Arnold Kling  

The Evolution of Macroeconomics, Part Two

Crosstalk on Financial Regulat... More Classic Jeffrey Friedman...

My latest essay:

We have accumulated more than six decades of macroeconomic experience since the end of World War II, yet neither stubborn Keynesians nor stubborn monetarists have encountered any data that would make them change their minds. Instead, since 2008, the Keynesians have effectively "taken back" all of the concessions that they made in the 1970s and 1980s.

Mark Thoma and I are supposed to be on a panel in a couple of weeks, but talking about something else. Still, when we get together outside the panel I hope I can coax a reaction to these essays out of him.

[UPDATE: Bill Woolsey's comment below makes me think of the straightforward argument for market monetarism:

1. What's the worst thing that would happen if the Fed tried to level-target nominal GDP at a 5 percent annual growth rate relative to when we were at full employment?

2. What's the worst thing that would happen if it didn't?]

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (7 to date)
david writes:

I think it was Nick Rowe who recently observed that Krugman seems to not realize that he (Krugman) isn't talking like a pre-Friedman Keynesian, but rather like what 60s Keynesians would have called a monetarist. Rhetorically Krugman wants to claim the old Keynesian mantle but he isn't actually in the same intellectual camp.

60s/70s Keynesians talked about the distinction between demand-pull and cost-push, and argued that the apparent correlation between output and money had causation going from the former to the latter rather than vice-versa. The battle between rules vs. discretion came only after the major battles of monetarism had already been won intellectually. I think you may be mischaracterizing the evolution of macroeconomic theory.

Friedman made all the self-described Keynesians push their own endogenous money theories down the memory hole, but you don't have to do the same thing.

On the monetarist side, note that Friedman's own simple %money-base-growth rule had to be rapidly renounced by Friedman himself, as soon as it became evident that the money multiplier was not in fact stable. A non-discretionary Taylor rule is already considerably more complicated than Friedman's original proposal (it requires some estimation of the optimal response constants). In a sense, the criticisms of the old Keynesians there were appropriate.

Bill Woolsey writes:

"Neither stubborn Keynesians nor stubborn monetarists see a need to take into account financial markets or structural unemployment. Instead, they take the view that any desired macroeconomic outcome can be achieved with the right fiscal and monetary policy approach."

"any desired macroeconomic outcome" is much too strong for market monetarists.

Monetary policy can keep deviations of nominal GDP from target minimal. Market monetarists consider this the least bad option.

But there is no notion that this "monetary" approach will stabilize inflation, output, or unemployment.

It is rather that fluctuations in nominal GDP will cause changes in production, employment, and inflation on top of any other sources of such changes. And that any fluctuations in inflation, output, or unemployment from other sources will not be significantly helped by shifts of nominal GDP away from a stable growth path.

BZ writes:

Thanks Dr. Kling -- I ate both of those essays up. Rating: 10/10 on the informative and entertainment scale.

JeffM writes:

It may be presumptuous of me to submit a comment at all because I make no claims to being an economist. However, I like to use economic theories, and I have developed (perhaps semi-developed) and found helpful a meta-theory about economic theories, namely that most economic theories are "theories of possibility."

That is, economic theories say what sorts of things may happen with some non-negligible probability. They are not like physical theories that say what will happen. An increase in demand means that quantity sold may increase, that relative price may increase, or both, but it does not exclude the possibility that relative price and quantity sold may both decrease because the whole theory is built on a ceteris paribus assumption that will almost certainly be violated in any real-world situation.

To take another example, if the money supply were to be increased by 20% in the course of a year, the probability would be almost 1 that the general price level, approximated by any reasonable method of measurement, would have increased between 5% and 35% over the same period. If, however, the money supply increased by 20% over the course of ten years, I doubt it is possible to assign a reliable probability greatly different from 50% to what even the sign of the change in the general price level would be over that period.

If it is true that economic theories are generally "theories of possibility," then the question becomes whether the possibility in question occurs with enough empirical frequency to be practically relevant. Under this criterion, Keynesianism becomes a very interesting case.

Obviously those who argue against Keynes on the grounds that his supposed equilibrium cannot possibly be even approximately a long-term equilibrium are technically correct. But the "General Theory" is really a book about dynamics dressed up as a static theory. If interpreted as a theory about "ghost points" in a dynamic system, I suspect it can be rescued at the formal level. (I am not going to try; my knowledge of mathematical dynamics is not up to it, and I am not an economist.) Such a rescue, if feasible, would necessarily be a theory only applicable as a close approximation to the short term because a dynamic system always escapes from a ghost point eventually. (Keynes himself would not have been bothered at all if his theory turned out to be only a short-term approximation.)

What interests me far more than what is "formally" wrong with Keynesian theory and whether it can be "formally" cured is that Keynesian policy has sometimes worked, at least in the short term, and sometimes failed, even in the short term. The Kennedy tax cut was a success. The Reagan economic policy was a success, and it contained (as is seldom recognized on either the right or left) a strong Keynesian element. Examples of Keynesian failure are the economic policies of Carter and Obama.

If all that is correct (and of course it may be nonsense), what would be useful is to ask under what circumstances has the Keynesian model been found useful and under what circumstances has the model been useless or even deleterious. To put it another way, the model seems to be missing one or more elements that would make it useful on a fairly consistent basis instead of unpredictably useful.

We are never going to get a "true" model of anything as complex as a modern economy. As Joan Robinson said, what tools are useful to have in our toolbox. I for one am perfectly willing to leave it to economists to design and test the tools and say when it makes sense to use which tool or tools. If, however, we have to wait for perfect tools, then economics is a perfect waste of intellectual resources.

I hope this intrusion by one not an economist is not considered untoward.

Blake Johnson writes:

I think while it may come off that Market Monetarists like Scott are saying that the economy will run perfectly if only monetary policy were done correctly, I think Bill Woolsey is correct in saying that is a mis-characterization. Even in the last few months, Scott has talked about other policies, some of them related to banking and others not, which are causing problems of their own.

The point of good monetary policy is not as a panacea to all of our macroeconomic woes. It is merely to make sure that we don't turn a mild recession into a Great Recession unnecessarily, or to turn what is a negative supply side shock into stagflation by obfuscating the activities of the central bank.

In point of fact, it is interesting to ask what the big difference is between economies which did not suffer through the worst of the Great Depression and the Great Recession, such as Canada and Switzerland, or Australia. One thing they have in common is (or at least did until this most recent trouble with the head of the Swiss central bank) is a credible commitment to an expansionary monetary policy if needed. They were credible because the central banks set explicit targets, and as a result the market did a lot of the lifting for them.

kio writes:

This is already a common place that the mainstream economics has failed to explain the current crisis. It is good to know that it has also failed to explain the Great Moderation, i.e. there is no measurable cohesion between monetary/fiscal policy and economic developement.

Do you really think that any qualitative discussion can help to understand the economy? Or one still needs a quantitative description to use some true cohesion to control the economy?

Mike Rulle writes:

This whole topic reminds me of that great album by the 1970's comedy troupe, Firesign Theater, "Everything You Know is Wrong".

If results like this existed in physics, it would be discontinued. While my comment obviously puts me into the "anti-modernists" category, "less is more" seems like a good over arching principle for government.

My wife and I raised 3 children. Like Hippocrates our first rule was "first do no harm". We always thought we had almost unlimited power to create havoc and harm, but considerably less power to achieve a "well intentioned" targeted result.

Have a few hard and fast rules that seem to make sense to all, and then within that zone let freedom and its consequences prevail.

In the grand scheme of things, how much difference can the Taylor rule or NGDP really make? Whatever difference that might be, I doubt we will or can ever know.

Comments for this entry have been closed
Return to top