Mankiw’s Text Hints at Yes, But The Answer is No
One of the standard claims made about the 1970s is that a major contributor to inflation and low growth was OPEC. In his excellent textbook, Macroeconomics, 6th ed., for example, on pp. 274 and 275, Greg Mankiw gives a table showing changes in oil prices along with the related inflation rates and unemployment rates. Of course, he shows a connection: large increases with oil prices are associated with high inflation rates and unemployment rates. President Jimmy Carter leaned heavily on this association in claiming, while president, that OPEC was a major villain.
But how major? It’s not surprising that Carter didn’t answer this but, more surprisingly, Mankiw doesn’t either.
Take as our starting point the quantity equation, MV = Py, where M is the money supply, V is velocity, P is the price level, and y is real income. A good rough way to measure the effect on P of changes in oil prices is to look at the effect of oil prices on y, real income. In 1973-74, OPEC raised world oil prices from $3 a barrel to $11 a barrel, an increase of $8. The effect on the United States, a net importer, was to reduce U.S. real income by about $12 billion a year. This comes from multiplying 4 million barrels a day (mbd) by $8 by 365 days. (I don’t look at the effect of higher prices on U.S.-produced oil because that is a transfer from U.S. consumers to U.S. producers, not a net change in U.S. income.) That’s a one-time drop in annual real income rather than a change in the growth of real GDP. In other words, as long as the higher price persists, real GDP in a given year is approximately $12 billion lower than otherwise. At the time, U.S. GDP was about $1.4 trillion. In 1973, inflation was 6.2% and in 1974 it was 11.0%. So the increase in the price level over those two years was 1.062*1.110 – 1 = 1.18 -1 = 0.18, or 18%. Only one percentage point of the 18-percentage-point increase in the CPI can be accounted for by the OPEC price increase.
Similarly, Mankiw states that the halving of oil prices in 1986 “led to one of the lowest inflation rates in recent U.S. history and to falling unemployment.” It helped, but how much?
Again, use the same method as with the price increase. In 1986, oil prices dropped by about $10 a barrel. At the time, the U.S. was importing 5mbd. So the increase in real GDP was $10*5mbd*365 = $18 billion. In an economy with a $4 trillion GDP, this fall in prices caused a less than 0.5% increase in real GDP and, therefore, accounted for only about 0.5 percentage points of the reduced inflation. Unfortunately, Mankiw’s textbook gives a different impression on both the upside and the downside.
READER COMMENTS
kebko
Nov 13 2009 at 3:16pm
I was surprised recently to realize how close to median wages the minimum wage had gotten by the late 60’s. Has there been any work done in seeing how much of a connection there was between that & stagflation?
spencer
Nov 13 2009 at 4:54pm
the import price of oil is the marginal price that also determines the price domestic producers receive.
Consequently, just weighting the impact of oil by using the weight of imports significantly underweights the impact of oil. All oil went up in price when imported oil prices rose so you should weight the impact by the weight of all oil in the economy.
spencer
Nov 13 2009 at 4:57pm
According to your rational for just using the oil imports weights leads to the conclusion that a closed economy can never have inflation.
I do not think you want to go there.
Jim Glass
Nov 13 2009 at 5:05pm
The big obvious data points here (as Friedman pointed out, of course) are that while the US had “oil price inflation” during the oil crisis years, Germany, Switzerland and countries following the German-led monetary policy did not — and Japan had disinflation. (Japan had an inflationary run up in years just previously, and the BoJ had adopted an explicit policy of reducing inflation that it continued).
This in spite of the fact that Germany and Japan were far more dependent on oil imports than the US was (lacking the US’s substantial “low price” domestic production).
So as to “stagflation”, the oil price shock was a real shock to the world economy that contributed to stagnation, but the “-flation” part was the result of domestic monetary policy, and varied country by country.
david
Nov 13 2009 at 5:06pm
@spencer
I think he does want to go there. “Inflation is always and everywhere a monetary phenomenon”, right? They wrote that to critique cost-push inflation, as I recall.
david
Nov 13 2009 at 5:07pm
*just in case: no, I’m not David Henderson.
8
Nov 13 2009 at 5:15pm
If Saudi Arabia said their reserves are wrong and their fields are in permanent decline, oil would shoot up. And then all the money the Fed’s been printing for the past year would break the dam and flood into the economy.
In 1986, investors wanted to buy stocks and bonds, not commodities.
When there’s asset inflation, its “benevolent” and the Fed looks smart. When there’s commodity inflation, the Fed looks stupid. They have the same policy the entire time, inflate the currency.
David R. Henderson
Nov 13 2009 at 6:05pm
Spencer writes:
the import price of oil is the marginal price that also determines the price domestic producers receive.
Consequently, just weighting the impact of oil by using the weight of imports significantly underweights the impact of oil. All oil went up in price when imported oil prices rose so you should weight the impact by the weight of all oil in the economy.
Spencer,
I handled that issue when I wrote, “I don’t look at the effect of higher prices on U.S.-produced oil because that is a transfer from U.S. consumers to U.S. producers, not a net change in U.S. income.”
And none of my reasoning implies that a closed economy can’t have inflation. Look at MV = Py. That shows you 3 ways a closed economy can have inflation: higher M, higher V, or, as in this case, lower y.
Best,
David
Richard A.
Nov 13 2009 at 7:41pm
Price inflation brought about by increasing scarcities will not cause wage inflation. An increase in the M (or V) growth rate will produce both wage and price inflation. In the 70s, we had both wage and price inflation.
Richard A.
Nov 13 2009 at 8:01pm
David Beckworth has a chart where he refers to the 60s and 70s as “The Great Nominal Spending Spree”.
http://macromarketmusings.blogspot.com/2009/09/great-nominal-spending-crash.html
Joey Donuts
Nov 14 2009 at 8:27am
One can treat payments to domestic oil producers as transfer payments with no effect on GDP if one assumes that all increases in receipts are immediately spent by the Oil producers and/or their stockholders. Although, the stockholders do become better off and one would assume that their expenditures increase at some time, this increase in expenditures does not happen instantaneously and while the expenditures do not occur, a drag on the GDP exists.
Arnold goes to great lengths to point out frictions in the economy. This is one worth watching, not only in 1975 but the oil price increase in 2007-2008 which in current dollars (counting all crude consumed in the US) amounted to a drag of $240 billion over some 15 months.
floccina
Nov 14 2009 at 1:35pm
It seemed then and now that petroleum was to small a part of spend to account for much of the inflation. I have though that unemployment in the auto sector due in part to higher petroleum prices could have contributed to the fed following an inflationary path.
David C
Nov 14 2009 at 2:59pm
1) What was your source for determining an $8 per barrel increase reduced real income by $12 billion?
2) If oil prices go up, this would cause the price of goods that depend on oil to be produced to also go up. This would in turn cause other goods to go up in price eventually leading to wages being cut. Might these effects take some time to disseminate through the economy? How much is stickiness a factor? How likely is it that $12 billion is a significant underestimate?
ThomasL
Nov 16 2009 at 9:35pm
I too always associated the “stag-” part of stagflation with a shock to commodity prices (particularly oil) and the “-inflation” half with intentionally pursued government policies.
The government wanted to stimulate employment and growth generally and pursued inflationary policies to do it, but increasing commodity prices prevented their actions from having the intended effects.
Using inflation to [mis]direct employment was always a bad idea, but the 1970s managed to chase a bad idea that also didn’t even work as intended.
Just for clarification, is Greg Mankiw really arguing that the oil prices _caused_ the general price inflation of the 1970s?
Comments are closed.