On February 10, Philip Harvey led a Columbia University Seminar on “Full Employment, Social Welfare and Equity” in the Faculty House. On February 21, he sent to the participants the following written responses to questions that were raised during the seminar.

(1) Do I believe that deficit spending caused the stagflation problems of the 1970s?

No, I do not. When I said that the failure of the Keynesian full employment strategy was revealed in the 1970s, I did not mean to suggest that deficit spending caused the stagflation crisis. I agree with Jeff that there were other causes. What the crisis exposed was the inability of the Keynesian strategy to combat high levels of unemployment that were accompanied by high levels of inflation. That, I believe, is what precipitated the sudden collapse of confidence in Keynesian macroeconomics during the 1970s among both professional economists and policy makers. Moreover, this collapse of confidence in the Keynesian strategy extended far enough into the left that full employment virtually disappeared from the progressive reform agenda for the next two decades, a trend highlighted by a comparison of the platforms on which Democratic candidates for the Presidency ran from 1944 through 1976 to those on which they have run since then. See Harvey,  Is There A Progressive Alternative to Conservative Welfare Reform? (2008) p. 173 note 49).

(2)    Do I think the Keynesian macroeconomic strategy is ineffective as a means of combating unemployment today?

No, I do not. I agree with other progressive economists that a full recovery from the so-called Great Recession could have been achieved before now with a larger dose of stimulus spending and that the recovery is still being delayed by the “austerian” response of policy makers to the recession.

(3) What then is my criticism of Keynesianism?

My criticism of Keynesianism differs depending on whether we are talking about the Keynesian response to the problem of unemployment during recessions or its strategy for achieving full employment (which I maintain is almost never achieved in market economies, even at the top of the business cycle—see Harvey,  The Trouble with Full Employment (2013)).

As a response to recessions, my argument is that the conventional Keynesian strategy requires more spending to achieve a given employment effect, is slower to achieve that employment effect, distributes the jobs it creates less fairly, and delivers its multiplier-driven stimulus to the private sector less efficiently than the direct job creation strategy pioneered by the New Dealers. See Harvey,  Back to Work: A Public Jobs Proposal for Economic Recovery (2011).

My criticism of Keynesianism at the top of the business cycle is that it is incapable of achieving genuine full employment because of its inflationary tendencies as the unemployment rate approaches the genuine full employment level of between 1% and 2%. This is something that the Swedish economists Gosta Rehn and Rudolf Meidner recognized from the beginning of the post-World War II era, but which progressive economists in general have been reluctant to acknowledge. See Harvey  (Why is the Right to Work so Hard to Secure (2013).

(4)    Do my cost estimates account for the fact that persons other than the officially unemployed would probably seek employment in the kind of program I describe?

First, I want to emphasize that I agree with Bill that the cost of the undertaking I propose is not precisely calculable. My estimates are only that—estimates. My goal is not to establish what the budget of a program designed to secure the right to work would be, but to demonstrate, based on a reasonable set of assumptions, how affordable the strategy is. In so doing I am merely responding to what I believe is a widespread assumption (among progressives as well as others) that it would be prohibitively expensive to actually guarantee decent jobs for everyone who wants to work. I think that assumption is false and would welcome more back and forth on the issue from anyone who thinks I’m underestimating the actual cost of the kind of program I am proposing.

My use of the “million job” unit of measurement in my cost estimate is nothing more than an accounting convention. I use it to estimate the average cost per job for a program serving a cross section of job wanters in the United States. Since there are no significant economies or diseconomies of scale for a program like the one I propose, I don’t think my use of that methodology poses a problem.

I also want to say that I agree with the observation that a direct job creation program like the one I propose would have to provide jobs for more people than just unofficially unemployed workers. All my estimates of program cost assume that jobs would have to be provided not only for officially unemployed workers, but also for involuntary part-time workers and for persons who say they want a job but are not actively seeking work and hence are not counted as unemployed in BLS statistics. In the end, I think my methodology is more likely to overstate than to understate the number of jobs that would have to be provided to secure the right to work; but no one should take my word for it. If you have questions about my estimation methodology, please ask them. I will be happy to respond.

Finally, I agree that securing the right to work would likely result in increases in the labor force participation rate over time that would exceed the immediate increase I assume in my cost estimates. But I don’t think this would increase the relative tax burden (per employed worker) of securing the right to work via the strategy I propose because increases in the labor force participation rate tend to create more macroeconomic space for private-sector economic growth to occur by natural or macroeconomically induced means before inflation becomes problematic. That’s why comparative labor force participation rates and employment to population ratios correlate so poorly with comparative unemployment rates. Once again, if anyone is interested in this issue I would be happy to explain my thinking in further detail.

(5) Would workers employed in a jobs program committed to paying market wages refuse to take private sector employment?

I would first like to point out that if this turned out to be a problem, there are a number of straightforward remedies for it. Hourly wages could be reduced in the program; hours of work could be limited; time limits could be placed on program participation; and/or administrative sanctions could be imposed on workers who refused bona fide offers of suitable employment (as UI laws already provide).

On the other hand, I think there’s a good chance these measures would be unnecessary except possibly at the top of the business cycle. First, the policy I favor would place workers in program positions only if it were determined (either administratively or via a job search/employee search requirement imposed on both employers and unemployed job seekers) that there were not enough jobs in the regular labor market to provide work for all job seekers with their qualifications. And even after their placement in the jobs program, both program participants and employers with suitable job openings to fill would continue to be referred to one another. In this environment I think it is reasonable to assume that private employers would grow accustomed to seeking candidates for job openings via the public employment service and public job creation program in addition to the places they now look—simply because it would be the cheapest and easiest way for them to access what Marx called the “reserve army of labor” when the burden of being unemployed no longer drove job applicants to seek them out. Employers have no trouble adopting aggressive recruitment measures when seeking to fill positions for which suitable applicants are not already knocking at their door. The only change is that they’d find it necessary to deploy these recruitment measures for all of their job openings. That’s what we should expect in a labor market in full employment equilibrium— notwithstanding employer complaints about labor shortages whenever they have to do more than pull out a stack of resumes from a file drawer to fill their job vacancies.

Would jobs program employees accept regular employment offered to them via this mechanism? I mentioned that program wages would be set at market levels, but the program’s wage scale would still be fixed in the short run—like public sector wage scales in general. That means an employer could always offer a program employee a better job simply by offering them a wage slightly above the program scale or with a slightly better benefit package, slightly better working conditions, slightly better hours of work, and/or slightly better opportunities for advancement. Again, employers already do this when seeking to fill positions for which the applicant pool is limited. We find it hard to contemplate only because we’re accustomed to most workers having to seek work in a job short environment.

Would employers complain? Of course they would until they became accustomed to the “new normal.”

(6) How would the Affordable Care Act (ACA) affect my program cost estimates?

It certainly wouldn’t make it any more expensive—since my existing cost estimates are already based on the assumption that the program would provide all employees the same health insurance benefits the federal government provides its employees. Whether a local program could shift some of its health care costs to the federal government by taking advantage of ACA subsidies and expanded Medicaid eligibility standards would require an analysis of the ACA and each state’s Medicaid and CHIP programs with that question in mind. I have not undertaken such an analysis, but it clearly should be done.

(7) Since regular public sector employment has declined due to the recession, wouldn’t it make more sense to support an expansion of the regular public sector workforce?

Of course it would, but that wouldn’t create enough jobs to secure the right to work, and since public sector hiring at the state and local level is constrained by balanced budget requirements, it’s not possible for state and local governments to maintain existing public sector employment levels during a recession without gutting other public services. What my proposal would do is permit state and local governments to maintain public sector employment levels during a recession by using jobs program trust-fund monies to pay the salaries of workers who otherwise would be laid off due to declining government revenues. Anti-displacement rules would have to be strictly enforced to prevent the inappropriate use of trust fund monies to pay the salaries of regular public sector employees in situations other than this, but that’s no more difficult to do than the enforcement of other regulatory limitations on the spending of public funds. The fact that Congress didn’t include adequate anti-displacement measures in CETA’s enabling legislation evidenced a lack of foresight rather than an inability to write and enforce such limitations.

(8) Since employers would oppose the strategy I propose with particular ferocity, doesn’t it make more sense to rely on more conventional strategies to expand employment and guarantee workers decent wages (e.g., with things like the EITC, raising the minimum wage, job training, the promotion of small business formation, and the deployment of the Keynesian macroeconomic strategy)?

Except for the Keynesian strategy, I agree that it would be theoretically possible to secure the right to work at the local level via these means, but progressives have been promoting that strategy for the past 50 years and the instances in which it has succeeded in achieving local full employment are as rare as hens’ teeth. I would support all such measures as an accompaniment to the direct job creation strategy, but I think the historical record is clear that they can’t do the job on their own.

At the national level, the problem is that none of these alternative strategies is capable of achieving genuine full employment. The effectiveness of the Keynesian strategy is limited by the

NAIRU (which I don’t think any progressive economist thinks is fixed, but which I also don’t think any progressive economist believes can be driven down to the 1%-2% level required to achieve genuine full employment). The other strategies would do little or nothing to close the economy’s job gap.

Moreover, even during a recession—when the conventional Keynesian strategy DOES work as an anti-cyclical intervention—it is a far more expensive, far slower, and much less equitable way to respond to the needs of unemployed workers than the direct job creation strategy; AND it also delivers no more macroeconomic bang for the buck in practice than spending the same stimulus dollars on a direct job creation initiative.

As I argue in my paper, if the money Congress allocated to the ARRA had instead been devoted to the strategy I advocate, the U.S. unemployment rate could have been reduced to the genuine full employment level of 1% to 2% as quickly as the jobs program could have been gotten up and running, and it could have kept it at that level beyond the 2010 midterm elections while simultaneously delivering a LARGER macroeconomic boost to the private sector than the ARRA did. My analysis in support of these claims is briefly explained in the paper I presented tonight (Harvey,  Securing the Right to Work at the State or Local Level) and more fully in the earlier Demos report from which it takes off (Harvey,  Back to Work).

One would think advantages as strong as these would inspire at least some interest on the part of progressive economists—notwithstanding the opposition of employers to the proposal.

Conservative opposition didn’t stop progressives from exploring and arguing the advantages of a single-payer national health insurance reform. Why is the direct job creation strategy verboten?

(9) What administrative structure do you propose for a local direct job creation initiative?

As explained in my paper, there are three “delivery models” that can be incorporated simultaneously in one centrally administered program, with each delivery model having certain advantages and disadvantages.

What I describe as the WPA model is one in which the program work force is employed on free-standing projects undertaken by the program itself. (And lest there be any confusion, when I refer to the WPA model I am NOT advocating the wage and hours policies the WPA adopted. They are in no way essential to the model—any more than CETA’s lack of adequate anti-displacement measures is essential to what I call the CETA model.)

The second model is the one CETA used—assigning program participants to jobs in regular government agencies where they work alongside regular government workers to expand and improve the quality of government services.

The third model is the one used in the College Work Study program. Under this model, not-for-profit employers are awarded grants that they can used to hire eligible workers to perform jobs in support of the not-for-profit’s mission.

For-profit corporations could also be paid to create jobs, but I don’t view this model as distinct from the College Work Study Model—since I assume it would be agreed that the jobs created by the corporation should not include profit making activities. This could be accomplished by including restrictions in the contract between the government entity and the corporation that would require the workers to be employed on public service projects that did not add to the corporations “bottom line” except via the PR value of their participation in the program and whatever contracted-for profit was negotiated up front to elicit the corporations participation.

I have described various aspects of the kind of program I am proposing in a string of writings on the subject going back 25 years. If anyone is interested in particular aspects of the administrative model I propose, I’ll be happy to respond to further questions.

(10) What kind of jobs could program participants reasonably be expected to perform?

I honestly don’t think this would be a problem. It certainly wasn’t for the CCC (which as a general rule only enrolled the children of families on local relief rolls) or the WPA (which in practice provided jobs for the poorest third of the nation’s population of unemployed workers). On those rare occasions when adequate numbers of jobs truly are available in local labor markets, the unemployment rate falls to the 1% to 2% level, so it’s clear that workers who remain unemployed when the unemployment rate reaches the 4% to 5% range are NOT unemployable.

That said, the jobs created by a direct job creation program designed to secure the right to work would have to be selected with the skills profile and trainability of the program’s workforce in mind, a strong preference for labor intensive projects, and a willingness to work with and accommodate private sector providers of similar goods and services. The following list is suggestive of the kinds of work that could be offered, consistent with these constraints.

• The operation of child care centers (on a sliding fee basis) and recreational programs for school-age children is an obvious choice because such services would be needed by the program’s own employees as well as the general public.

• Access to summer and school-holiday-break day and sleep away camp experiences could be offered as an entitlement (and on a sliding fee basis) to all children.

• The renovation of dilapidated housing and the construction of new low-and moderate income housing is another obvious choice because of the shortage of such housing in urban, suburban AND rural communities throughout the United States. The goal of program activities in this area could be to turn access to decent housing into an entitlement—with the subsidy required to make this possible furnished by the construction and renovation work of the job creation program combined with conventional financing secured by the sale or rental of the housing units.

• Permanent (or temporary) housing with needed support staffing could also be provided to homeless individuals and families.

• The dreary, over-crowded quality of government office spaces, including those where the public comes to receive public services, could be remedied. Three is no reason these spaces should not be as appealing as their private sector counterparts.

• Energy conservation improvements could be offered at nominal or no cost to existing home owners and renters.

• The CCC’s contribution to the conservation, expansion and improvement of public parklands and outdoor recreational areas could be emulated and expanded to include existing and newly created suburban urban parks.

• Recycling and related conservation measures of all types could be expanded.

• The work study program could be turned into an entitlement, with the participation of not-for profit organizations in addition to schools guaranteeing after-school and summer employment not only to post-secondary students but to high school students as well.

• Supplemental staffing could be provided to any unit of government that could use the extra help.

• Schools and libraries could be provided additional educational and support staff.

• The Public Works of Art programs of the New Deal era could be emulated to provide work on a competitive basis (as the New Deal program did) for aspiring writers, artists and performers.

• Community-based support services for the elderly and disabled populations could be expanded—including transportation services, shopping assistance, home maintenance and upkeep, meals on wheels, etc.

• Support services could be also be furnished for program workers (and others) with special needs.

(10) What if any residency requirements would a local job creation program have to impose?
This is a challenging issue—perhaps the most challenging issue a locally-funded job creation program would have to face—as I conceded in response to the question when it was posed during the seminar. Nevertheless, I think it can be worked out. There are various ways to limit eligibility based on social insurance principles. These include but are not limited to the following.

• Residents of the locality could be required to live in the locality for a certain period of time before becoming eligible in order to deter people from moving to the locality simply to obtain program employment.

• Non-residents of the locality who are employed in the locality could either be excluded from eligibility or given the opportunity to obtain eligibility by working in the locality and paying taxes in support of the program for a specified period of time (as is currently the case for non-residents of a state under the state’s UI program).

• To the extent program slots had to be rationed, preference could be given to applicants based on the length of their residency (or employment) in the locality.

In the long run, of course, there is reason to hope the problem would solve itself by incorporating progressively larger units of government into the program (cooperating localities, counties, states, and ultimately the federal government). It’s also worth noting that even a nationally administered program would face this issue through its impact on unauthorized immigration and the many problems posed by the existence of two work forces in the economy—one legally authorized to work and the other lacking such authorization. From a human rights perspective I think immigration restrictions are as problematic as restrictions on emigration and the free movement of people within a country—and in the long run I think the remedy for the problem is the same at the national level as it is at the local level: expand the program beyond the polity’s borders through cooperative arrangements with other polities.