Weekend read: What caused the stagflation of the 1970s? Answer: Monetarism

from Philip George

 If any one event can be said to have knocked Keynesian economics off the high perch it occupied during the 1950s and 1960s it was the stagflation of the 1970s.

Keynes held that when aggregate demand fell due to a fall in business investment the government could stimulate the economy by increasing its own spending and lowering interest rates. Inflation would kick in only when the economy approached full employment. Monetarists questioned the ability of the government to control the business cycle through fiscal spending or loose monetary policy. They held that loose monetary policy could affect output only in the short run, and that in the long run it would only result in inflation. When the US economy simultaneously experienced both high unemployment and high inflation in the1970s, monetarism seemed to have triumphed.

If Keynesians had no policy or theoretical response to monetarists it was because they thought of money in the same way as monetarists did, as exemplified in the image below.

The economy is like a boat floating in a tank into which money is being poured. Monetarists argued that the top of the tank rose each year by an amount proportional to the growth rate of GDP. Only when the money being poured into the tank exceeded that level would the inflation meter start ticking. Keynesians held that money could be poured into the tank as long as the level was below “full employment”. Only after that level would the inflation meter start ticking. Allowing for differences of emphasis between monetarists and

Keynesians the model was substantially the same, which was why Keynesians had no answer to the monetarist claim that when unemployment fell below a certain level, inflation would accelerate.

But while the model seems to work for closed economies it fails for open economies where inflation comes in, for example, through external oil price increases. In such cases, we have to analyse the economy from a completely different viewpoint, by regarding money as a medium of exchange, whereas both Keynesians and monetarists regard money as an asset.

The figure below plots year on year changes of West Texas Intermediate crude price against consumer price inflation. From 1974 onwards it can be seen that inflation is always triggered by changes in oil prices. I have used West Texas Intermediate as a proxy for international oil prices.

The problem is to estimate the effect of oil price increases on inflation. For that, consider the economy as a firm which has an input A accounting for 5% of total input costs. If the price of this input goes up by 20% while the prices of all other inputs remain unchanged then the firm has to raise the price of its own output by 1% to keep its profit at the old level.

The graph above shows that in 1974 the price of crude went up by an average of 170%, year on year. At that time energy costs accounted for 10% of the US economy. A rough-and-ready calculation shows that a crude oil price increase of 170% would translate into inflation of 17%. The higher cost of imports and, by transmission, the higher prices of goods and services would call for a greater amount of money to effect exchanges. If the central bank allowed money supply to grow by 17% then we would have inflation of 17% (assuming a one-to-one relationship between money supply growth and inflation) but no reduction in growth. If the Fed instead curtailed money growth to less than 17%, then we would have both inflation and a contraction of GDP (and hence higher unemployment). This is what happened after global oil prices began to shoot up. In 1974 the average rate of growth rate of M1 was cut to an average of 5% during the year from 9.1% at the start of 1973. The average inflation in 1974 was 11%, less than our calculated 17%, but the cost was a severe recession.

Inflation in 1974 was caused by higher prices of oil imports. The contraction in growth occurred because the Fed, under the mistaken view that inflation was caused by too large increases in money supply, curtailed money supply growth instead of allowing it to expand by the amount required by imported inflation. The rise in unemployment was caused not by too much money but by too little money.

In the first six months of 1980, oil prices rose by an average of 127%, year on year. By then energy accounted for 13.5% of the economy, so inflation should have been 17%. If money supply had been allowed to grow to this extent, there would have been inflation of 17% and no effect on growth. The actual money supply growth (M1) was 6%, roughly the same as in the previous year. Inflation was on average 14% but the reduction in inflation from the calculated 17% was paid for in the form of a recession, which lasted from January to July1980.

From July 1980 the price of oil began to fall steadily. Between July 1981 and November

1982, the price of oil fell 6.5% on average each month, year on year. Oil accounted for about 13% of the economy, so inflation should have fallen to about 1% over time, without any intervention. Instead, the Fed squeezed M1, whose growth during this period was about 6.3%, a fall from 11.5% in April 1981. Inflation fell sharply from 10.8% to 3.8%, but the cost was a severe recession. Again, stagflation was caused not because of too much money but because of monetarist ideas.

If an oil-importing refiner was importing 1 million barrels of crude a day at $15 and the price rises to $30 then it needs to pay twice the amount of money to import the same amount of crude that it did earlier. Also, those to whom it sells the refined gasoline will need more money than they spent earlier to buy the same amount of gasoline. Money supply needs to be expanded to permit this. Else, less crude will be imported and less gasoline sold down the line, resulting in a fall in GDP and a rise in unemployment. Instead, central banks typically reduce the growth rate of money, which explains the stagflations of 1973-75, 1980 and 1981- 82.

The 1969-70 inflation was not caused by either high oil prices or too much money but by increased expenditure on the Vietnam war. The figure below shows expenditure on defence as a percentage of GDP. In 1968 spending on war rose to 8.6%. To understand the effect of this on inflation, it is again necessary to view money in a way different from how Keynesians and monetarists have customarily viewed it.

Consider an economy working at full employment. It produces $90 worth of consumption goods. An equal amount is therefore paid out in income flows: wages, rent, and profit. If the saving rate is 10%, $9 of these income flows is saved and $81 is spent on consumption goods. That leaves $9 worth of consumption goods unsold. But the economy also spends $10 on investment, which means an equal amount is paid out as income. Since the saving rate is 10%, $9 of this income is spent on consumption goods and $1 is saved. The $10 of total saving is channelised into investment by financial institutions.

Now assume that this economy goes to war. $5 now has to be spent on defence goods and services. Since the economy was operating at full-employment, real resources newly spent on defence have to be diverted from the production of consumption goods. But an equal amount of income, $100, is paid out as before. Assuming the same saving rate as before, $90 is now available to be spent on consumption goods but only $85 worth of consumption goods is now being produced. The result is inflation, which is caused not by too much money (the central bank has not expanded money supply) but by a shortage of consumption goods.

The central bank, however, believing that the inflation was caused by too much money curtailed the growth of money supply from March 1969. This caused a reduction in output, which exacerbated inflation instead of reducing it, and increased unemployment. The result: inflation and high unemployment or, in other words, stagflation.

Thus, the episodes of simultaneous higher inflation and higher unemployment in 1969-70 as well as the subsequent three recessions were caused by the central bank’s mistaken belief that inflation was caused by an unwarranted increase in money supply. Stagflation was caused not by too much money but by the central bank’s belief in monetarism.

When money is viewed as a medium of exchange, as above, we can also question the concept of NAIRU. When unemployment falls the money being put into the economy in the form of wages goes up. But since the newly employed are also producing goods and services on which wages can be spent, inflation will not accelerate.

What is a bit difficult to explain at first glance is the situation in 1953, when defence spending was 11.3% of GDP but inflation was less than 1%? A publication of the Federal Reserve Bank of Minneapolis, Monthly Review, dated January 31, 1952, offers clues to an explanation. “Perceiving the worsening international situation at the close of 1950 and studying the defense programme, most consumers in the early weeks of 1951 rushed into the market places to buy goods which they feared might soon be in short supply. They then observed that the shelves were quickly refilled with merchandise,” it noted.

“Their awareness of this made people less eager to spend and more willing to save.” Personal net saving rose sharply. “… many families had become rather well stocked with consumers’ durable goods in the period June 30, 1950 to February or March of 1951, and were in a strong position to postpone further purchases.” Tightening of credit controls too may have helped. This explanation seems to have carried on into 1953 as well and explains the low inflation of that year.

What is interesting is that in 1952, six years after the death of Keynes and 12 years since the publication of his How to Pay for the War, central bankers, in seeking to explain inflation or the absence of it, took into account defence expenditure, consumer spending and saving in addition to money supply and credit controls. By 1970 all explanations of inflation centred on a single point: money supply. And Keynesians had no answer to the monetarist claim because by then, the transformation of the “economics of Keynes” into “Keynesian economics” was complete.

We may add in, passing, that expectations cannot explain inflation. One expression of the expectations thesis is that if a central bank signals its intent to run an expansionary monetary policy, then agents will raise prices if they trust the central bank’s pronouncements, whether the central bank expands money supply or not. But this cannot be correct. If one set of agents raises prices of goods and services, then in order for those prices to become effective, another set of agents must have a larger quantity of money (as medium of exchange) to purchase those goods and services at the higher price. And since that money must be real and not merely exist in someone’s expectations, the purchases cannot be carried out unless the central bank increases money supply.