Ben Bernanke on the New Deal

I’ve been rereading some of the essays in Ben Bernanke’s Essays on the Great Depression, which of course is self-recommending.  I thought this passage summed up some relevant truths:

Our [with Martin Parkinson] own view is that the New Deal is better characterized as having "cleared the way" for a natural recovery (for example, by ending deflation and rehabilitating the financial system), rather than as being the engine of recovery itself.

Bernanke notes that there were "remarkably strong" productivity gains throughout much of the 1930s, even though there was no capital deepening.  This is a central puzzle which any account of the New Deal, or New Deal recovery, must incorporate.  These gains seem to span more sectors than could be accounted for by New Deal policy alone, and note that most government interventions, even good ones, don’t bring productivity gains over such a short time horizon and in such a regular and sustained fashion. 

Bernanke does suggest that some of the gains came from forced unionization and "efficiency wage" effects and yes that would credit the New Deal.  But I doubt that is the best hypothesis and of course it contradicts the traditional account of profit-seeking behavior from businesses (why weren’t they paying the higher wages in the first place?).  Rick Szostak’s work suggests that the New Deal saw lots of labor-saving, process innovations, which meant both high productivity gains and pressure on labor markets at the same time.  In my view most of these gains were simply the result of working through the implications of the earlier fundamental breakthroughs of the preceding twenty years.

Whatever is the case (and we genuinely don’t know), these productivity gains are central to the story of New Deal recovery.  Roosevelt may deserve credit for some of them, or for allowing them to proceed, but don’t assume that the New Deal caused such gains just because you see them in the gross data. 

You can find different drafts of the relevant Bernanke-Parkinson paper here, with various forms of gating.


The increase in productivity was due to the depression not to the New Deal. When 20% of the workforce lose their jobs you expect the least productive to be on the street - raising labor productivity among those remaining at work. When thousands of banks, small firms and small farms go under you expect productivity to be higher at the remaining banks, firms and farms because the least productive are selected out. "Crediting" the New Deal for raising productivity by creating unions and throwing the least productive out of work is absurd. Combine all of the above with ongoing exogeneous technological advance, which was fast at the time, and you have a perfectly reasonable explanation.

John, your point is well taken but note that productivity was rising even when employment was rising...


The unemployment rate throughout the 30's was still much higher than it was during the 20's. If I had to make a guess, it looks like the average rate in the 20's was 8% and the average rate during the 30's was 15%. Here's a graph I found from google.

Higher productivity in the 1930's reminds me of France having a higher productivity per hour than America. That's easier to do when the most unproductive workers, such as African immigrants in the Paris suburbs, remain unemployed.

Ben could be correct that the New Deal helped to stop deflation and let an economic recovery take place. GDP did rebound very quickly after FDR took office and it does look like workers became more productive in the 30's. However, this economic recovery was limited to those employable under FDR's union-friendly government.

My biggest fear with an Obama registration is the ending of union secret ballots. If we go back to a card system and see unionization rebound substantially, the effects on employment could be devastating.

Occam's razor?

Maybe productivity gains were simply the fruit of the Efficiency Movement. As for capital, I always thought it did try to take more credit for increases in prosperity than it was entitled to. Increased exchange of information and ideas deserves far more credit than it gets.

re: "Bernanke does suggest that some of the gains came from forced unionization and "efficiency wage" effects and yes that would credit the New Deal... it contradicts the traditional account of profit-seeking behavior from businesses (why weren't they paying the higher wages in the first place?)"

OK, haven't read the piece but I think I can clear up the logic here.

(a) An efficiency wage effect is an automatic by-product of high unemployment. The higher unemployment, the worse the alternative is to any particular job, so workers will work harder.

(b) Forced unionization OFFSETS the efficiency wage effect, because that means it's more difficult for bosses to fire the least productive guy, and prevents "arms races" between workers trying not to be the least productive guy.

(c) Forced unionization might, however, INDIRECTLY create an efficiency wage effect-- by creating high unemployment. It's a conspiracy of insiders against outsiders, and many of the latter end up unemployed, which (see point (a)) creates an efficiency wage effect.

(d) The reason that profit-seeking firms might not have done this on their own is that the optimal rate of unemployment for purposes of creating efficiency wage effects might be lower than the rates that resulted from the stupid macroeconomic policies of the 1930s. (Anyway, it might not be individually rational for business to create the rate of unemployment that business as a whole would want. Employing people has, in effect, the *negative externality* of reducing other workers' incentives to work, but no given entrepreneur will take this negative externality into account.)

It's also worth noting that Europe, during its various periods of post-war high unemployment, has not generally seen such high rates of productivity growth during those times. That is another reason for thinking the productivity effect is not just a residual from a change in labor quantity.

I agree with John. The increases in unemployment seen in regular recessions, such as Tyler points to, are typically accompanied by labor hoarding (firms keep their best workers when they expect the recession to be temporary even if they can't fully employ them). Thus in a regular recession you get a combination of two opposing effects - the most productive people are employed but they are not working at full capacity. In the Great Depression, however, the decline was so large and it lasted so long there was no possibility of labor hoarding - thus those who remained employed were the most productive and they were working at full capacity.

I also think exogenous technological change was high during this period (i.e. for reasons having to do with the growth of science and not directly related to the economy.)

With high unemployment, workers are aware that their positions are not secure and therefore they work harder to prove their worth. I am puzzled about why this effect wasn't observed in postwar Europe though.

Job insecurity will only inspire workers to greater personal productivity if they believe that increased personal productivity is, on the margin, the most efficient way to improve their job security. If those who make hiring and firing decisions aren't paying attention to individual productivity, then job insecurity may only inspire greater attention to workplace politics.

re: "Why does everyone keep chiming in with a reiteration of John's original point even after Tyler has pointed out that 'productivity was rising even when employment was rising...'?"

Both Tyler's objections have some force but are not decisive. The problem is that we're missing a theory of the *dynamics* of productivity as a function of unemployment. Does a jump in unemployment cause a discrete jump in productivity, or an acceleration of productivity growth? That productivity kept rising after employment was rising (but was still low) only refutes John's point if the effects of unemployment on productivity are instantaneous. If high unemployment causes a higher equilibrium level of productivity, but it takes time to get there, Tyler's counter-argument loses its force.

As for the reason that high unemployment in Europe didn't generate high productivity growth, there's an easy explanation for that. Not that it's necessarily right, but it's very simple and entirely consistent with the unemployment-efficiency-wage story. Post-war Europe had HEAVY WELFARE BENEFITS. If large welfare benefits are available, losing your job isn't so bad, and you have less incentive to work hard. The alternative to a job was a lot scarier in 1930s America than in 1980s Europe. Also, think about the dynamics. The huge, fast jump in unemployment in the US in the 1930s is different from the gradual rise in unemployment in Europe, and that has different implications for dynamics.

My knowledge of this period is very modest, but perhaps Tyler's point on productivity rising as employment was rising could be partially explained by taking into account both John's and Nathan Smith's points.

If high unemployment led to the least productive workers being thrown out of work, along with the remaining employed workers becoming more productive out of fear of losing their jobs, then as employment rose again, the re-employed workers would adopt the new high-productivity standards established by the initial high-productivity workers who kept their jobs. This wouldn't raise productivity necessarily but it might slow or stall its decline as more people enter the workforce.

You can also look at material costs. During the Great Depression you had an increase in labor and a fall in the price of commodities. If highly productive workers are given low cost inputs to work with doesn't that look like an increase in productivity.

In Europe, after the war you had a surplus of labor but they also have a shortage of raw materials. Not to mention that war destroys infrastructure.

Next, take the basic monopolistic competition model. If demand falls, you see lower prices and lower output plus you move to the left on the average total cost curve, a region where per unit costs are falling as you increase output. If other factors of production (commodities) are declining at the same time, then the average total cost curve shifts down. Once again we see per unit costs falling. The combination of decreased demand and decreased cost of inputs can lead to a substantial reduction in the cost per unit of output as you start to increase output. Especially if you are trying to return to former economies of scale for the factories you own.

I assume manufacturing tends to have dedicated resources (physical plant etc) that are not easily converted to other uses, so they stay underutilized until production returns to the former optimal output level. And it is possible that the only surviving firms able to rebound are the ones with some advantage in labor, location, technology. political connections, or dumb luck. i.e. firms that are able to keep their ATC curve (or average variable costs for some period) down compared to competitors.

Until you return production to the optimal level for your current factories you don't invest much to increase output. You are using underutilized capacity, which can look like increased productivity. In contrast Europe after the war, was dealing with shortages and a lack of physical capacity.

My amazement that Bernanke says the New Deal cleared the way for recovery was seconded only by Tyler Cowen agreeing with him, and specifically saying "forced unionization" was a "credit" to the New Deal.

People who told me Tyler Cowen was the smartest libertarian on the planet were obviously wrong. He may well be the smartest social democrat.

1) It would help the overwhelmingly abstact and theoretical comments in this thread if a comparative perspective were applied to the four related sides of the debate:

* a strong growth in productivity in the 1930s Great Depression in the US after 1933, and the causes of it: the New Deal or otherwise

* the debate on efficiency wages --- and, with little discussion, its connection to the New Deal

* the surprising lag in private investment: at the trough of the fall in GDP, it had declined to about 25% of its 1929 level. Investment did rise between 1933 and 1937 to about 59% of its 1929 level, only to fall to around 39% in the short-lived 1937-38 recession (with about a 13% dip in GDP), and by the end of 1939 it had risen again to only 49% of its 1929 rate.

--- All this, mind you, even though GDP itself reached 100% of its level in 1936 and, despite the 1937-38 recession, climbed to 120% of its 1929 level in 1940.

* And finally, of course, high persistent unemployment. It fell from 21% in 1933 to 9.1% and then climbed and stayed around 11-12% until WWII.


2) The impact of the New Deal figures in all four of these debates. The generally agreed upon good side of the FDR innovations from 1933 are:

* taking the US off the gold standard (which enabled interest rates to be lowered by the Federal Reserve, no longer having to protect the value of the dollar);

* ending quickly the banking crisis; later, adding FDIC insurance.

* stopping deflation;

*ending efforts to sustain governmental surpluses: FDR and his team moved initially to neutral fiscal policies, then permitted a slight deficit, then --- when the new social security taxes flowed to the federal government in 1937 --- moved back to a slight surplus (a major cause, along with the Federal Reserve tightening monetary policy to fend off inflation, of the 1937-38 recession).

* allowing the US dollar first to float, then a year later fixing it at a much lower level than before 1933

* moving toward free trade in the late 1930s --- though with little response from the British (the Imperial-Commonwealth system in place); Nazi Germany and its economic hegemony over East and Central Europe; the French empire; and the Japanese empire. With the Soviet Union closed off to the capitalist system.


3) The tangled and harmful side of the New Deal: the NRA, NIRA, AAA and other programs that, mistakenly, tried to cartelize industry and later to rationalize and limit competition, keep wages high (how much effect the FDR policies here had is debatable), and SDC. Though short-lived, these policies probably created uncertainty in the private sector, and --- if the work of Cole and Ohanian is sound (it uses real-business cycle theory and constructs a hypothetical trend-line growth in GDP for measuring the effects of the New Deal on GDP and investment growth) --- were positively harmful. The WPA, which employed about 3 million people --- and form the basis of Darby’s 1973 important revisions of the official unemployment statistics (by counting those 3 million as employed) --- was far more positive and kept, among other things, American arts, theater, literature, and photography flourishing.

Here's a later article of Cole’s and Ohanian’s work than the one Tyler links to:


4) Enter my key point here --- which will be brief, but nonetheless add some substantively weight, hopefully, to the discussion in this thread. (It will be useful if you look at the various trends in GDP during the Great Depression in the countries discussed here: a good chart is found at this URL:

4b) Start with Canada.

If you look at Canada, the Great Depression hit it even harder than it did in the US: higher unemployment, higher GDP drop. And recovery was slow. But --- it was the opposite recovery compared to the US's: employment recovered much quicker there, but productivity growth languished. What’s more, private investment fell even further and faster than in the US --- to 15% of its 1929 level --- and it rose finally by 1940 to about the US’s recovery-level: 52% vs. about the same for the US.

The conclusion: there was no New Deal in Canada, and yet its productivity growth --- unlike the US’s --- never recovered nearly to the level of 1929. Even if the Canadian economy was heavily dependent on commodity exports --- and especially dependent on the US economy’s recovery --- it appears that the absence of a New Deal in Canada did not facilitate any bursts in productivity growth there.


5) Germany’s Recovery.

To be brief, it was fast. It rose quickly from a trough only slightly higher than the US’s in 1933 when Hitler took power (January 1933, with Roosevelt coming to office two months later). And its speedy recovery was due to huge government spending --- 50% more per year than government income from taxes, with most of it in transportation and private investment (in arms not least), though armaments-spending itself was about a quarter of overall government spending until 1936 . . . after which more and more of the German economy was devoted to preparing for war.

Productivity increased throughout the decade between 1929 and 1938 at about 1.3% a year. Nothing spectacular, mind you --- but a steady increase. And German statist intervention into its increasing militarized economy --- along with a big welfare state that built on the Bismarckian innovations --- did not impede the growth of productivity. Score another likely point for the New Deal role, possibly anyway, for its stimulating effect on productivity in the US: directly or indirectly.


6) Britain’s recovery. It started slowly, then picked up considerably from a smaller downfall in GDP after 1929. There was no New Deal, mainly a boost to growth after the pound-sterling was taken off the gold-standard, devalued, and exports flourished as a result of the Imperial-Commonwealth mercantile system.

Productivity growth was strong throughout the 1930s. In effect, British industry --- which had lagged behind German and American industry in the previous decades --- now regrouped, modernized, and was rationalized by the National Governments (Conservatives, Liberals, some renegade Labourites) that ruled Britain in the 1930s. In effect, they facilitated the decimation of the traditionally small and mid-size British firms; encouraged in a variety of ways to regroup them into giant firms more like their American and German counterparts; and encouraged too a steady increase in productivity.

Britain, too --- though avoiding New Deal programs like the WPA --- had a more advanced welfare state than the US, and so its increases in productivity and modernized industry were not hindered, but (like their American counterparts), seemed to be aided by governmental programs. British economic growth, note, never fell as much as American or German after 1929, but it had languished throughout the 1920s, and British recovery from the Depression was further pushed along by rearmament in the late 1930s.


7) Japan, finally: Very briefly now, though the Great Depression caused widespread misery among the small peasant farmers, Japan’s GDP did not fall very much. It quickly recovered, not least owing to large deficit spending and a big devaluation of the Yen . . . followed by an increasing state-role in preparing Japan for world war.

By 1941, when war in the Pacific engulfed no longer just the Chinese, but the Americans, British, French, and Dutch and their colonial peoples, Japan’s economy --- under the pressures of its command militarized and cartelized economy --- had been transformed from light industry and agricultural exports to heavy industry prepared for war.


8) The outcome?

If a comparative perspective is applied to the Great Depression, it appears that in all the rich countries surveyed other than Canada --- which was still largely an agrarian and commodity-oriented economy --- a large state role emerged . . . far larger, in fact, in Germany and Japan than in the US, and with Britain’s National Government pursuing policies not much different from the New Deal except for more orthodox fiscal policies. And everywhere in those countries save Canada, there was a major increase in industrial production and productivity . . . aided by rapid rearmament in Germany and Japan, somewhat less so in Britain, and hardly any until 1939 in the US (when naval and aircraft production was increased even before we entered the war).


Michael Gordon, AKA, the buggy professor

PS. The discussion in this thread of efficiency wages is, on the whole, way too simple. There are numerous variants of the theory, its nature, and its impact on unemployment. The most recent work reflects the behavioral revolution carried out by a small but increasing number of behavior economists who actually study the nature of human behavior in concrete situations, using work derived from psychology and social psychology. The summary and analysis of these variants at Wikipedia is admirable and easy to follow:

Pople who think unionization results in a more productive workforce, I'm guessing aren't working in one.

The notion of New Deal recovery is not only not confuted by Tyler, but is even vaguely affirmed.

Bad Tyler!

Big productivity increases in the 1930s were related in part to no net immigration during the 1930s.

Very grateful to a bunch of much better skills. I look forward to reading more of the future of the subject. Keep the good work.thanks.

Comments for this post are closed