Arnold Kling  

The Austerity of 1945-1947, Again

Economic Creationism... Paul Samuelson's Prediction fo...

In 1991, Richard K. Vedder and Lowell Gallaway wrote,

The smooth transition to peace was accomplished despite the existence of a fiscal policy that was the very antithesis of Keynesian economic prescriptions to deal with falling aggregate demand. The most dramatically contractionary fiscal policy in modern American history failed to materially alter the pace of economic activity.

Thanks to Alex Tabarrok for forwarding me the link.

The authors look at real GDP, which I did not show you in my last post. Here it is (as of today, that is; Vedder and Gallaway were looking at different numbers):

Nominal GDP223.0244.1
Real GDP2012.41776.1
Nominal Government Spending93.036.3
Real Government Spending1402.2409.5

Some points stand out:

1. Real GDP reportedly fell by more than 10 percent, even though nominal GDP increased.
2. In 1945, government spending was 40 percent of GDP in nominal terms, and almost 70 percent (!) of GDP in real terms.

My reading is that the attempt to inflation-adjust numbers from the 1940's to try for comparability with today introduces tremendous noise in the yearly movements in estimated real magnitudes. I believe that this phenomenon is known as base-year drift. As you change the base year for calculating real GDP, you get anomalies in years that are not close to the base year. Vedder and Galloway put it,

Between 1960 and 1990, government economists approximately doubled their estimate of the inflation occurring from 1944 to 1947

I see that not as a conscious change in estimated inflation but instead as a mechanical result of base-year drift. If what you want to do is compare the standard of living in 1945 to the standard of living today, then you have no choice but to use real GDP. However, if what you want to do is estimate the trajectory of the economy from 1945 to 1947, I think you are better off using nominal GDP. What you miss in terms of inflation adjustment is way less than the error that you introduce due to base-year drift.

Another point, which Vedder and Galloway also make, is that with government spending accounting for so much of GDP (even forgetting the suspect estimates for real GDP), one really has to worry about whether the estimated value of government purchases is correct. Private-sector spending is subject to market discipline, which leads people to pay no more for something than the opportunity cost of the resources used to provide it. With government, we have no such assurance.

In the end, I do not know how to value what the government bought with its credit card in the 1940's ("Defeating Hitler: priceless"). Did government pay draftees the opportunity cost of their time? I'm guessing not, which would mean that government GDP was under-estimated. On the other hand, government probably overpaid for other resources.

In a comment on my previous post, Dan Klein points to Robert Higgs, who takes the view that wartime GDP was overstated and postwar growth was understated. Higgs has thought about this much more than I have, but I would like to poke around in the history myself before taking a strong stand.

Comments and Sharing

CATEGORIES: Economic History

COMMENTS (9 to date)
Arthur_500 writes:

One aspect of this situation may be important - pent up demand and savings.
What I have read is that there was a higher level of savings during this period of time simply because many goods were not available. After several years of austerity the late 1940's were a period of playing consumer catchup.
Additionally, there was little competition from Europe and nothing from Asia. We exported everything possible to a devestated Europe and a great deal to Asia also.
I agree with the fundamental idea but I must take into consideration the unique aspects of that recovery period.

Foobarista writes:

When have Keynesian strategies ever provably "worked"?

Noah Yetter writes:

"Defeating Hitler" aside, the true value of government spending on war goods isn't even zero, it's negative. What's a tank worth? A battleship? A bomb? It's no wonder at all that the economy boomed like crazy as soon as we stopped spending so much money on blowing stuff up and killing each other (not to mention the end or phasing out of rationing, wage & price controls, etc.).

Doc Merlin writes:

The natural answer that the Keynesians will give is that this only applies to military spending. How do we look at other spending to counter that?

Jeremy, Alabama writes:

The point about it being military spending is how little this distorts everything else. The government competes for inputs, but the output does not distort markets, and meanwhile the government is distracted from say housing policy, banking policy, agricultural policy, Wall Street, there is no problem with full employment.

Also, all kinds of innovations are rushed to market. I wonder what the economic value of this is?

When the government is focused on blowing up other people instead of micro-managing our economy, we are much better off.

Travers writes:

UGH! The Keynesian prescription is for _liquidity traps_. The war broke the economy out of the liquidity trap, so Keynesian fiscal spending WAS NO LONGER REQUIRED!

I keep hearing "Keynes said spend and spend like there's no tomorrow because he said, 'We're all dead in the long run'". Misquoting and misunderstanding Keynes seems to be all the rage right now.

'We're all dead in the long run' meant "If you know you are going to die in the end, does that stop you from making good choices right now for your life?"

Zach writes:

[Comment removed for supplying false email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Benny Lava writes:

Question: What percentage of spending reduction was in the military, and what percentage of spending reduction was in social programs? And is this important when comparing 1947 to 2010?

Chris writes:

Technical points: The comment about government being 70% of real GDP also suffers from base-year drift. Measured in 1945 prices, government was 40 percent of GDP. If you read the fine print on the BEA's release table it says, "Warning: Do not add chained dollars." The same thing goes for subtracting or multiplying.

One nice feature of using chain-weighted real GDP (the currently used measure) is that you can actually compare real GDP in 1945 and 1947 without having to take prices in the year 2005 into account. Yes, output was a lot lower in 1947 than in 1945.

The amazing thing about that episode is what happened to labor markets. Employment, like output, crashed like a rock but unemployment rose only moderately. A large number of people left the labor force at the end of the war. This really was a weird episode.

Comments for this entry have been closed
Return to top