by Mike Kimel
Unemployment During the New Deal
Cross posted with the Presimetrics Blog.
This is not the post the post I was planning to write this weekend, but frankly every time I post something that touches on FDR, people get ahold of me with the same canard: unemployment in the U.S. was somewhere around 20% all through the 1930s, which is a sign that the New Deal failed. It’s something repeated at length by the Glenn Becks of this world, but also by a lot of supposed scholars.
Now, there’s a reason I haven’t written this post in the past and wasn’t planning to do it any time soon; it’s been done before, and well enough that there isn’t much for me to add. I think the first time it was done was back in 1975 as a working paper Michael Darby (sorry this is behind a paywall – as is the later version published in the JPE). In blogs, Eric Rauchway at The Edge of the American West covered the topic very well (distilling down some of the post-Darby work in the process) a couple of years ago. I’ll be drawing extensively from the Rauchway piece in this post, including stealing both of the graphs that appear below.
So to begin, as Rauchway noted, there was a time that everyone knew that unemployment in the 1930s looked like this:
The reason everyone knew that was because those were the figures Stanley Lebergott published in
Enter Darby. He reached the conclusion that people who are building infrastructure (roads, dams, sidewalks) and getting a regular paycheck for that labor should not be considered unemployed. (Go figure!) And if you’re inclined to argue that point, consider this – what exactly do you think the unemployment figures would be for China today if you counted everyone who worked for a company controlled by the PLA or some other government agency as unemployed? What would the unemployment rate be in Singapore if anyone working for a government owned company was considered unemployed? And how realistically do you think those figures would reflect the situation in China or Singapore?
So Rauchway includes this second graph, showing both the Lebergott figures for unemployment and Darby’s corrected version.
Big difference, eh? Of course, unemployment was relatively high all the way through the start of World War 2. The Great Depression caused a huge hit on the economy, and the Dust Bowl didn’t help. It is not a surprise that it took six years for the unemployment to drop in half, and that even that figure was elevated. But the reason the unemployment rate was dropping from its high of 24% (picture how bad the economy has to be for unemployment to get that high) was that some serious real economic growth had set in. In five of the eight years from ’33 to ’40, real GDP grew by more than 8% a year! In terms that the folks who criticize FDR’s performance would understand, that’s about a percent faster than Reagan pulled off during his best year.
So while the unemployment rate remained high from ’33 to ’40 (the private sector wasn’t hiring the unemployed, much less competing with the government for the services of those who worked for the WPA), the thing we should be looking at is the rate at which the unemployment rate changed. Otherwise, we have to conclude that GW did a better job on the economy than any other President, never mind the lukewarm growth rate intro leading up to the Great Recession finale. After all, the peak in real GDP per capita (so far) came during his term.