Recent Posts

Saturday, July 25, 2020

Primer: The Job Guarantee

The Job Guarantee is a programme that offers all citizens willing to work for it a guaranteed wage – which becomes the de facto minimum wage in the economy. It creates a floor for incomes and working conditions. It also gives the government a workforce that can undertake non-time sensitive projects. (Permanent government jobs and most infrastructure jobs would be unaffected.) The implementation of the programme would be country-specific, as the objective would be to re-use existing job seeking infrastructure. The central government would pay the workers and oversee the programme, but local governments and non-profits would likely be the employers, at least in larger countries.

Note: This is an unedited draft from my upcoming Manuscript. Rather than write a review of Pavlina R. Tcherneva's The Case for a Job Guarantee, I have created my summary of that text.

The Job Guarantee is a core part of Modern Monetary Theory, both as a policy recommendation, as well from a theoretical perspective. The realisation that governments can use the pool of labour within the Job Guarantee programme in a fashion analogous to a commodity price support programme to stabilise the price level was independently observed by Warren Mosler and Bill Mitchell.

My description of the Job Guarantee leans heavily on The Case for a Job Guarantee, by Pavlina R. Tcherneva (Amazon affiliate link). That book describes the programme, as well as outlining the research behind it. I make some editorial comments herein which does not necessarily align with Tcherneva’s views, but those digressions are noted. Since most internet commentary from critics is concentrated on implementation difficulty, that is the focus herein.

Implementation Outline

The underlying premise of a Job Guarantee is that the central government guarantees that a job exists for all citizens willing to work at a fixed wage. Beyond that, the implementation could vary widely, and different authors might have different suggestions.

Pavlina Tcherneva argues that the best strategy is to not reinvent the wheel; existing programmes can be repurposed to fit within the Job Guarantee. In the United States, in addition to existing job centres, state and municipal governments already have job creation programmes, and non-profits are already recognised and “regulated” (their activities have to be certified by tax authorities). Meanwhile, these activities are typically under-funded and under-staffed.

The ease of implementation is going to be related to cyclical conditions. During the worst of the pandemic, it was difficult to organise new activities (given the unfamiliar physical distancing rules), and the number of potential entrants would have been extremely large. Conversely, the number of potential applicants would presumably be reasonable if the economy had been expanding for some time.

One common discussion point is whether these are really jobs, and whether people can be fired from them. Planting trees or cooking in a homeless shelter are certainly real jobs, and so it is not a question of lying around playing video games. Disruptive workers can be fired from the Job Guarantee, and they will end up on last chance social welfare programmes. (Since hardship cases of sick or disabled individuals will exist, such a safety net is still needed.) The guarantee is a job for people willing to work, and is not an income guarantee.

With the exception of some youth apprenticeship programmes, the workers would not be made available to for-profit firms. Presumably, other training or education programmes would be coordinated with the Job Guarantee, but the details would vary widely by jurisdiction.

Tcherneva pointed out that governments already run some complex activities, such as kindergarten to high school in North America, or significant portions of health care outside of the United States. Shying away from manageable complexity is a defeatist attitude.


I will keep the discussion of benefits brief, on the basis that most reasonable people understand that this programme benefits works, their concern revolves around the costs/side-effects of the programme. I summarise some of the benefits as follows.
  • A powerful automatic stabiliser and acts to stabilise inflation (unlike many alternatives).
  • Unlike Old and New Keynesian aggregate-demand approaches, the spending is flowing exactly to the areas where demand is weak. By contrast, lowering interest rates helps fuel a property bubble in the big cities, while rural areas stagnate in poverty (the pattern see in the United States in recent decades).
  • Across the social sciences, the consensus is that having people socialising in a workplace generally results in better psychological outcomes.
  • Society pays other costs associated with unemployment, and the existence of the programme would reduce them.
  • Employers do not like hiring the long-term unemployed, as their “attachment to the workplace” is lost over time (e.g., habits like punctuality, or conformity to the social hierarchy from the perspective of an anarchist). Savvy employers that pay relatively low wages will develop relations with supervisors inside the programme (such as volunteers at non-profits) to get an honest appraisal of workers’ abilities. (Warren Mosler emphasises this point, referring to the Job Guarantee as a “transition job.”)
  • The state has labour power to undertake non-time sensitive projects, such as long-term environmental remediation. A century ago, it might have made sense to send workers out with shovels to dig ditches, but the current low cost of energy means that such projects should be done in a more capital-intensive fashion.
  • The authorities have a new policy lever that allows it to influence wage differentials. (Some might view this a negative.)

Steady State Inflation Stabilisation…

The basic assumption of analysis is that private sector firms will generally need to offer compensation that is at a premium to the Job Guarantee compensation package – or at least offer starting-level positions that offer the possibility of a promotion to better-compensated jobs. (Intern positions already draw in students at below-market pay as a way of getting office experience.)

Note that I wrote “compensation” in the previous paragraph, as the Job Guarantee could have benefits (like medical insurance or daycare) that are not already mandated/provided by the state. (By offering benefits or advantageous working conditions – like guaranteed job schedules – the government can push employers to match the new standard.). For simplicity in the rest of this text, I will refer to the “wage differential” between the Job Guarantee and private sector jobs, embedding other aspects of compensation into “wages.” The argument for just focussing on the wage is that the other aspects of Job Guarantee compensation will not move with the business cycle (unlike in the private sector, where employers will change the compensation mix based on job market conditions). This matches the terminology of conventional economic models, which just uses “wages” instead of “compensation” when modelling the labour market.

In the steady state, the inflationary effect of the Job Guarantee is not a concern under the fundamental argument that the government is bidding at a fixed price for wages. If wages in the private sector rise relative to the Job Guarantee wage, the relative wage differential rises, and so there will be tendency for workers to move to the private sector. This creates two mechanisms for inflation stabilisation: the workers add supply of labour in response to rising prices, and the total spending on Job Guarantee wages by the central government falls, reducing fiscal stimulus.

By offering a fixed price bid for labour, the possibility of wage deflation for lower-paid workers is almost certainly eliminated – rather than taking a pay cut, they will quit. (Highly compensated workers can have their pay cut, but this process will be short-circuited by the fact that their wages would eventually converge to the Job Guarantee wage.) Meanwhile, people moving to the Job Guarantee pool raising government spending, and thus automatically creates new fiscal stimulus.

Critics have noted that the Job Guarantee will stop damping inflation if the pool of workers (that are perceived as employable) drops to zero. Since this represents true full employment, and that is hardly an outcome that I would spend a great deal of time worrying about. That said, the discussion is deferred to the chapter of MMT criticism, in Section 5.5.

… Initial Price Level Shock

The lack of worry about inflation in the steady state does not imply that there will be no price consequences of the Job Guarantee. The expectation is that there will be a one-time adjustment of wages for workers at the bottom end of the wage distribution.

Tcherneva (and her colleagues) currently suggest a $15/hour Job Guarantee wage in the United States – while the Federal minimum wage is $7.25/hour. (That Federal minimum wage is a bit of a dead letter; some states have much higher minimum wages already.) Given the likely possibility that many existing low-wage employers will have to pay a premium over the Job Guarantee wage (although that is hypothetical; we do not have empirical evidence of this effect), they will have no choice to either raise wages and prices, suffer lower profit margins, or cease to exist.

Tcherneva argues in Chapter 5:
However, this one-time increase should not be confused with inflation – which is a continuous increase in the price level. Such a significant one-time wage increase would not be unprecedented. In 1949, the minimum wage was nearly doubled without accelerating inflation, at a time when the economy was as close to true full employment as it has ever been in the postwar era.
It is highly predictable that employers that are directly affected by rising wages at the bottom of the income distribution will argue that this change would be cataclysmic. Given that low-wage occupations are almost by definition low productivity ones, the shrinkage of that “sector” should raise aggregate productivity. Otherwise, relative prices will have to adjust, which is what they are supposed to do. Given the tendency of downward stickiness in wages and prices, the net effect should be a one-time upward adjustment of output prices to match the higher cost of labour. That said, it appears implausible that the entire price structure would move in parallel with the rise of wages at the bottom. (For example, I would have chortled if any analysts  that reported to me used such an excuse to ask for a pay raise, and I doubt that I would have been the only one with that attitude.)

Real-World Programmes

Pavlina Tcherneva discusses similar programmes that were implemented in Chapter 5. She notes that most were temporary, other than India’s National Rural Employment Guarantee Act (NREGA), although NREGA is not a universal programme.

Well-known temporary or small programmes included the New Deal (Depression era) and the Youth Incentive Entitlement Pilot Projects (1978-1980) in the United States, Argentina’s Plan Jefes, Youth Job Guarantee in Brussels, zero long-term unemployment regions in France, and the first iteration of the 2009 Future Jobs Fund in the U.K. (it transitioned into “a punitive workfare programme”).

Delving into the lessons that were learned in these programmes is beyond the scope of this text.

Standard Criticisms

The Case for a Job Guarantee notes a number of commonly raised criticisms. I will summarise some that were of interest.
  • Tcherneva’s response to the administrative difficulty is to observe that we do not believe that there is an optimal level of illiteracy; governments already undertake complex tasks to deal with pressing social issues.
  • Some economists are worried about productivity. Tcherneva responds that leaving people unemployed is going to be less productive than giving people almost any somewhat useful job. A related question that I have come across (but not discussed in the book) is the argument that the number of workers in the Job Guarantee pool would be larger than (hypothetical) equilibrium unemployment. The argument was based upon neoclassical arguments about job searching. This is an interesting question, but it relies heavily upon the assumption that neoclassical beliefs about the labour market are correct. Since MMTers dispute the notion of an “equilibrium” unemployment rate, resolving this debate would require a deep dive in the different approaches to labour market analysis, which is beyond the scope of this text.
  • Another argument is that the jobs will be “make work” jobs. Tcherneva notes that there are two approaches: we can either look at whether the projects need to be completed, or we can approach the problem as being a guarantee for individuals. If we focus on projects, the calculus then naturally evolves towards hiring the best people for the job based on efficiency arguments. This mode of analysis makes sense for standard governmental programmes, but is not a good fit for the analysis of the Job Guarantee. It was one of the weaknesses of the New Deal.
  • People are (or at least) were concerned about robots taking all the jobs, and humans will not have any menial tasks left to do. I think that this is largely delusional propaganda pushed by the tech industry, and so the validity of the argument depends on whether one agrees with that view.
  • There are concerns that it would be difficult to respond to changing cyclical conditions. Firms already deal with the cycle (although they have the slightly simpler task of shutting down activities on short notice, rather than starting them up). Certain activities – like daycare – are not going to be possible with a rapidly rotating workforce. (Although one might imagine that under the assumption that the Job Guarantee offers childcare, some additional staff could be added. However, licensing/safety concerns would mean that people could not walk off the street straight into a child care position.)
  • Those on the left have several criticisms. One argument is that the Job Guarantee is yet another workfare scheme. Another arguments relies upon the logic found in Michal Kalecki’s essay “Political Aspects of Full Employment.” In brief, Kalecki’s argument that capitalists would be deeply opposed to any policy that diminishes their power over workers. Pavlina Tcherneva observed that every increase in the rights of workers that we now take for granted (e.g., minimum wage, 5-day working week) were fought against by free market ideologues. In my view, this is just an observation that is as unremarkable as saying that the sun rises in the east. Certain employers will be forced or radically change practices (or fail), but competitors who pay a good wage benefit. Meanwhile, rising aggregate demand is good for business conditions. The coalition building is difficult, but in my opinion, not impossible.

The Job Guarantee vs. the Universal Basic Income

One policy that is often discussed in the context of the Job Guarantee is the possibility of paying a Universal Basic Income (UBI). Some people view the programmes as being complementary, while some MMT proponents are skeptical of the UBI. I am one of the people who are skeptical about the UBI. Please note that this discussion represents my views and is not based on The Case for a Job Guarantee. Note that I am discussing a UBI that lefts someone out of poverty without any other income sources; it is obvious that governments can make small transfers to all citizens without causing too much disruption (or do one-time cash transfers to counter recessions, which is currently popular).

This is an old debate. For example, Hyman Minsk demonstrated the difficulties with Milton Friedman’s Negative Income Tax in the 1960s, while he preferred an Employer of Last Resort programme (which is like a Job Guarantee to an outside observer). (Proponents of a UBI argue that the Negative Income Tax is different, but to an outside observer, once again, the differences are cosmetic.)

A UBI has the advantage that it is simple – all citizens (of a certain age?) receive a monthly cheque from the central government (for the same amount). That is, no system of means testing (which is normally attached to any modern “liberal” programme). In Canada, senior citizens already receive such a benefit (in addition to the universal pension plan that pays based on contributions, as well as means-tested hardship benefits), a UBI just reduces the age at which the benefit is received.

My concern with the UBI is not “paying for it,” rather basic multiplier analysis. A UBI has large implications for domestic demand. My argument is that the evidence is that a significant portion of the population has a high propensity to consume out of income, with most savings being generated by a smaller number of households (often with higher incomes). The implication is that there would be a much higher multiplier on UBI expenditures than is the case for policies like income tax cuts.

This aligns with Minsky’s relatively traditional analysis that suggested that “… a negative income tax is expansionary or inflationary, even if budgets are balanced.” That is, it is not enough to raise taxes to match the total spending on the UBI, the taxes raised would need to be larger. This would particularly true if something like a wealth tax was used instead of raising income tax/value-added tax rates, since the brunt of the wealth tax would hit households with a low propensity to consume out of income.

If as I suspect, a broad-based income tax hike would be required to contain the inflationary pressures of the UBI, the result is that a significant portion of households would face much higher marginal tax rates and end up with almost net benefit. (That is, the UBI payments are largely matched by increased income taxes.). This is a very unattractive policy mix, and it would be quite easy to agitate to lower the real value of the UBI in exchange for lower marginal tax rates. Since the premise of the UBI is that it would replace welfare state programmes, the ultimate result is the demise of the welfare state (possibly explaining why Milton Friedman backed the proposal).

Concluding Remarks

Developed economies have been sluggish in recent decades, and persistent underemployment seems to be tied to this. A Job Guarantee would eliminate the ability of firms to force workers into precarious jobs; they need to provide legitimate work positions. A transition to a higher-pressure labour market would force true innovations, and not just regulatory arbitrage.

References and Further Reading

  • The Case for a Job Guarantee, by Pavlina R. Tcherneva. Polity Press, 2020. ISBN: 13-978-1-5095-4211-6.
  • "Political aspects of full employment," by Michal Kalecki. New York and London (1943).
  • “The Macroeconomics of a Negative Income Tax.” By Hyman P. Minsky (1969). Reprinted in Ending Poverty: Jobs, Not Welfare, 2013. Levy Economics Institute of Bard College. ISBN: 978-1-936192-30-4.

(c) Brian Romanchuk 2020

No comments:

Post a Comment

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.