Home / General / People Must Live By Work Book Club: Week 4 (Chapter 3)

People Must Live By Work Book Club: Week 4 (Chapter 3)


Last week, the online book club of my new book (available to the reading public here and here) discussed how the Works Progress Administration triumphed over the Public Works Administration’s traditional public works strategy in a battle of the acronyms. (Incidentally, I’m throwing a book party next Friday.)

This week, I’ll examine the track record of the WPA in its 1935-1943 career, and make the somewhat controversial argument that the jobs programs of the New Deal ended the Great Depression before WWII.


The third chapter of People Must Live By Work has two distinct and related halves. In the first half, I take something of a break from the narrative account of the development of direct job creation policy to make a larger argument that the New Deal successfully combatted depression levels of mass unemployment. In the second half, I examine how the WPA used politics and ideas to accomplish this.

Argument #1: The New Deal ended the Great Depression

Talking about when the Great Depression “ended” is difficult, because we have different ideas in our heads about what a depression is, especially because we don’t have a formal definition for it the same way that we do for recessions (thanks to the National Bureau of Economic Research). If by depression we mean a severe contraction of the economy, we might well say that the Great Depression ended in 1933, because the U.S economy stopped shrinking that year and started growing quite rapidly. But when we talk about the Great Depression ending in 1941, what we’re really talking about is not GDP statistics but the experience of mass unemployment.

And this is where we get to the tricky problem that modern unemployment data starts after World War II, thanks to the many innovations in statistical methods that came out of the total war effort. The data we do have for the 1930s is only a set of estimates that was commissioned by the Labor Department in the late 40s, and put together by an economist named Stanley Lebergott.

Unfortunately, Lebergott “deliberately ignored the public-sector jobs created by the New Deal and instead counted the people who held them as still unemployed. He drew an analogy between WPA workers and Nazi concentration camp inmates or Soviet gulag prisoners who also “worked.”” As I discuss in the book:

As social science, this is problematic, blurring two categories of people who experienced the 1930s in radically different terms. It separates people who got paychecks from the private sector or “regular” public sector workers from those whose pay packets owed their existence to New Deal programs, despite the fact that both groups went to work, got paid, spent their money on consumer goods, and generally felt themselves to be productive. Accordingly, this is a questionable distinction as a matter of economics. Arguably, it is even worse as policy analysis. With this omission, Lebergott ignored the primary method by which the Roosevelt Administration attempted to reduce unemployment, making it much harder for scholars to gauge the true effectiveness of the New Deal’s economic policy.

After thirty years of historians and economists passing on the false impression that the New Deal had failed to make much of a dent, with double-digit unemployment continuing throughout the 1930s, economist Michael Darby produced a new set of estimates that “found” the three-and-a-half-million workers whom Lebergott had “mislaid”:

Darby’s data is the foundation of this first argument.  It gives us a different set of unemployment rates and thus a different narrative of the New Deal’s economic impact, with unemployment dropping below the double-digit line twice to sit at 6% prior to the U.S entering World War II. But even more importantly, it breaks down private sector workers, direct job creation workers, and the unemployed, allowing us to see the specific contributions of the New Deal to the overall decline in unemployment rates. Throughout the 1930s, New Deal job programs created more new jobs per year than the private sector, producing a public-sector-led recovery. The scale of the New Deal’s efforts is especially staggering when considered in relation to the American workforce as a whole; across FDR’s first two terms, direct job creation workers made up an average of 6.43% of the national workforce (for sake of comparison, the current Federal government makes up 5.7% of the whole).

So where does this leave us with our original question? If a depression is measured by a sharp contraction in economic growth, then the Depression had ended well before WWII, with the mid-to-late 1930s seeing an average GDP growth rate of 5.8%. If a depression is measured by unemployment at or near double-digits, and the U.S entered WWII with an unemployment rate of 6% (a rate which is basically the average unemployment rate from 1948-2014), then the depression was over before the war. And if, as we’ve demonstrated above, the New Deal’s jobs programs made an over-sized contribution to recovery of unemployemtn, they deserve the credit for ending the Great Depression.

Argument #2: The WPA accomplished this with both politics and ideas

So how the hell did the WPA pull this off? Well, as we saw last week, the experts and administrators of the WPA were adept at both political and intellectual coalition-building and applied the same skills they had used to achieve dominance over the PWA to future fights within Congress and the Roosevelt Administration.

In Congress, Harry Hopkins, Aubrey Williams, Jacob Baker, Corrington Gill, and Alan Johnstone executed a nimble two-step, cultivating a “pro-spender” caucus in Congress which reliably gave them solid majorities for appropriations to maintain and expand the WPA between 1935-1938. At the same time, these WPA bureaucrats pursued an ambitious strategy for facing down conservative opposition after 1937 in which they would propose ambitious expansions of a million workers above their current levels, or creating a permanent cabinet department for direct job creation, or abolishing the needs test and increasing wages beyond “security” levels. By pushing the edge of the debate leftwards, they made more modest expansion the moderate alternative to conservative calls for cuts, and so WPA workforces continued to grow or hold steady, long after other New Deal programs ran aground.

Simultaneously, these activists recruited new allies within FDR’s economic advisors through a meeting of the minds on the brand-new field of Keynesian economics. To American Keynesians like Lachlan Currie and Leon Henderson, the WPA represented the fastest way to boost consumer spending and renew economic recovery; to the WPA’s experts, Keynesian economics provided a new language of purchasing power, marginal propensity to consume, and multiplier effects to justify direct job creation. Combined with the allies they had already made during their time on the Committee of Economic Security, Hopkins et al. were able to reverse FDR’s decision to balance the budget and commit the Roosevelt Administration to deficit-spending through the WPA.

Finally, I look at how the WPA assessed their own work through a combination of quantitative and qualitative surveys: in their Inventory, the WPA used the photo-documentary techniques of the Farm Security Administration (where Dorothea Lange did her famous photos of Dust Bowl refugees and southern sharecroppers) to make a bold modernist statement about the products of their labor. In their Final Report, which relied more on statistics, the experts of the WPA assessed their work as a successful experiment, “an advance over traditional poor-law methods of providing relief,”and provided their report as a how-to manual for the next generation of Federal bureaucrats who would carry out the policies required by the “acceptance by the Federal Government of a portion of the responsibility for assistance in the provision of work and wages in a time of mass unemployment.”

Thus, they confidently looked to the future that the end of the war would bring.


For those of you who have followed the work of Eric Rauchway, both in his published academic work and at the excellent and much-lamented Edge of the American West, the graph I presented above is quite familiar. When I was beginning my research into the history of direct job creation policy, Rauchway’s work on Lebergott and Darby and especially his excellent fisking of Amity Shlaes’ The Forgotten Man, was incredibly influential. Here was an established scholar who had pulled back the curtain to explain why academic opinion on the New Deal seemed to diverge so strongly from what I was learning about the WPA and other job programs, and who pointed directly from that theoretical mismatch to its impact on our current political debate.

At the same time, Rauchway’s work confirmed me in a source of irritation I had with liberal commentators in the wake of the Great Recession of 2007-8, especially during the debates over the stimulus bill. Conservatives from George Will on down would confidently assert a consensus that the New Deal failed and thus stimulus couldn’t work, and the most that liberals like Paul Krugman would do to rebut was that the New Deal failed because it didn’t go far enough.

Needless to say, I found this response lacking in a doorstop-pitch sense; why concede the point at all, why not push straight back by stating forthrightly that the New Deal had ended the Great Depression and should be imitated now?


Unsurprisingly given the arguments I’m making, Chapter 3 is probably the most controversial chapter in my book. The article it started as was repeatedly rejected at various journals, in no small part due to economic historians and economists who didn’t much appreciate a political historian playing in their backyard. However, I think the intervention is a critical one, and after revisions I think the point stands.

Much of the dominant work by economists on the Great Depression is profoundly shaped by a methodological approach that focuses on the relation between federal deficit levels and GDP growth rates; following Okun’s Law that a 1% increase in the unemployment rate leads to a 2% drop in GDP, major thinkers like Ben Bernanke, Christina Romer, Larry Summers, and Brad DeLong tend to conclude that New Deal deficits weren’t big enough to produce the recovery that was observed, and so point to other factors, whether that’s monetary inflation, banking reforms, or going off the gold standard.

The problem with this approach is that, unlike conventional fiscal stimulus, direct job creation doesn’t rely solely on fiscal multiplier effects to achieve its goals; in addition to putting more money in workers’ pockets, direct job creation has a direct effect on unemployment, as workers move from the unemployment line to the job site. In addition to leaving this significant factor out of their equations, many economists have yet to fully grapple with Michael Darby’s intervention, especially the way that the new data set allows us to observe both the direct and fiscal impacts of New Deal job programs.

Historians, for their part, continue to use Lebergott’s data in both scholarly books and articles and in mainstream textbooks, and are generally loathe to question the judgements of economists on the New Deal.

moreover, historians have tended to follow A. earlier historians who only had Lebergott’s data to work from, and B. economists, to whose expertise they defer


As the above section suggests, the relevance of this chapter to modern politics is that we need to be very careful in how we analyze the efficacy of different forms of stimulus. In weighing “bang for the buck” – as we did with tax cuts vs. more money for food stamps in the 2009 stimulus – we need to include both the direct and indirect effect of direct job creation policy. The result, as Modern Monetary Theory economists have been figuring out, is that DJC is subtantially more efficient in creating jobs than traditional stimulus spending. In other words, when we face the prospect of the next recession, we need to be thinking more Job Guarantee than ARRA.

The second point of relevance has to do with how we analyze past public policies. One of my frustrations with academics on this subject is that, out of the best of intentions, they have tended to conclude that, because the New Deal didn’t help 100% of the unemployed, it was a failure. This is clearly an insane standard, that we don’t bother to employ with regards to current policy. Obama’s stimulus didn’t help 100% of the unemployed either, but we know what the difference between 10% unemployment and 4% unemployment feels like in our lived experience.

Now more than ever, we need to apply the same kind of standards to the past as to the present, because the popular historical memory of major reform efforts like the New Deal and the War on Poverty play a powerful role in shaping people’s beliefs about the capacity of government to help them today. Conservatives understand this, and have waged a fifty-year campaign to convince people that government cannot help them. In this war for hearts and minds, a “yes, but” strategy will not work.

  • Facebook
  • Twitter
  • Google+
  • Linkedin
  • Pinterest
It is main inner container footer text