The Economics Babel
One thing that the current crisis has done is to expose just how much economists differ in their mental models of the macroeconomy. It seems as if we often talk right past each other, not really understanding the points the other side is trying to make. And I don’t think this is an inevitable part of academic discourse. In microeconomics, for instance, Chicago economists may differ from salt water economists about just how competitive most markets are, but at least they can debate the issue using a commonly understood language. Not so in macro.
Take the writings of Keynes, which have about as many possible interpretations as the Bible (and are about as consistent as the Bible as well.) Krugman has recently been ridiculing distinguished economists who don’t understand Keynes, or at least don’t understand Keynes the way he does. Tyler Cowen responded by noted that Keynes had spoken out against using fiscal policy to smooth out the business cycle. So perhaps even Krugman isn’t really a Keynesian. I wouldn’t deny that Krugman is sharp enough to spot the occasional lapse in logic by conservative economists, but I also think that all sides of the debate tend to underestimate just how much macroeconomic worldviews differ, and the extent to which we even lack even a common language to debate issues. Here are a few brief examples:
1. What does the term “monetary policy” mean? A Mundellian might visualize monetary policy in terms of the price of money (in terms of gold or foreign exchange.) Thus Mundell argued that the price of gold was set too low after WWI, and that that was a root cause of the subsequent deflation. Friedman might visualize monetary policy in terms of the quantity of money (M1, M2, etc.) and see the cause of the Depression as being the sharp fall in the monetary aggregates in the early 1930s. Most economists visualize monetary policy as the rental cost of money, the interest rate. For them, money is too tight if interest rates are set at a level too high to allow for full employment. So there are at least three very different indicators of policy. But even worse, it’s not clear how to interpret any individual indicator. Michael Woodford has argued (correctly in my view) is that what matters is not changes in the current setting of the policy indicator, but rather changes in the expected future path of that indicator. He also argued that what matters is not the level of interest rates, but rather how they compare to the (unobservable and always changing) Wicksellian equilibrium rate.
2. What is the Phillips Curve model? Friedman developed a “Natural Rate Model” where unexpected changes in inflation pushed unemployment above or below the natural rate (which was itself unobservable.) In his view, an expansionary monetary policy would raise M*V, or nominal spending. In the short run some of that increase would show up as increased output (due to sticky wages), and some as higher inflation. Today, however, most economists see the Phillips Curve model very differently. Expansionary monetary policy might lower interest rates, and if interest rates fall that might boost investment and real output, and if real output rises that might boost inflation. What Friedman saw as the direct effect of expansionary money, is instead a long run and very uncertain effect. Because I take Friedman’s view, I find I have trouble explaining to other economists why the current problem is simply to get an expansionary enough monetary policy to boost nominal spending. They see the problem more in real terms–will monetary stimulus lower interest rates and raise real investment spending? Inflation is an effect of growth, not of monetary expansion.
I often observe people switch from the Keynesian mental framework to the Friedman mental framework in a matter of seconds. In an earlier post on hyperinflation I mentioned that most people today do not believe that monetary expansion can boost AD. But when you give them a reductio ad absurdum example of running the printing presses and buying up everything in sight, they envision hyperinflation. To do so they switch to the monetarist excess cash balance model (assuming they are not a skilled macroeconomist who understands Krugman’s “expectations trap” model.) The transition is not smooth, it’s as if they are forced to speak a different language. One can can briefly observe the cognitive dissonance on their face when they make the transition.
Bennett McCallum identified no less than 10 different versions of the sticky wage/price model, each with different theoretical implications. And all that is within mainstream macro. If one moves into Austrian macro, post-Keynesianism, or even the writings of Keynes himself, the language changes much more radically. What do phrases like “savings exceeds investment” really mean?
Elite macroeconomists may have a common mathematical language with which they can communicate via DSGE models (a language I do not speak) but it is becoming increasing apparent that when thrown out in the open to face complex real world problems, they revert back to their instinctive worldview, which varies greatly from one economist to another. Some say monetary policy is obviously impotent today, others cannot imagine a more absurd idea under a fiat money regime. And no one knows whether familiar models like IS-LM are of any use in resolving that debate.
Tags:
12. February 2009 at 10:24
It is hard to see how an increase in the money supply can increase nominal expenditure unless it somehow increases some element of real expenditure. I don’t find it plausible that an increase in M will result in a matching decrease in V, but someone must be responding to the higher M by trying to purchase something, otherwise, it seems to me that V will drop to match an increase in M. Someone must be spending excess money balances on something, or the demand for money must be rising to match the added supply.
The relationship between open market operations, nominal interest rates, real interest rates, and so efforts to buy something, seems reasonable. What is the alternative?
Helocopter drops is a combined monetary/fiscal policy–which is what tax rebates will amount to under current conditions.
On the other hand, the notion that we have lowered the target for Federal Funds close to zero, and the T-bills that are usually purchased are closer to zero, so, monetary policy can’t work, appears to me to be confused. The traditional approach to monetary policy isn’t working. Something must change. It has always been obvious to me, anyway, that only after the entire 5 trillion national debt has been purchased using base money, can we even begin to say that there is a “problem” with monetary policy being effective. And then you start looking at private securities.
And, of course, as long as the Fed pays interest on reserves, how can anyone take seriously any claim that monetary policy isn’t working?
As for the phillips curve business… really? Demand raises output and maybe prices? I thought everyone figured that firms sell more, and they raise prices and production–with the added production being a “mistake” in some kind of global sense.
I think you are being fooled because everyone only cares about the added output, and the inflation is an unfortunate side effect. But, maybe I just am assuming they must see it like I do.
15. February 2009 at 06:11
Bill, Here are a few random observations about the money transmission mechanism.
1. How does money raise prices during hyperinflation? Surely it is not through the Keynesian mechanism of higher real output?
2. How does money raise prices in a primitive economy with no financial system–say an island nation using shells for money? It’s not through lower interest rates, it is through the “excess cash balance” mechanism. If one thinks of money as just another good, then an increase in the money supply will lower the relative value of that good in the same way it would for any other good. We don’t say that more apples lowers the relative value of apples only if interest rates change, why is money any different?
3. Of course we do have a sophisticated financial system and interest rates do play a role. But consider the other following mechanisms. If we suddenly have 20% more money, and it leads to higher expected future prices, then there will be an immediate effect on all flexible prices set in auction-style markets. The the dollar will immediately depreciate–pushing up the prices of imports. The prices of commodities will rise immediately if their future expected prices rise. The price of stocks will rise immediately. the price of real estate will rise. This is exactly what happened when FDR depreciated the dollar. The reason that it is usually hard to see, is because monetary policy is now very endogenous, and it is very hard to identify monetary shocks.
17. February 2009 at 04:57
1. I find it hard to believe that anyone believes that increased demand only impacts output and then perhaps prices. I think the natural idea is that increased demand impacts both price and quantity at the same time. If one is trying to raise output, then perhaps it is natural to also think–we get the increase in output, good. Ah, but what is the cost in terms of undesirable price increase. Of course, there is also another way of looking at it. In some sense, only prices “should” change, adjusting the real supply of money to the demand. Why is it that there would be an increase in output?
As for hyperflation, the impact of expected prices on supply and demand for particular goods is only going to happen if people believe that increased money will impact the real demand for something. If sellers raise their reservation prices enough, then there is no impact on real output. But, this is conditional on them believing that if they didn’t raise their prices, there would be an increase in real demand. They are preemptively choking it off, right?
2. Where do the new shells come from? Any helocopter drop story is like fiscal policy financed through money creation. I do expect that this will increase demand. And as the money passes through cash balances as it is earned and spent, it appears no different than any other increase in income. In the shell economy, there isn’t anything like money being lent into existence (so that someone must pay it back) or else an open market purchase, so that those receiving the money initially have given up an asset they were holding.
3. 20% more money will increase the expected price level. Why? Why? It seems to me that arguments through expectations assume some other transmission mechanism. Well, if expecatoins are rational. Anyway, if the current price level is 20% above equilibrium, then the 20% increase in the money supply simply raises the equilibrium price level back to equilibrium. There is no increase in prices. I suppose, of course, that this story is inconsistent with “auction-type” markets.
I don’t believe that interest rates actually need to change for money created through lending or else open market operations (which of course involves lending on the secondary market) to increase real demand. If interest rate elasticities are really high==perfectly elastic, then interest rates don’t change at all. But I think that transmission mechanism generally does involve changes in interest rates.
I also agree that thinking about the supply and demand for money is a powerful tool. (spending excess cash balances.) If additional money doesn’t cause additional spending, then there must be some account for why people are willing to hold the additional money. Usually, claims about inneffective monetary policy do provide such a rationale. Too bad they aren’t more explicit about it.
17. February 2009 at 10:51
Bill,
You are probably right that people generally don’t believe AD impacts only prices in the short run. I think that was Keynes’ assumption in the GT (at least for less than full employment), but modern Keynesians would presumably see some price level impact. Even so, I think many Keynesians over-estimate the extent to which money can only raise prices by raising spending.
I see the hyperinflation case differently. Consider what happens if the supply of apples rises sharply and the nominal price falls by 80%. In that case one could say “the general price level in apple terms has just increased five-fold.” If asked to describe why the price level in apple terms increased, no one would talk about an increase in the real AD for non-apple goods. It’s a relative price change. The same is true for a five-fold increase in money, with one exception, money is the medium of account. If wages and prices were completely flexible, then I’d say it’s exactly like the apple case–the price level rise would have nothing to do with an increase in the real demand for all other goods. (Although as you say, the nominal supply and demand curves do shift.) In fact, wages and prices are sticky, so real AD does rise in the short run. But this is incidental to the process that causes prices to be higher in the long run (which would occur even more quickly if prices were not sticky and real AD did not increase.) The effect on real output and real AD is a consequence of wage/price stickiness, and is not essential to the monetary transmission process in regard to prices. (Does my apple analogy make sense?)
You are right about the shells–I forgot the fiscal angle. Even so, I think the fiscal effect is far less important than the monetary effect. The effect (in normal times) of a $80 billion tax rebate is trivial compared to an $80 billion increase in the base (which would have been 10% before the recent crisis.)
On the interest rate transmission mechanism, I don’t think it’s merely a question of elasticities. In the example I gave in the post on Ratex, the Dec 2007 contractionary surprise lowered bond yields from 3 months to 30 years (due to the income and Fisher effects.) These are the maturities that matter to investors. But even though there was a surprise fall in these rates, the markets saw the contractionary nature of the announcement, and stocks fell sharply. I view the fact that contractionary monetary surprises do raise overnight rates as an incidental effect flowing from wage/price stickiness, that plays almost no role in the monetary transmission mechanism.
14. August 2011 at 13:38
[…] monetary policies (some of which is tangled up in macro-economics’ lack of a common analytical language); law-and-order conservatives endorsing torture (relabelled, in true PoMo style, as ‘enhanced […]
19. January 2012 at 07:42
[…] of my very first posts compared macroeconomic debate to a Tower of Babel. One reason we think our opponents sound so stupid, is we use completely different frameworks […]
11. July 2012 at 21:26
[…] was aware that modern macroeconomics is bedevilled by a lack of a common analytical language in talking about money. But the problem goes deeper than that; money is a multi-dimensional […]