The Trouble With Full Employment

Columbia University sign in subway station IMG 0974 / CC BY-SA

By  Philip Harvey
Presentation at Conference on An Economic Bill of Rights for the 21st Century
Columbia University, Oct. 18, 2013

One of the things that’s always puzzled me about President Roosevelt’s economic bill of rights is how little impact it has had on the way progressives have conceived and promoted their reform agenda. While not averse to rights based policy analysis and rights based political claims—the civil rights movement demonstrated that—progressives have been hesitant to follow Roosevelt’s lead in embracing human rights norms in their analysis and promotion of economic and social entitlements.

I believe the reasons for this hesitancy are an important topic of inquiry and self- reflection, but it is not the topic I intend to pursue in my remarks this morning. Instead I want to comment on the consequences of this tendency for one key component of the progressive reform agenda—the struggle to achieve what Roosevelt referred to as the right to a useful and remunerative job.

Before I start, however, I need to say something about terminology. Specifically, I want to point out that the right to a decent job is generally referred to as the “right to work” in international human rights discourse. Indeed, this usage dates back to the French Revolution and Roosevelt himself employed it—most famously in his acceptance speech at the 1936 Democratic National Convention where he thundered against the “royalists of the economic order” who “granted that the Government could protect the citizen in his right to vote, but . . . denied that the Government could do anything to protect the citizen in his right to work and his right to live.”1

I am perfectly aware of the fact that the “right to work” phrase has very different connotations in the United States today, where it has been appropriated by right wing organizations for use as an anti-union slogan. I simply want to make it clear that when I use the term, I am using it the way President Roosevelt did.

At the same time, it is worth noting that the ability of conservatives to appropriate the term, and turn its meaning on its head, illustrates how little claim American progressives have laid to the human rights language that both Franklin and Eleanor Roosevelt championed.

In addition to illustrating Roosevelt’s use of the right to work phrase, the 1936 speech from which I just quoted also shows his predisposition—long before he delivered his economic bill of rights speech—to view access to a decent job and a secure livelihood as basic human rights.

But did he believe the right to work could actually be secured? He certainly never wavered in his insistence that it could, and the period leading up to his 1944 economic bill of rights speech seemed to give him good reason to feel confident. By 1942 the nation’s unemployment rate had fallen to 4.7 percent, in 1943 to 1.9 percent, and in 1944 to 1.2 percent. The contrast with the 1930s could not have been starker, and progressive economists were virtually unanimous in touting this achievement as proof of the Keynesian proposition that deficit spending could in fact achieve what they called “full employment”—a term just then being popularized.

It was a seductive idea because it seemed so simple and promised so much. If war-time spending could achieve full employment with such apparent ease, why couldn’t peace-time spending do the same—with American workers employed building homes and schools and hospitals instead of warships and tanks?

This was the economic and intellectual environment in which Roosevelt proclaimed his economic bill of rights. It is therefore hardly surprising that Roosevelt’s congressional allies chose to rely on the Keynesian full employment strategy when they undertook the task of drafting legislation to implement the right to work, the first entitlement on Roosevelt’s list and the one he considered the “most fundamental” of all the rights he enumerated, because it was, in his words, the “one on which the fulfillment of the others in large part depends.”2

In fact, the bill Roosevelt’s allies drafted to secure the right to work was so thoroughly Keynesian that the right itself was mentioned only once in the bill. The rest of the bill— beginning with its title—spoke only and repeatedly of achieving full employment.

No one questioned this substitution of the Keynesian full employment goal for the Rooseveltian goal of securing the right to work, because everyone—including Roosevelt— assumed that full employment meant the under-2 percent unemployment the country was experiencing at the time—a level of unemployment that was, with good reason, considered synonymous with securing the right to work.

This full employment bill never made it into law, but it did establish the strategy progressives have promoted ever since then to secure Roosevelt’s right to work goal. And that, I submit, is the reason we have failed over the past 70 years in our efforts to do so.

The problem, in my view, is that by unqualifiedly embracing the Keynesian full employment strategy, progressives hitched their wagon to a horse that simply could not make it to the top of the hill. The Keynesian strategy can drive unemployment rates down to the level at which inflation becomes problematic, but except in very unusual and hard to replicate circumstances, that level of unemployment is nowhere near low enough to insure the availability of work for everyone who wants it.

Some progressive economists, most notably Gösta Rehn and Rudolf Meidner in Sweden, were aware of this problem from the start and took steps to address it. But they were the exception, and the strategy they developed relied on institutional features of Swedish collective bargaining that do not exist elsewhere and ultimately proved unsustainable in Sweden itself.

So when the stagflation crises of the 1970s finally exposed the inflationary Achilles heel of the Keynesian full employment strategy, neither progressive activists nor progressive economists had a fallback plan. What’s worse, instead of going back to the drawing boards to develop a new strategy for securing the right to work, they simply dropped the full employment goal from their reform agenda while soldiering on in the vain hope that the progressive reform project could survive without achieving even the attenuated form of full employment that progressives had aspired to achieve in the 1950s and 60s.

But aren’t highly-respected progressive economists like Joseph Stiglitz, Paul Krugman, and Jared Bernstein once again promoting full employment? They are. But there’s a catch. The full employment goal they’re promoting is not the full employment goal that progressives viewed as synonymous with securing the right to work in 1944. Instead, it’s simply the lowest sustainable rate of unemployment they these economists believe is achievable via the policies they advocate. In other words, they’ve redefined the top of the hill as that point on its slope that the horse they’re flogging actually can reach.

This explains why Krugman and Bernstein, for example, feel comfortable referring to 5 percent unemployment as full employment, notwithstanding the fact that the last time the U.S. achieved that rate, American employers were seeking to fill only 3.9 million job vacancies while 7.5 million officially unemployed workers were actively seeking work, another 5.2 million workers were employed part time but wanted full-time jobs, and an additional 4.7 million people said they wanted a job but were not counted as unemployed because they weren’t actively seeking one. So while it’s true that the full employment goal is once again being promoted by prominent progressive economists, that’s only because they are using the term in a way that completely severs its link to the right to work.

I know there are plenty of progressives who still think of full employment the way the term was understood in 1944. Indeed, I’m quite certain this room is filled with such people. I believe, though, that the only way to reclaim that definition is for progressive activists to insist, and vociferously so, that progressive economists explain their usage of the term clearly enough and often enough that there will be no confusion as to whether the full employment goal they are promoting would in fact secure the right to work.

This does not require a rejection of Keynesian economic analysis. I want to make that perfectly clear. It only requires that the limitations of the Keynesian strategy for achieving full employment be recognized and frankly acknowledged.

And for progressive economists willing to reaffirm their commitment to securing the right to work, it means taking up the challenge of developing a new strategy for achieving that goal.

My own view is that a good place to begin that effort is with a reexamination of the New Dealer’s own efforts to secure the right to work before they invested their hopes in the Keynesian strategy.

An exploration of this history leads back to a cabinet level task force called the Committee on Economic Security that President Roosevelt appointed in the spring of 1934 to develop a set of legislative proposals that would address, in a comprehensive fashion, the economic security needs of the American people—not just during the Great Depression, but on an ongoing basis in the future.

The Committee’s final report, which was formally conveyed to the President in January 1935, proposed the establishment of the Nation’s Social Security and Unemployment Insurance systems, an income support program for poor children that evolved into the AFDC program, an income support program for poor seniors and a strategy for improving the provision of income support for the disabled poor that eventually evolved into the SSI program, and a variety of public health initiatives, including a national health insurance program (though Roosevelt asked the Committee to delete that particular proposal from its final report when it became clear that the AMA would be able to block its implementation in Congress).

In short, the Committee on Economic Security and its staff, were the architects of the American welfare state as we know it, and the Committee’s 1935 report constituted the first blueprint of that edifice.

There is, however, one element of the Committee’s overall plan that has not survived. This missing element was their proposal for achieving what they called “maximum employment,” a goal they described as the “first objective in a program of economic security.”3

Remember, the term “full employment” had not yet been popularized and it would be 13 months before Keynes’s General Theory was even published. So how did the Committee on Economic Security propose to achieve “maximum employment” without the benefit of the yet to be postulated Keynesian strategy?

Their proposal was straightforward. The federal government should back up its efforts to stimulate private sector employment (whatever form those efforts might take) with a commitment to provide temporary jobs in the public sector for those workers the private sector was unable to employ at a particular moment in time or in a particular place. In other words, the Committee proposed the use of direct government job creation to close the economy’s job gap, thereby providing workers with what the Committee called “employment assurance”— another term for securing the right to work.

Sadly, instead of fully implementing this recommendation, President Roosevelt decided for a variety of reasons to adopt the more modest goal of providing jobs for only those unemployed workers who were able to demonstrate financial need. Nevertheless, since 40 to 50 percent of all unemployed workers were able to satisfy this requirement, Roosevelt’s commitment promised a dramatic reduction in unemployment and an even more dramatic boost to the living standards and morale of the neediest members of the nation’s unemployed population and their families.

The programs used to implement this strategy were the Works Progress Administration (the WPA) and its smaller, 2-year old cousin, the Civilian Conservation Corps (the CCC). In 1936, the WPA’s first full year of operations, employment in these two programs reduced the nation’s unemployment rate by 6.3 percentage points—almost double the 3.3 percentage point decrease attributable to new private sector hiring, notwithstanding the fact that the latter was fueled by a 12.9 percent rate of real GDP growth, the fastest annual growth rate of any peacetime recovery on record.

Unfortunately, the singular ability of the WPA and CCC to drive down the nation’s unemployment rate has gone largely unnoted. In part this is because the most frequently cited unemployment estimates for the 1930s count WPA and CCC workers as unemployed rather than employed—contrary to generally accepted statistical practices today.

Another reason the job creation effect of these programs has been underappreciated is because the Keynesian framework of most such assessments focus on the impact of program spending on private sector hiring—via the multiplier mechanism—while ignoring the direct job creation effect of program hiring.

This chart shows the normally cited unemployment rates for this period in blue and the actual rates, counting WPA and CCC workers as employed, in red. Notice that the normally cited figures indicate that the nation’s unemployment rate stood at 17.0 percent in 1936, whereas the actual rate, counting WPA and CCC workers as employed, was 10.8 percent. No wonder Roosevelt was reelected that year in the largest electoral vote landslide in the nation’s history.

The same bias is apparent in the tendency for progressive to attribute the achievement of full employment during World War II to increased government spending. If we view the armed forces as a replacement for the WPA and CCC, which in a real sense they were, it immediately becomes apparent that the achievement of full employment during the war could just as easily be attributed to the direct job creation effect of military enlistments. The federal government created three times as many jobs in the armed forces during the last three years of the war than it had in the WPA and CCC combined at their peak.

So, I ask you, did World War II prove the effectiveness of the Keynesian full employment strategy or of the Committee on Economic Security’s employment assurance strategy? If this causes you to wonder what a comparison of the advantages and disadvantages of the Keynesian and New Deal strategies would show, I can only say that I think it confirms the superiority of the New Deal strategy in almost every respect. As an anti-cyclical policy this superiority can be summarized in the following five points. [SLIDE 4] 1. First, the New Deal strategy delivers aid to unemployed workers in a far more valuable form—actual wage paying jobs—than the social insurance payments favored by most progressive economists today.

2. Second, the New Deal strategy also creates far more jobs per dollar of stimulus spending. If, for example, the $787 billion fiscal stimulus authorized by the American Recovery and Reinvestment Act had instead been spent on an updated version of the WPA, paying market wages for 40-hour per week full-time jobs with full health insurance benefits, it could have reduced the nation’s unemployment rate to less than 2 percent for over two years, while simultaneously producing a larger multiplier effect on private sector employment than the ARRA did.

3. Third, the direct job creation strategy also creates jobs much faster than the Keynesian strategy. Its job creation effect is front loaded, whereas the job creation effect of the Keynesian strategy is back loaded. Page 6 of 8

4. Fourth, it is far easier to target the job creation effect of the New Deal strategy, insuring, as the New Dealers did, that if not enough jobs are created to provide work for everyone who needs it, at least those who need it the most will get the jobs that are created.

5. Fifth, whether deployed during a recession to promote recovery or as a prophylactic against recessions, the direct job creation strategy also delivers its fiscal stimulus to the private sector in a more even and effective way—by short-circuiting the process that causes unemployment to lead to more unemployment during an economic downturn, and which permits lingering unemployment to act as a drag on the economy’s recovery once the recession is formally over.

Turning to the relative merits of the two strategies as a means of securing the right to work, the superiority of the New Deal strategy is even clearer. Why? Because it has three features that should enable it to circumvent the inflation barrier that cripples the ability of the Keynesian strategy to achieve that goal.

1. First, the direct job creation strategy has a natural tendency to limit its job creation effect to those geographic locations or segments of the labor force that actually do suffer from a shortage of jobs. It is not prone, as the Keynesian strategy is, to aggravating labor shortages in some sectors of the economy while others are still mired in recession.

2. Second, because they would remain available for private sector employment, jobs program employees would perform the same inflation-fighting buffer-stock function that unemployed workers do—but without requiring anyone to suffer unemployment.

3. Third, unlike the Keynesian strategy, the direct job creation strategy can boost labor demand at the top of the business cycle without increasing aggregate demand. The trust fund financing of the nation’s unemployment insurance system illustrates how this can be accomplished.

Finally, I also want to mention one political advantage of the direct job creation strategy that I believe is extremely important—that is, its suitability for implementation at the state or local level as well as the federal level. It would cost more per capita, but progressives would not have to wait for Washington to act in order to implement the strategy in places where they possessed the political power to do so.

The only significant disadvantages of the direct job creation strategy is the difficulty of administering it. Like a public education system, a national health insurance system or a standing army, the operation of a direct job creation program capable of securing the right to work would pose an enormous administrative challenge. But that doesn’t mean it can’t be done well. The New Deal demonstrated that as well.

So, in conclusion, I think it is past due time for progressives to reassert the link between their full employment goal and the goal of securing the right to work and to acknowledge their need for a new strategy to achieve that goal. Fortunately, I think both of these tasks are doable.


1 Acceptance speech at Democratic National Convention, July 1936.
2 State of the Union Message to Congress, January 1935.
3 Report of the Committee on Economic Security, January 1935.


For fuller explanations of the arguments advanced in this talk, copies of the following may be accessed at

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.