a constraint in real terms but in nominal terms it isn’t a constraint at all. In nominal terms the “budget constraint” is an equilibrium condition that determines the real value of the stock of nominal bonds, the price level adjusts to make the real value of the supply of bonds equal to the demand for real bond holdings.
Now, it is true that (EulerIS + CalvoPC + Taylor Rule) only determine inflation but that is not the whole model. That is just its reduced form, (actually there is also a Fisher equation).
” FTPL introduces a non-neutrality of money, so you can’t talk about “the natural rate of interest” any more, since the equilibrium rate of interest depends on the price level. That’s how it makes P determinate.”
That’s not true.
A better way to say this is to say that in the Ricardian fiscal regime the government is assumed to adjust the level of real surpluses to accomodate changes in the value of its debt. Thus the “budget constraint” acts as a constraint on what it can do in nominal terms. Since one then needs to know P to evaluate this constraint it’s clear the constraint can’t be determining P.
In the non-Ricaridan fiscal regime the government doesn’t change surpluses to accomodate changes in the value of its debt, thus the “budget constraint” doesn’t act as a constraint in nominal terms, it simply specifies the real value of the outstanding debt as a function of the stream of real surpluses and the price level adjusts so the real value of the nominal stock of bonds equals the real value in the budget constraint.
In the fiscal theory the real value of the debt acts as a constraint in real terms, the value of the debt can’t exceed the present value of the real surpluses. If the government issues more nominal debt the price level rises to enforce the real constraint.
In the Ricardian regime, since the government adjusts primary surpluses to accomodate changes in the value of the debt it is not adjusting the nominal stock of debt.
Adam: I have replied in more detail over on your blog http://canucksanonymous.blogspot.com/2011/06/price-level-determination-without-money.html
Here’s the short version:
Under Calvo pricing, the current price level cannot jump. The central bank’s inflation target plus the pre-determined price level determine the future path of the price level.
Suppose that path of the price level contradicts what FTPL would say, based on the prospective path of government primary surpluses and the existing level of debt. There are 4 possibilities:
1. Canada 1995. Ricardian regime. The government gets its act together and increases taxes or cuts spending.
2. Greece 2012(?). The central bank sticks to its inflation target, the government sticks to its (real) tax and spending plans, and the government defaults on its debt.
3. Zimbabwe. The government grabs an AK47 and forces the central bank to print money and change its inflation target.
4. Woodford. The central bank earns zero profits from seigniorage in a cashless economy. The central bank’s assets are government bonds, and it has zero net worth. If the government goes insolvent and defaults on its bonds, the central bank goes insolvent too. So the central bank is forced to inflate to keep both the government and itself solvent.
I just replied:
http://canucksanonymous.blogspot.com/2011/06/price-level-determination-with-one.html