Since others agreed in the prior post that the depression chart would be better in real dollars rather than nominal dollars I though I would take this opportunity to post this chart.
In business cycle analysis if you use NBER official terminology the cyle should be divided into three phases. The first is recession from the peak in economic activity to the bottom. The second is recovery from the bottom in economic activity until the the previous peak is reached.
The third phase is the expansion from when the prior peak is surpassed until the next peak.
The second part of this chart is a basic forecasting rule I developed years ago that recessions– recoveries are symmetrical. That is the recovery period is almost always the same length as the recession. If the economy falls for 4 quarters it should take it 4 quarters to regain the prior peak and on average that is what happened in post WW II recession. In other words, mild recessions have weak recoveries and deep recessions have strong recoveries. You can see that in the dashed line in the chart above. Also note that the dashed line is quarterly data while the solid line showing the 1930s is annual data. The BEA only published annual data for the 1930s.
If you look at the 1929 to 1936 period it is obvious that this rule worked well. The downturn was four years from 1929 to 1933 and the recovery was three years from 1933 to 1936 when real GDP surpassed the 1929 peak. According to the NBER the bottom was in March, 1933, the very month FDR took office. If we had quarterly data it would look even more like the 1929-36 recession–recovery was symmetrical ,just like the post WW II recession — recoveries.
We have a massive sub-industry in the economics profession claiming that the depression recovery was weak. But this chart actually suggest that the 1933-36 recovery was very strong by post WW II standards because after falling for four years the economy only took three years to surpass the 1929 peak. In the chart I put the bottom in 1933 and in post WW II recession at the same point on the scale so the 1929 peak is placed earlier then the post-war peaks. So this chart says that all the analysis blaming Roosevelt and the new deal for a weak recovery is just plain incorrect. The recovery was not weak in comparison to other recoveries.
Where you get the difference is in 1937-38. In a normal economic cycle this would be the expansion when growth is normally above trend and in the 1930s the economy clearly had the capacity to grow at 8%-9% rates in 1937-38 as it did in 1939 and 1940.
But in 1937-38 the US had a recession rather than growing at 8% to 9%. this difference is why the good economists are correct in saying the recovery in the depression was weak. Of course, this raises the question of why there was a recession in 1937-38 and that is open to debate.
I personally blame the Fed just as the Fed was the single most important reason for the 1929-33 collapse. In late 1936 the Fed became very worried that the economy was too strong and would generate inflation. So they raised reserve requirements, a tightening move that caused the 1937-38 recession just as such tightening moves generally cause recessions. If this analysis is correct it implies that the analysis by Amy Shlaes and other blaming the New Deal are completely without any basis.