Scott Sumner  

Milton Friedman wouldn't have been confused

Sunday Morning Audio: Henderso... What a Wonderful World!...

Stephen Kirchner pointed me to a very infuriating Financial Times article on inflation. Here is the title:

Nobody seems to know why there's no US inflation
Nobody? Seriously?

How about market monetarists who point to the very slow growth of NGDP, caused by a contractionary monetary policy?

At times it seems we've gone back to the 1970s, as completely discredited theories of cost-push inflation are being revived:

A five-month-long string of weak figures this year has prompted some Fed policymakers, including Bill Dudley of the New York Fed and Rob Kaplan of the Dallas Fed, to start asking searching questions. Are globalisation and technological advances restraining inflation at a time when a cyclical pick-up might otherwise drive it higher?

"These two forces are colliding," Mr Kaplan told the Financial Times last month. "The cyclical forces we have understood historically; the structural forces are somewhat new, particularly technology-enabled disruption."

Some policymakers cite the increased ease with which shoppers can compare prices on the internet and the impact these changes have on brand loyalty and pricing power. Amazon is entering grocery retail via Whole Foods, while the hotel sector is being overturned by Airbnb.

If inflation were held down by rising aggregate supply, then you'd expect rapid GDP growth. Instead we have the opposite.

Fortunately some media outlets seem to get it. Here's Caroline Baum at MarketWatch:

Almost every discussion on this subject begins with a statement of fact that the tight labor market, as evidenced by the 4.4% unemployment rate, should be lifting wages and prices.

Note the order: wages and prices. A tight labor market leads to higher wages, which lead to higher prices. This is one of those myths that never dies: cost-push inflation. Milton Friedman was adamant that both prices and costs rise in response to an increase in aggregate demand, which is a function of the Fed's money creation. Wage and price increases are a reflection of inflation, not a cause of it.

As to the relationship between prices and wages, they generally move together. But prices lead wages, not the other way around. You can read about the relationship in academic literature, or you can think about how businesses operate.

Let's start with a small-business owner whose company produces widgets. He begins to see a pickup in sales. Pretty soon, his products are flying off the shelf, and he can't accommodate the increased demand.

What does he do? A rational businessman raises his prices to allocate the limited supply. If he still can't satisfy the demand for widgets, he may try to increase output using his existing staff, paying overtime if necessary. If he still can't meet his customers' demand for widgets at the higher price, he will most likely look to add staff.

In the mythical cost-push-inflation world, a businessman responds to increased demand by paying a higher wage to attract additional workers -- and then tries to raise prices to preserve his profit margins.

Which of those examples describes how businesses operate? Good. Let's move on.

It's kind of scary when top Fed officials have forgotten that inflation is a monetary phenomenon.

I was also puzzled by this comment in the FT:

Despite a recovery that is now the third-longest on record, America is trapped not in a 1970s-style, double-digit inflationary upsurge, but a slow-inflation quandary.
I suppose it seems like a "trap" if you don't believe that monetary policy determines the inflation rate. Fed officials look everywhere for the culprit behind "lowflation", when all they need to do is look in the mirror.

PS. I've continued my discussion of modern art over at TheMoneyIllusion.

Comments and Sharing

COMMENTS (12 to date)
Effem writes:

When I'm told we understand inflation but have no idea what drives velocity, it doesn't give me confidence that we understand the system we've created.

Scott Sumner writes:

Effem, Why do you say we have no idea what drives velocity? Surely nominal interest rates (and interest on reserves) play a major role. There are many studies of velocity that do a pretty good job of explaining it.

Benoit Essiambre writes:

Wait even with "demand pull" inflation, as I understand it, can't wages lead price increases?

>"Let's start with a small-business owner whose company produces widgets. He begins to see a pickup in sales."

This begs the question what monetary mechanism would cause this pickup in sales?

Central banks interest rates moves don't affect consumer credit much. Would you change your purchase behavior if your credit card APR went from 19% to 18.5%?

For larger investment projects, factories, large equipment, a half percentage point can make a significant difference in the cost of capital equation about whether a project is worth doing or not. Knowing that prices are going to rise faster also influences decisions and might cause businesses to hire more people at higher wages now to build capacity for the future.

My intuition is that the main causal chain of monetary stimulus goes something like:
1. Easier financing and expected higher future prices leads to ->
2. More investment and new capital creation ->
3. Which requires more employees ->
4. Which raises wages ->
5. When these wages rise people try to buy more stuff ->
6. Which cause prices to go up (self-fulfilling 1.).

I'm not saying there aren't other channels such as consumer demand going up a bit but they don't seem to me like they would be as important. Wages and intermediate goods prices should go up before final goods prices. Businesses shouldn't immediately pass the full cost increase from wages and intermediate goods to the consumer because lower returns will be partially offset by the easier financing and expected growth in nominal revenues.

Jeff writes:

The Fed is a bureaucracy. It will never admit responsibility for anything.

Benoit Essiambre writes:

For example think of housing, which embodies both a consumption and a investment component. Under monetary easing, rents or mortgage payments, which represent more the consumption side, may actually go down initially. At the same time, demand for housing goes up and house builders get more jobs and wage increases. The price of houses may go up immediately but that largely represents an investment asset price increase.

bill writes:

Tangent. How stupid is the notion of a low inflation quandary? Oh no, unemployment has fallen to 1% and we still don't have 2% inflation. Woe is me.

Hazel Meade writes:

This seems like the sort of thing that would demand some empirical research - as in go ask retailers why they aren't increasing prices.

It might be that they face more competition today, or that demand has not really kept pace with increasing production.

I wonder if perhaps modern information technology allows producers to track or predict demand so much more rapidly and accurately that they can get goods on the shelves faster demand can drive up prices. If supply is perfectly tracking demand, wouldn't that result in 0% inflation?

Mark writes:


I think then we would see the effect of this in GDP growth, which I guess we don't, as Scott mentions.

Scott Sumner writes:

Benoit, Monetary policy does not affect spending via interest rates, it works through the hot potato effect. Even if interest rates did not exist, an expansionary monetary policy would boost AD.''

If there is an expansionary monetary policy, the price of housing will rise before the wages of workers. These higher prices and sticky wages will combine to produce more housing construction. Interest rates have nothing to do with it.

Jeff, When he was at the Fed, Bernanke admitted that the Fed was to blame for both the Great Depression and the Great Inflation. They don't tend to take blame for recent mistakes, although the Minneapolis Fed President just did so.

Hazel, Supply can equal demand at any inflation rate, even 1000%.

Brian Donohue writes:


Sometimes progress is too much to ask for, and God's work entails merely trying to preserve hard-won knowledge. Cheers!

Hazel Meade writes:

Scott, I think you are misunderstanding me. Obviously if prices rise, demand will drop to equal supply, but the direction that prices move depends somewhat on what order things happen in. If demand increases before supply, prices will rise - and bring demand back into balance. But if supply comes first, outpacing demand, prices will fall, bringing up demand. So the inflation rate that you actually get should depend somewhat on whether supply is leading or lagging demand. Right?

Scott Sumner writes:

Hazel, Again, you can have any inflation rate with supply equalling demand.

Comments for this entry have been closed
Return to top