Saturday, February 18, 2017

How an economist thinks

I usually like reading David Glasner for snippets of economic history. But I was a little disappointed when I read his old post on 1970s Stagflation. Here is Glasner's opening:

Karl Smith, Scott Sumner, and Yichuan Wang have been discussing whether the experience of the 1970s qualifies as “stagflation.” The term stagflation seems to have been coined in the 1973-74 recession, which was characterized by a rising inflation rate and a rising unemployment rate, a paradoxical conjunction of events for which economic theory did not seem to have a ready explanation.

The term was not coined during the 1973-74 recession. Wikipedia:

On 17 November 1965, Iain Macleod, the spokesman on economic issues for the United Kingdom's Conservative Party, warned of the gravity of the UK economic situation in the House of Commons: "We now have the worst of both worlds—not just inflation on the one side or stagnation on the other, but both of them together. We have a sort of "stagflation" situation. And history, in modern terms, is indeed being made."[3][5] He used the term again on 7 July 1970, and the media began also to use it, for example in The Economist on 15 August 1970, and Newsweek on 19 March 1973.

Stagflation is not an incidental or inconsequential part of econ. I find it distressing that Glasner has so little concern about stagflation that he doesn't know the origin of the term, is willing to guess the origin, and is satisfied to make a bad guess. (If my view seems extreme, let it reflect the importance of stagflation in my economic thinking.)

Glasner does offer a good description of stagflation as "a paradoxical conjunction" of rising inflation in a stagnant economy. And he is right about the reason stagflation was significant: Economic theory could not explain it.

Glasner links to Karl Smith, who says

The 1970s were definitely an era of stagflation

and to Scott Sumner, who quotes that line and disputes it. As Glasner has it:

Scott observed that inasmuch as average real GDP growth over the decade was a quite respectable 3.2%, applying the term “stagflation” to the decade seems to be misplaced.

What is the argument here? Given that we did have rising inflation in a stagnant economy at times in the 1970s, the argument seems to be whether it is okay to call the decade an "era" of stagflation. Sumner says it is not okay, because growth on average was "normal" in the 1970s.

But Karl Smith does not say growth was slow in the 1970s:

Scott is correct that we remember the 70s as an era of slow growth but indeed GDP growth was rapid.

Smith agrees with Sumner that growth was good in the 1970s. Where is the problem?

They agree that growth was good. But Smith points out that there was stagflation. And Sumner says No, growth was good. Quite the non sequitur.

Hey... It is important to get the facts right. If growth was good in the 1970s, we need to know it. But Karl Smith does not dispute that growth was good in the 1970s. So what's the problem?

Maybe the problem is semantics: For Scott Sumner the word "era" implies "the whole decade". And then he mistakenly equates "stagflation" with "stagnation". Thus, Smith says there was stagflation in the 1970s, and Sumner apparently takes him to mean there was stagnation for the whole ten years of the 1970s. Sumner responds, saying real growth was normal and averaged a 3.2% annual rate in the 1970s.

The kindest interpretation of their dispute that I can offer is that Sumner knows that many people say growth was slow in the 1970s. And he knows that growth was not slow in the 1970s. And the discrepancy is a sore point for him. I sympathize.

Scott Sumner says it is incorrect to think that growth in the 1970s was slow. But so does Karl Smith. Come to think of it, Paul Krugman said it too. And David Glasner:

... if one looks at the periods of rapid increases in aggregate demand in which oil price shocks were absent, we observe very high rates of real GDP growth.

So Scott Sumner and Karl Smith and Paul Krugman and David Glasner and I agree that growth was good in the 1970s. And look at this graph from Marcus Nunes:

Graph #1 Source: Marcus Nunes. See also here and here.
Marcus shows inflation-adjusted GDP on a log scale, so that a constant growth rate appears as a straight line. In red, he shows a constant-rate trend line. And he marks up the graph to identify different periods. Look at the period labeled "G.I." for "Great Inflation" -- the inflationary years from 1965 to 1980.

Marcus's graph shows real GDP (blue) at or above trend for the entire inflationary period. By contrast, before 1965, and again after 1980, the blue line is at or below trend. The inflationary period shows particularly good economic performance.

So that's Scott Sumner and Karl Smith and Paul Krugman and David Glasner and Marcus Nunes and me. And the third economist named in Glasner's post, Yichuan Wang, says real growth should get a boost from inflation like we had in the 1970s (though he doesn't see it himself, according to Glasner).

So yes, there's good reason to be careful when talking about stagflation in the 1970s, inflation and stagnation in the 1970s. Good reason. And yet, Karl Smith *is* careful. He explicitly says growth was good in the 1970s. And he says it immediately after he calls the 1970s "an era of stagflation". Why, then, does Sumner choose to disagree with Smith when Smith is trying to agree with Sumner?

Why? Because Sumner has an agenda:

Rather than arguing over semantics, I’d rather focus on the important issue; what does the 1970s tell us about NGDP targeting?

You might have guessed. Sumner wants to talk about NGDP targeting. He doesn't want to argue over semantics. The whole "semantics" argument is a straw man that Sumner set up so he could say he doesn't want to argue over semantics. Sumner is out to get attention for his hobby horse, his NGDP targeting hobby horse. He says so himself.

And Sumner will stop at nothing to get that attention. He even re-defines "stagflation" to suit his purpose. Here's Glasner:

The term stagflation [means the combination of] a rising inflation rate and a rising unemployment rate ...

Yes. Stagflation is the increase in the two rates that, when added together, give the "Misery Index". But Sumner criticizes Karl Smith for using the same definition Glasner uses. Here's Sumner:

Karl seems to think [stagflation] means high inflation plus other bad things, like high unemployment.

Scott Sumner sticks a parenthetical finger in the definition of stagflation:

I had thought the word ‘stagflation’ meant high inflation plus slow output growth (due to slow growth in AS.)

By Okun's law, "slow output growth" is equivalent to "high unemployment". That change of wording is only a distraction. But Sumner modifies the concept when he adds the words "due to slow growth in AS". He puts those words in parentheses, as though they don't really change the definition. But those words change everything! By adding the causal factor to the definition of stagflation, Sumner changes everything.

Sumner's "AS" means "aggregate supply". Sumner's extra words make stagflation explicitly and exclusively a result of supply-side factors, by definition.

Sumner himself points out that he is changing the definition:

I think if the term ‘stagflation’ is going to mean anything useful, it has to refer to a periods where, for any given rise in AD, slower than normal AS growth leads to higher inflation. The 1970s do not meet that definition.

Sumner re-defines stagflation to suit his agenda, and suddenly the world is different. You have been bullshitted.

Glasner describes stagflation as

a paradoxical conjunction of events for which economic theory did not seem to have a ready explanation.

Glasner is right. The lack of explanation is the reason stagflation was such a big deal: Stagflation didn't fit with what economists knew about the world. The whole point of raising interest rates to reduce inflation is that it works by slowing the economy. Anti-inflation policy creates stagnation. It still does. But in the 1970s, inflation and stagnation were thought to be mutually exclusive. It was "inflation on the one side or stagnation on the other", as Iain Macleod said. Except, during the inflationary 1970s, we got inflation and stagnation at the same time. It meant there was something wrong with economic theory. It was a big deal.

It opened the door to Milton Friedman and Monetarism and Paul Volcker and all that came after. That's why stagflation is important. That's why it matters. To show that Milton Friedman and everything since Friedman is wrong, if that is what you might want to do, it is necessary to go back to stagflation and review what happened then, and think it all through. Instead of blindly accepting the notion that the whole decade of the 1970s was a time of stagnation. And instead of blindly accepting Scott Sumner's agenda-driven alternative.

No comments: