A government’s job is to protect its people from misfortune, in particular want of work, and one that abandons this duty of care to the market deserves to be cast out. This is the best argument for giving a government job-guarantee program a fair trial.

“Any government,” writes the economist and hedge fund manager Warren Mosler, “can achieve full employment by offering a public service job to anyone who wants one at a fixed wage.” Versions of this idea have received powerful endorsements from prominent Democratic politicians in the US, including presidential candidate Bernie Sanders and Rep. Alexandria Ocasio-Cortez, who has linked a government job guarantee to a Green New Deal. Moreover, versions of a job-guarantee program (JGP), more or less connected to green economics, have been implemented in Argentina, India, South Africa, and – whisper it quietly – Hungary under its illiberal populist leader, Viktor Orbán.

In the United States in December 2017, 6.6 million people were deemed to be officially unemployed, 4.9 million were working part-time but wanted full-time work, and 5.9 million wanted to work but were not counted in the official statistics. In other words, at least 17.4 million Americans wanted but were unable to find stable, well-paid work.1

Image: Levy Institute

Economists L. Randall Wray, Flavia Dantas, Scott Fullwiler, Pavlina R. Tcherneva, and Stephanie A. Kelton have proposed that the US government guarantee all such jobseekers as much work as they want – up to a weekly cap of, say, 35 hours – at a fixed wage of $15 per hour. Today, such a JGP might employ around 16 million American workers and cost about 2% of US GDP. Unwanted unemployment, as we have known it since the Industrial Revolution, would no longer exist.

But, of course, JGP proposals raise many questions: How feasible is such a program in a fully developed capitalist society? How is it to be paid for? Is there likely to be enough political support for it? What are the main objections?

Keynes with a twist

Contemporary JGPs are inspired by the British economist John Maynard Keynes, but differ from classic Keynesianism in three important ways. First, although JGP advocates accept Keynes’s argument that the private sector may not generate enough jobs to maintain full employment, they argue that the government should provide extra jobs directly instead of trying to stimulate private spending to create them.

Second, a JGP is more radical than classical Keynesianism in its social objectives. For Keynesians, full employment was an end in itself. But to today’s post-Keynesians, a JGP is a means to address a host of other socio-economic problems – notably poverty, inequality, and climate change.

Third, a JGP is supported by Modern Monetary Theory, which brushes aside the problem of “financing” full-employment programs. Government spending, MMT claims, creates its own tax base. It is this monetary underpinning of JGPs, rather than the proposal itself, which has drawn most of the criticism.

Practice and theory

Governments have always provided or subsidized “public works” during economic slumps, though never on a scale large enough to match the rise in unemployment. In the 1930s, US President Franklin D. Roosevelt’s Public Works Administration paid private construction firms $7 billion to build airports, roads, dams, bridges, schools, and hospitals. In addition, the Works Progress Administration (WPA) provided local jobs directly to the unemployed.

Advocates of the two approaches competed for influence within the Roosevelt administration. The outcome – a victory for the WPA and direct job creation – opened the way for more substantial fiscal intervention and a more profound connection between the New Deal and the “one-third of a nation” struggling with economic deprivation.

It was Keynes who provided the theoretical rationale for a permanently larger public presence in the private economy. He argued that in an economic slump, if the government were to set people to work rather than pay them unemployment benefits, the multiplied effect of their earnings would restore the private sector to full employment. As a permanent policy, a public job guarantee would greatly dampen fluctuations in the business cycle.

Moreover, Keynes insisted that providing employment would have a multiplier effect regardless of whether the extra employees produced anything of economic value. In typical tongue-in-cheek fashion, he wrote: “If the Treasury were to fill old bottles with banknotes, bury them at suitable depths in disused coalmines which are then filled up to the surface with town rubbish, and leave it to private enterprise [...] to dig the notes up again [...] there need be no more unemployment and, with the help of the repercussions, the real income of the community [...] would probably become a good deal greater than it actually is.”

The golden age?

Persuaded by Keynes’s masterwork, The General Theory of Employment, Interest, and Money, and the political circumstances of the time, post-war governments committed themselves to maintaining “full” or “high” employment. Although there was no official employment target, full employment was generally reckoned to have been reached if the unemployment rate was 2% in the United Kingdom and 4% in the US. The level of unemployment would be determined by “macropolicy,” or the discretionary manipulation of interest rates, taxes, and public expenditure to maintain a full-employment level of total spending.

This worked pretty well for 20 years or so, but ran into two problems. First, governments lacked the forecasting power to undertake the expertly timed monetary and fiscal interventions on which macropolicy depended. Governments would forecast the difference between actual and potential output (the output gap) or the excess of total spending at the full-employment level of output (the inflation gap), and set their interest-rate, tax, and spending policies to close whichever gap threatened. But how large were these gaps at any point in the business cycle? As Milton Friedman pointed out, fiscal effects were subject to “lags,” during which any initial gap might have grown larger or smaller. Policymakers were thus as likely to destabilize the economy as steady it.

The second, more significant problem was one of political economy. Because Keynesian governments attached more importance to keeping unemployment low than to suppressing inflation, budgets became increasingly unbalanced amid growing political pressure to prevent any rise in joblessness.

This was the entry point for the monetarist counter-revolution. In 1968, Friedman claimed that excessive expansion of the money supply to hold down unemployment was the cause of rising inflation. The attack was lethal. The demise of Keynesian “demand management” was signaled by Britain’s Labour Prime Minister James Callaghan in a speech in 1976. “We used to think that you could spend your way out of a recession and increase employment by cutting taxes and boosting government spending,” Callaghan said. “I tell you in all candor that that option no longer exists, and insofar as it ever did exist it only worked by [...] injecting a bigger dose of inflation into the economy.”

Let the market decide

The new economic orthodoxy, endorsed by US President Ronald Reagan and British Prime Minister Margaret Thatcher in the early 1980s, held that the market, not government policy, should set the unemployment rate. The market would guarantee that all genuine jobseekers would find jobs at “market-clearing wages,” and if people declined work at those wages, that was their choice.

If the resulting level of unemployment was deemed socially unacceptable, the answer was to improve incentives to work by weakening labor unions and reducing unemployment benefits. With fully flexible labor markets, unemployment would automatically be at its “natural” or equilibrium rate – the rate at which inflation had no tendency to rise or fall. In fact, the notion of involuntary unemployment disappeared entirely.

What survived from the wreckage of Keynesian fiscal policy was the much weaker default position that monetary policy (if managed by an independent central bank) could do all the stabilizing that a market economy needed, but without the constant temptation to print money for political purposes. To justify discretionary monetary policy, the neoclassical (and anti-Keynesian) full-employment assumption was relaxed to allow for the effectiveness of a short-term monetary stimulus. This so-called New Keynesianism owed little to Keynes, but by allowing central banks to vary their official lending rates in pursuit of their inflation target, it did at least keep a tiny space open for “monetary” policy to influence the real economy.

Monetarists are in retreat

The new dispensation, however, ran into problems of its own. Between 1980 and 2008, average unemployment increased from its “golden age” average of 1.6% to 7.4% in the UK, from 3.1% to 7.5% in Germany, and from 4.8% to 6.1% in the US. Underemployment – people forced to work fewer hours than they wanted – was rising, poverty and inequality increased, and economic growth slowed.

The Great Recession that started in 2008 reinforced the Keynesian lesson that the private sector, beset by uncertainty, could not continuously deliver full employment. Moreover, the failure of monetary policy to prevent this economic collapse or bring about a durable recovery had discredited the monetarist approach to economic management.

These neoliberal failures led to a revival of interest in fiscal policy. An important early initiative was the 1978 Humphrey-Hawkins Act in the US, which “authorized” the federal government to create “reservoirs of public employment” to balance fluctuations in private spending. The reservoirs would deplete or fill up as the economy waxed or waned, creating an “automatic stabilizer” without any “management” of the business cycle. But the Humphrey-Hawkins Act was never implemented. It was the last gasp of New Dealism.

Following the 2008 recession, radical policymakers had to undergo what Pavlina R. Tcherneva of Bard College called a “fundamental reorientation” if they were to achieve “true full employment, long-term macroeconomic stabilization, better income distribution, and improved socioeconomic outcomes.” The JGP is the attempt to do just that.

False constraints

A JGP, the most radical of the post-Keynesian responses to the failure of neoliberal policies, represents a fusion of two lines of thought. The first asserts government responsibility for full employment, but in the form of a direct job guarantee rather than by managing aggregate demand. The second, based on MMT, contends that “deficits don’t matter” in the pursuit of full employment.

MMT attacks the orthodox view that governments, like households and firms, can spend only as much money as bondholders or taxpayers allow them. Orthodox theory identifies two types of “fiscal constraint.” The first is financial: the size of the government deficit is limited by the willingness of bondholders to lend the government money. This constraint bites hardest in a downturn, when the government deficit automatically grows.

The second constraint is said to arise from the limit of “real resources.” As stated by University of Chicago economist John Cochrane in 2009: “Every dollar of increased government spending must correspond to one less dollar of private spending.” The clear implication, as I noted in my 2010 book, Keynes: The Return of the Master, is that “jobs created by stimulus spending are offset by jobs lost from the decline of private spending.” This constraint is portentously called Ricardian equivalence.

The two constraints, taken together, mean that governments have only a limited amount of “fiscal space” for employment policy. Either the interest on public debt will go up, or the private sector will increase its saving in expectation of higher taxes. This was the intellectual foundation for austerity policy after 2008. It boiled down to the proposition that the faster the government cut its spending, the quicker the economy would recover.

But such “crowding out” arguments are deeply flawed. It was nonsense to believe that in 2008-09, with the global economy in free fall, every dollar of increased government spending on employment was a dollar subtracted from private spending. And it is odd to claim that a sovereign can run out of money that it produces itself. A government is not like a private firm or household that has to go to the bank manager for a loan. It has its own bank.

A JGP, conjoined with MMT, is thus an attempt to re-establish the case for fiscal policy, not as an emergency measure in a slump, but as a permanent part of employment provision. Like Friedman, MMT envisages “helicopter money,” scattered by the government over a parched landscape. But it deviates from monetarism by insisting that extra money affects economic activity not by being printed, but by being spent. As Kelton has said, “the helicopter can drop money, gather bonds or just fly away.”

The only surefire way to guarantee that new money is spent is for the government to spend it. That is why MMT supporters see the theory as part of fiscal rather than monetary policy. It uses a version of the quantity theory of money championed by Friedman to support the case for the sort of fiscal intervention that he abhorred.

Some Keynesians who reject orthodox “crowding out” arguments also consider MMT arguments too extreme. Specifically, they would not dismiss questions of “confidence” as cavalierly as do modern-monetary theorists. As Keynes himself noted, “Economic prosperity [in a capitalist society] is excessively dependent on a political and social atmosphere which is congenial to the average businessman.” Substitute “financial markets” for “average businessman” and you have the world as it is, not as MMT theorists would like it to be.

The six-point guarantee

For starters, MMT’s target is not aggregate demand but labor demand, because it is much easier to target employment than output. Output and employment converge only in the very short run: one can have jobless recoveries. And it is very difficult to calculate output gaps and multipliers. The direct job guarantee eliminates such problems. Further, by easing workers’ transition to private-sector jobs as the economy recovers, a JGP helps to relieve the shortage of skilled labor that can produce inflationary pressures when the economy is below full employment.

Second, JGP advocates attach great importance to Keynes’s emphasis in 1937 on the “need today of a rightly distributed demand [more] than of a greater aggregate demand” (my emphasis). A satisfactory average level of employment can be consistent with some parts of the economy overheating and others having high unemployment. A JGP can thus be used to influence the distribution as well as the level of employment – the aim of the Green New Deal, for example.

Third, and consistent with this, JGP advocates say that their programs, while federally funded, would be administered locally by various entities, including municipal governments, NGOs, and social enterprises, which would seek to create employment opportunities where they are most needed, matching unfulfilled local needs with unemployed or underemployed people. To ensure that projects can be rolled out on demand, job banks and job centers would maintain inventories of communal needs under three main headings: care for the environment, care for the community, and care for people. Addressing looming environmental threats, for example, could generate millions of public-service jobs for years to come.

Fourth, the public-sector job pool would be a buffer stock of labor that could expand and contract automatically with the business cycle, thus eliminating discretionary fiscal policy. Such a buffer – an idea introduced by the Australian economist Warren Mitchell – would be a more powerful automatic stabilizer than unemployment benefits, because it maintains demand more reliably. And although replacing unemployment benefits with wages would cost more money, it would deliver a more powerful economic boost. A job pool would also better maintain workers’ employability, and could easily be coupled with on-the-job training – an important contributor to economic recovery and growth prospects generally.

Fifth, JGP workers would be paid at a fixed rate, which the government can set at any level it chooses above that of unemployment benefits. A fixed wage sets a floor for private firms’ wages, even without minimum-wage legislation, since private employers would always have to pay at least the JGP wage to attract workers; and in periods of strong private-sector demand, they would have to bid for scarce labor at above the JGP wage. The availability of employable JGP workers at a lower wage would restrain inflationary wage demands.

Finally, MMT provides monetary backing for the proposed fiscal policy. The basic idea is that because sovereign governments typically have a monopoly over the issuance of their own currency, they can issue money whenever they want and thus never face the threat of bankruptcy. “In reality, then,” MMT advocates claim, “the size of the deficit is irrelevant.” For monetary sovereigns, therefore, the composition of government liabilities is of no importance. It doesn’t matter whether the central bank is “printing money” or the Treasury is “printing bonds,” because the central bank will always ensure the right amount of money to support its targeted interest rate.

When JGP meets MMT

What has been most shocking to orthodox economists is MMT’s claim that a sovereign government does not have to “finance” its spending by issuing debt or raising taxes. This claim has its proximate roots in Abba P. Lerner’s theory of “functional finance,” which rests on three key principles.

First, governments should aim for the right amount of total spending to prevent unemployment and inflation, and not worry whether their policies conform to any longstanding economic doctrine about what is sound or unsound. Second, policymakers should adjust the issuance of money to the desired rate of interest – meaning the one consistent with the desired rate of investment. Finally, the government should print, save, or destroy the money needed to implement the first and second principles.

On this view, governments should pay for their spending by issuing their own money, and borrow from the public or raise taxes only in order to “drain” the economy of excess money. The only constraint governments face is inflation, and not the fiscal limitations emphasized by orthodox economists.

The inflation constraint is binding because it limits the amount of extra employment the government can generate. If a government spends too much money, it will no longer be able to buy real resources, because its extra spending will merely raise the prices of all the resources (including labor) that it buys. Only when the economy can no longer deliver more real goods and services to the state can the government be said to have “run out of money.” As Kelton, Andres Bernal, and Greg Carlock put it, “inflation isn’t triggered by the amount of money the government creates but by the availability of biophysical resources that money tries to go out and buy – like land, trees, water, minerals, and human labor.”

JGPs and MMT, although logically independent, support each other. By linking them, advocates can make the politically appealing claim that the pursuit of full employment is in no way constrained by the vested interests of bondholders or taxpayers.

At the same time, however, Keynes’s comment on Lerner’s “functional finance” approach is worth quoting: “His argument is impeccable,” Keynes said. “But, heaven help anyone who tries to put it across the plain man at this stage of the evolution of our ideas.” And the notion that government “prints its own taxes” sounds even more paradoxical to the average citizen after a half-century of neoliberal propaganda.

What the critics say

Critics can be divided between those who object to a JGP per se and those who object to MMT, the macro theory supporting it.

A familiar criticism of JGPs is that they generate quasi-employment – rather like Keynes’s example of digging holes and filling them in again. Local work programs, critics claim, offer “pretend” work that does not really need doing. In response, JGP advocates such as Tcherneva have outlined the type of jobs that a JGP could provide. A JGP, she says, will address the urgent “environmental and care needs of communities” – for example through a Green New Deal, initiatives in the arts, youth apprenticeship programs, child and elder care, and special-needs programs for veterans, at-risk youth, and former inmates. The essential point is to identify “needs gaps” and devise national and local programs to fill them.

Others have objected to the potentially punitive nature of a JGP. If a JGP job is alternative to unemployment benefits, it seems logical that a JGP entitlement should replace the unemployment benefit entitlement. In practice, it is already widely accepted that unemployment benefits should be time-limited. The problem has been a lack of alternatives to unemployment. The JGP fills this gap. The only real issue is how much time on benefits should be allowed for job searching.

A much stronger theoretical attack is directed at MMT. Harvard luminaries Kenneth Rogoff and Lawrence Summers have claimed that, by denying the existence of a fiscal constraint, MMT is inherently inflationary. They have a point, of course: it seems naive to believe that taxes can be increased whenever necessary to stop inflation. But their argument would have been more interesting had they attacked MMT not for “voodoo economics” (Summers) but for “voodoo politics.” To say that there is no theoretical reason why sovereigns need to “apply to the people” for money is correct, but it fails to consider why the rules of “sound finance” were invented in the first place. The rules were put in place precisely to prevent governments from spending money at will. That governments should apply to parliaments (ultimately, “the people”) for “supply” is part of the theory of the limited state.

For his part, economist Thomas I. Palley accuses MMT economists of lacking a class perspective, and criticizes MMT’s attribution of economic crisis to financial instability. The root cause of financial instability, Palley argues, is the skewed distribution of wealth and income, which generates under-consumption. Although money-financed deficits have helped to reflate asset prices since 2008, they have not addressed structural problems, and will thus lead to new crises.

In a reply to critics of MMT, economists Scott Fullwiler, Rohan Grey, and Nathan Tankus make four points. First, excess demand should be controlled by tightening credit conditions as well as by raising taxes. Second, speculative (supply-side) sources of inflation should be met by regulation and price controls, which will shift some of the task of controlling inflation to other government agencies. Third, assessing the inflationary impact of new spending proposals requires a disaggregated approach that considers variations in demand between sectors and regions. Moreover, budgetary policy should address not just the level but the distribution of output. Fossil-fuel industries, real estate, defense, and financial services are either too large or too dirty, or eat up too many natural resources, and should be curtailed. Fourth, governments should not slap on taxes suddenly when inflation appears: they should progressively raise taxes and tighten lending as needed.

The duty of care

The case for a JGP hinges on three interrelated propositions. The first is Keynes’s contention that the market for labor cannot be expected to clear continuously even if wages are perfectly flexible. The second is that, if that is the case, it is the state’s responsibility to fill the employment gap as it arises. The third is that the best way for it to do so is to provide a direct employment guarantee. The first is a matter of economic theory, which challenges the microfoundations of orthodox economics. The second is a matter of political economy. One can argue that even if the first is true, some unemployment may be a price worth paying to restrain inflation or prevent an undue expansion of government intervention. The third is a matter of technique: which of the many possible policies for maximizing employment is most effective? The main question concerning MMP is political: is this the best way to present the Keynesian argument that the aim of fiscal policy should be not to balance the budget, but to balance the economy?

On these matters, my view is the following.

First, the contention that the market system cannot guarantee continuous work for all those who want it seems to be indisputable. An unmanaged market system has a built-in tendency to underactivity; the growing financialization of economic life adds to the instability of employment, for reasons pointed out by Hyman Minsky.

Second, I believe that maintaining full employment (which includes preventing unwanted unemployment) is one of the chief responsibilities of modern government. I am persuaded that a direct employment guarantee is better targeted to this purpose than discretionary policy.

Third, I am not persuaded that MMT is the best way to challenge the orthodox theory of sound finance. Though Michael Kalecki was no doubt right to point out in 1943 that the social function of sound finance is “to make the level of employment dependent on […] the confidence of those with money,” MMT is too dismissive of the constitutional case for limiting government spending. This does not mean these limits should be set by flawed theories of orthodox budgeting: the “bonds of revenue” are more elastic than either orthodox theory or MMT suppose.

Although the JGP debate involves questions of economic theory, the main issue is one of politics – the size of the state, its proper economic role, and the resources it needs.

Naturally enough, the wealthy and powerful have generally preferred a limited state that leaves them free to do what they want with the money and power they control. The poor and powerless, by contrast, have looked to the state to protect them against the insecurity of the market and the predations of the wealthy. Keynes was on their side. He was right to say that unregulated capitalism guaranteed neither full employment nor an “equitable distribution of wealth and incomes.”

There are undoubtedly problems with the design and implementation of a JGP. But I applaud its spirit and intentions. A government’s job is to protect citizens from misfortune, and want of work is the greatest misfortune a population can suffer outside of war and natural catastrophe. An elite that abandons this duty of care, on the spurious grounds that people “choose” their own level of employment, deserves to be cast out. FDR understood this, and so did Keynes. For the first time since the collapse of the Keynesian revolution, it is being suggested by serious politicians that a government has the moral and financial responsibility to maintain full employment. That is the best argument for giving a JGP a fair trial.