Thursday, February 23, 2017

Oil shocks?

Yesterday I said the oil shocks of 1973 and 1979 could explain the stagflation of the mid and late 1970s but they do not explain the stagflation of the early 1970s which, I said, was "plain as day" on this graph:

Graph #1: Quarterly Data Suggesting Times of Stagflation. Early 1970s rung up.

Then I came across Oil and the Macroeconomy since World War II by James D. Hamilton.

Hamilton says

All but one of the U.S. recessions since World War II have been preceded, typically with a lag of around three-fourths of a year, by a dramatic increase in the price of crude petroleum.

My knee-jerk was No, that's not right. I was there in the 1970s. That doesn't help, though, because Hamilton is talking mostly about before the 1970s.

My problem is that maybe there was an oil shock that created the stagflation of the early 1970s. If so, maybe what I said yesterday is wrong. Maybe the oil shock of (say) 1968 could explain the plain-as-day stagflation of the early 1970s.

You shoulda seen me scurry! I went right to FRED to look at oil prices. I grabbed the first two datasets I found that go back to the 1940s, where neither one hides the other. Why two? Because sometimes you can look at two different measures of the same thing and they are nothing alike. For me, that means I've got something wrong. If I get two datasets that show a similar pattern, I gain confidence that I've picked good data.

Here's what I got:

Graph #2:The Price of Oil, 1946-2017
Similar pattern. Oh, one of 'em is an index. Eh, that's all right I guess.

Hard to see any "dramatic increase in the price of crude" before 1973-74. But then, Hamilton's paper is from 1983. Benefit of the doubt: Maybe the big increases after '83 make the increases of the early years relatively small and hard to see.

Here's a look at the same data up to January 1983:

Graph #3: The Price of Oil, 1946-1983
Yeah I dunno, I still don't see much price activity before the 1974 recession.

Maybe it's me. Well, we can look at percent change from year ago. Maybe that'll help:

Graph #4: Percent Change from Year Ago in Oil Prices, 1946-1983
Not really. Not much activity there, from the 1950s to the 1974 recession. There are maybe four times when the red line jogs up a little bit, holds steady for a while, and then falls. The biggest one of those four is the one at the 1954 recession. It shows an increase of less than ten percent -- and that's the biggest one. If the price of gas was 30 cents a gallon before that increase, it was less than 33 cents a gallon after. That's not an oil shock.

The price of a barrel of "West Texas Intermediate" went from $2.57 in May of 1953 to $2.82 in June of 1953. The price stayed at $2.82 until February of 1957. The dates do seem to be related to recession dates. But I don't see any "dramatic increase".

On Graph #4 the red line stays high for a year after the increase because we're looking at "change from year ago" values. It's not like the price increased every month for a year. The price increased once in 1953, and did not increase again until 1957. And those were pretty small increases, in my book.

But we were talking about the stagflation of the early 1970s, and any possible "oil shock" that might have caused it.

I don't think so. Here. This graph compares the price of oil (blue) to the general price level (red):

Graph #5: Crude Oil (blue) and Consumer Prices (red)
The two lines start out the same, in January 1959. Both lines go up from there. But the red line goes up faster, until 1974. The price of oil (blue) lags behind.

You can see the increase in oil prices just before the 1970 recession, and the one just after. (Those two increases are also visible on Graph #4.) Those increases don't seem to have any major impact on the path of the general price level. I'm more inclined to say the increases in oil prices came in response to increases in the general price level.

Hamilton considered that possibility. In the PDF he calls it "an inflationary 'catch-up' effect". He says the evidence is not strong but the "catch-up phenomenon was surely operative".

James Hamilton also says his findings suggest that "oil shocks were a contributing factor in at least some of the U.S. recessions prior to 1972." Okay, but everything contributes to everything. Hamilton's dinky little oil shocks do not by themselves account for the stagflation of the early 1970s, far as I can see.

Oh, and I don't find the word "stagflation" anywhere in Hamilton's PDF.

No comments: