Share |

A Response to Paul Krugman’s Depiction of 1937

Amity Shlaes | 9/7/10 |

Click here for a printer-friendly version of this article

Paul Krugman draws an analogy between United States in 1937-1940 in his Labor Day column.

A summary: Dr. Krugman depicts 1937 as a year when FDR tried to slash the deficit. Dr. Krugman suggests that stimulus, especially fiscal stimulus, ought to have been applied. Instead, the Roosevelt administration ``pulled back stimulus too soon.” Voters sided with caution as well, Dr. Krugman notes. A Gallup poll of 1938 that more than half of voters wanted no stimulus and more than half backed tax cuts.  But they were wrong, says Dr. Krugman. The economy crashed. Only when the war came, and the federal government borrowed massively to spend, Dr. Krugman suggests, did the economy recover.

Dr. Krugman’s argument is a strange one, worth retracing year by year.  For a systematic review of that period does not suggest that massive spending at the current point in U.S. history makes sense.

Dr. Krugman starts out right: it is true after the election of 1936, policy changes were made that yielded an effective tightening of money. Some of the changes were intended and some were policy half-accident.  The first of these was the staged doubling of bank reserve requirements, which had the effect of pulling money out of the economy. A second factor was Social Security payments; the Social Security Act became law in 1935, but 1937 was the first year government collected, taking spending cash from consumers. In 1936 and 1937 large bonuses were given to veterans; no such bonuses were scheduled for 1938.  For 1937, total government spending as a share of GDP was 13.2%, down from the 15.4% of the preceding year.

But Dr. Krugman fails to tell of other policy changes, some of which were also made as far back as 1935. These other changes also dragged down the economy in 1937 and 1938.

The first of these was change in wage policy. In 1935, Congress had passed the Wagner Act, the basis for modern labor law. The Wagner Act of the 1930s was a tiger compared to the kitty cat we call labor policy today. John L. Lewis of the CIO and others shortly used their new power to launch sit-down actions against steel companies and the automakers. The phrase “sit down” sounds vague to us now, but it is important to remember that what it meant was that the workers might occupy employers’ private property. Violent strike actions accompanied the sit-downs. Employers, squeezed by unions as never before, lost days of production. Whenever they could, companies raised wages to appease the unions and Washington. This was the period when U.S. Steel and all auto producers but Ford were unionized. It seemed organized labor had no limits.

The effect of the Wagner Act was to push up the real wage and make employers hesitant to hire or rehire. Scholars Lee Ohanian and Harold Cole, who have studied the period extensively, reckon that New Deal wages were 20% higher than they would have been without New Deal policy. (For a fuller explanation see their Milken Review article on this topic.)

What’s just as important is that people knew labor policy was making finding a job harder. The song “Nice Work if You Can Get It” became popular in these later 1930s years, in part because it adequately captured the divide between those who had union jobs or government jobs and those who could not find work. The whitewashing of Keynesianism is what has kept us from knowing about it all these years.

Taxes are also overlooked by Dr. Krugman. In 1935 likewise, Congress and Roosevelt had raised income and other taxes. That shift discouraged already enervated businesses. So did a frightening new tax passed in 1936, the creepy undistributed profits tax. The tax, which proposed to eat away at companies’ essence if they did not disgorge their money, was Roosevelt’s revenge on business; if business would not spend to spur the economy, then he would force the money out of them.  There had been a period in the early New Deal when Roosevelt’s policy seemed to favor big business – the 1933 National Industrial Recovery Act’s codes advantaged larger companies and hurt smaller ones. Now however, FDR turned on business, blaming corporations for the downturn.  His prosecutors, Robert Jackson and Thurman Arnold, systematically went after big companies such as Alcoa.

A third factor unmentioned by Dr. Krugman is political uncertainty. Roosevelt didn’t merely win a historic victory in 1936, taking 46 of 48 states. He committed to translating that electoral power to policy.  The President spoke in his 1937 inaugural of “fashioning an instrument of unimagined power” with government, a phrase so bold no modern president, Democratic or Republican, would dare to read if he saw it on a Teleprompter. Net private domestic investment went negative as companies sat on their hands rather than invest in this environment.

It is the combo of all these factors that caused the Depression within the Depression of 1937 and 1938.  My book, The Forgotten Man, stresses the taxes, labor costs and uncertainty. Since writing it, I’ve come to believe more emphasis ought be placed on the labor price.

There’s an additional point where Dr. Krugman’s narrative doesn’t make sense. From 1938, Dr. Krugman leaps ahead to World War II, to suggest that it ended the Depression. In doing so he omits to describe the recovery, incomplete but strong, of the second half of 1938 and 1939. In those years unemployment dropped a smidgen, but was still achingly high, due to the aforementioned labor policy.  But industrial production started to strengthen. After shrinking 3.4% in 1938, real GDP grew 8.1% in 1939 and 8.9% in 1940, before Pearl Harbor.

What might have caused this upturn?  There were monetary factors. Gold flowed to the United States as the crisis in Europe worsened. Since the U.S. was on a form of the gold standard this meant an effective loosening of money. The inflow was not sterilized, but whether the heads at Treasury and the Fed understood the full import of the non-sterilizing decision is not clear. In a 1991 National Bureau of Economic Research working paper Christina Romer, who just retired as chair of the Council of Economic Advisors, stresses the importance of these gold inflows for recovery. You can also argue there were spending factors – the government grew as a share of the economy, although not in the massive fashion that Dr. Krugman prescribes.  

But other factors in the 1938-1939 upturn were taxes and the diminishment of that uncertainty. In 1938, the political tide began to turn against Roosevelt. In the spring of 1938, lawmakers gutted his undistributed profits tax and dropped the graduated corporate income tax in spite of Roosevelt’s objections. Their bill became law without his signature. In the midterm Congressional elections of 1938, Democrats lost 81 seats, not enough to lose control of the House, but enough to chasten them. Bored and frustrated with the New Deal, FDR turned to foreign policy, an area to which he was better suited in any case. The Supreme Court ruled against sit down strikes, limiting the scope of union power. Washington’s war on business was suspended, in part because the president knew he would now need the same industrial giants he had prosecuted if he was going to arm the U.S. and Britain. Treasury Secretary Henry Morgenthau, who had personally sicced attorneys on his predecessor Andrew Mellon, now put a sign on his desk to signal friendship for business. The sign read “Does it contribute to recovery?” The policy mix of the late 1930s was far from ideal, but the direction was enough to cheer everyone.

The real question is not how war spending ended the Depression. It is why the Depression lasted so long. Spending, in any case, didn’t have much to do with the Depression’s end. As Dr. Romer herself summed up in that 1991 paper, “it is hard to argue that changes in government spending caused by the war were a major factor in the recovery. The recovery was nearly complete before the war had a noticeable fiscal impact.”

It’s not clear whether Dr. Romer, Dr. Ohanian or Dr. Cole has analyzed the later 1930s precisely right. The 1930s were not a monocausal period. Still, what is clear is that evidence of the late 1930s does not support Dr. Krugman’s case for unprecedented outlay today.

Amity Shlaes is a senior fellow at the Council on Foreign Relations.

e21 Projects & Partnerships