The Green New Deal is a radical proposal to tackle the global challenge of climate change and reshape the American economy at the same time. It's defining the US policy agenda for 2019, and spilling over to places like Australia. The first of its principles is that every citizen should be guaranteed a job with a living wage. And it might be funded using an approach to public finance called Modern Monetary Theory (MMT), instead of raising taxes.

Both of those ideas—a job guarantee and MMT—have recently been prominent in their own right. They were given a boost by the advocacy of Stephanie Kelton, economic advisor to the Bernie Sanders 2016 campaign, so people have already debated what their possible effects might be. But not much attention has been paid to the fact that they've effectively been tried before. And their results, in practice, were not good.

MMT versus modern monetary practice

MMT is hard to pin down precisely. Paul Krugman unkindly calls it Calvinball, the game where the rules change whenever you like, and Matt Bruenig wonders if the obfuscation is a deliberate tactic.

The gist of the theory is that tax policy should be uncoupled from spending decisions, and that any resulting inflation can be managed through fiscal policy. Consequently, any new public policy should be judged by whether or not it's a good idea, not by whether or not we could afford it.

Since MMT became popular, politics has changed for the better. It's helped to give left-wingers the confidence to propose bold ideas, and to brush off the bad-faith complaints about deficits coming from their right.

Please take a minute to appreciate how amazing that is. Since the financial collapse more than a decade ago, every centrist economic technocrat has been saying that budget deficits are a low-priority issue that can safely be ignored for a while, but nobody seemed to care. When Australia reacted to the GFC with aggressive fiscal policy, conservative politicians talked ad nauseam about a "debt and deficit disaster", at a time when public debt was literally zero.[1] "LOL," commented the central bank governor, to no effect. Across the Pacific, experts continually raged about Paul Ryan's phoney concern for budget balance throughout his time as Speaker, but nothing budged.

At last, AOC and her supporters have shifted the public debate by giving the "how will you pay for it" question the lack of respect it deserves. That is a genuine achievement by MMT, which should not be diminished by the mere fact that the theory doesn't work.

For the sake of argument, though, what would happen if the Green New Deal put MMT into practice? Here's the problem in a nutshell:

Deficits and inflation don't matter.
Therefore,
Deficits and inflation never matter.

Everybody agrees on the premise, but the fallacious reasoning to the conclusion is the tricky part. That may seem like an uncharitable summary of MMT, because future governments are meant to handle any inflationary problem by increasing taxes. But that hand-waving towards the future is utterly implausible, because of how things played out 50 years ago.

LBJ wanted to fight the war in Vietnam at large scale, and also launched a war on poverty at home through the Great Society programme. To pay for either of them—not in the literal sense of handing over money, but in the real sense of raising taxes and ultimately causing resources to be diverted from what they would otherwise be used for—would have been a massive undertaking. But LBJ wanted to do both of them, and because inflation and the deficit were low he had the technical capacity to take action and postpone the consequences.

I once came across a vignette in which some of LBJ's economic advisors, holdovers from Camelot, were trying to show him a PPF diagram and explain about tradeoffs, like the choice between producing guns and butter. Then some urgent news from Vietnam arrived and LBJ swept out of the room, pausing only to announce "I want guns and butter," and that was the end of their discussion. At least, I could have sworn I read that somewhere, though it may have been a grad-school fever dream. Here's the closest thing I can find that makes the same point:

Question: how will we decide how to balance these priorities? Answer: yes, we will.

With no major changes in tax policy, something had to give, so the debt was monetized and the US had high inflation for the next decade or two. To reiterate, LBJ's successors could have solved the inflationary problem by tightening up policy, as MMT asserts, but they didn't want to.

Is inflation bad?

If the main strike against MMT in practice is that it's likely to cause inflation, then maybe that's fine. Compared to persistent unemployment, growth in prices seems abstract and painless, especially now that we've had more than 25 years of low inflation.

When you ask why inflation is bad, many of the stock answers only appeal to the wealthy. It discourages investment; degrades the value of capital; penalises savers; and such and such. Those may all be true, but with wealth inequality on the rise, maybe it would be good for the rich to share some of the pain.

The most cogent reason to avoid inflation is that you'll need to cause unemployment to fix it. Policymakers can choose any of several different methods to lower inflation, but they all shrink the level of aggregate demand.[2] That's common ground with MMT advocates and neoliberals.[3] In other words, when inflation accelerates until it's too high to bear, then whatever the unemployment rate is, you need to put more people out of work involuntarily.

Not that people didn't try to avoid that. The 1970s were awash in price controls, wage controls, collective action plans, subsidies, self-sufficiency drives, and other ideas to escape the malaise. For instance, Gerald Ford's Whip Inflation Now scheme flailingly gestured to a wartime spirit of sacrifice as it urged people to cut kitchen waste, carpool, take the bus, and grow their own vegetables. "Collective obligation, in the key of homespun earnestness," as Rick Perlstein put it. It didn't work.

The extensive range of W.I.N. merch was an inspiration for Buzzfeed. (Image: Wikimedia)

Every government of the 1970s baulked at making unemployment higher than it already was. And why wouldn't you? It would have caused pain and anger, and all the credit would have gone to whoever took over after you were turfed out of office. The status quo was also terrible, but you have to play the hand you're dealt. Plaster on a shit-eating grin, give a thumbs up to the camera, and pray that commodity prices fall. (They did, but too late.)

Having given up using fiscal tools to reduce inflation, policymakers fumbled their way over to the monetary side. They experimented with monetary aggregate targeting, current account targeting, a monetary "checklist", and several other variants, without much success. The answer finally arrived in the early 1990s with the New Keynesian enlightenment. Since then, OECD economies have mostly had independent central banks with a policy of inflation targeting. It may not solve every economic problem, but it solves that one.

In the New Keynesian framework, inflation stays low because everyone knows that the central bank will react strongly if it ever gets too far out of whack. That's why central bankers are obsessed with inflation expectations and policy credibility—they get most of their power from signalling what they would do. Provided that everyone believes them, they rarely have to actually do it. "Why should we care about 5% inflation?" people sometimes ask, and here's the answer. If the central bank lets it slide, then 5% will become the norm, and we'll be wondering about 8% inflation, and so on, and so on. By reacting early, they can stop that ratchet effect from ever getting started.

Is hyperinflation bad?

"You can't solve a problem by just printing money," say people who've never studied macro. It seems like a common-sense idea, but it's wrong.

But I think, in the case of MMT, the common-sense reaction might actually be correct. MMT advocates scoff at comparisons to Zimbabwe and Weimar Germany, but I'm not sure they're entitled to, if their hope of preventing runaway inflation is a policy change by hypothetical future governments.

Of all the leaders who presided over a hyperinflation, not one of them wanted to start one. They began by implementing a fiscal programme which they couldn't pay for directly, choosing to monetise public debt instead. As inflation accelerated, and inflation expectations began to run ahead of it, they opted not to inflict the painful measures that might have stopped it because they lacked the political capital, or because they had other priorities.

If a large-scale programme is put in place with MMT, it's easy to imagine it ending either in a hyperinflation, or with a conservative electoral mandate for neo-Thatcherite austerity. If there's a plausible third alternative, I'm yet to hear it.

Implementing a jobs guarantee

Meanwhile, a jobs guarantee sounds like the kind of ambitious policy that is "capable of uniting both the winners and losers of globalisation", as Tim Ayres put it. There are people in their prime working years who live in towns where the single large employer has shut down. Why not figure out a way to give them a job where the market has failed?

The trouble comes when you start trying to spell out exactly what the goals of the programme should be. Do we want to offer people the minimum wage to do something at any time, regardless of how useful it is? Should they be higher-quality jobs with higher-than-minimum wages, to counterbalance precarious casual work? Or should the wages be barely enough to survive on, in effect a work-for-the-dole scheme?[4] Trying to straighten out these conflicting priorities generates not just a dilemma, but what Australian researcher Hugh Sturgess (quoted by Matt Bruenig) has called an "impossible quadrilateral".

Even if those tensions could be resolved, implementing a jobs guarantee would be another matter. Advocates keep repeating a phrase: the scheme would be "federally funded but locally administered". That sounds reasonable at first, but I've come to think that it's too glib, a way of ducking the issue.

Let me describe a particular jobs guarantee programme, and see if you can spot any problems. (I mean, obviously this is going to be a trap, but please play dumb for a minute.) The objective of the scheme is to generate jobs for people in a depressed area, and it's forecast to create about 3,000 of them. It's funded by the Federal government, with an ample budget already agreed on, and will be administered locally. There is no political controversy about the plan, and minimal publicity. Local community and business leaders agree that it's a good idea. Would you vote for it?

I was describing a pilot programme by the US's Economic Development Administration, part of LBJ's Great Society, to refurbish the port of Oakland. It wasn't explicitly a Job Guarantee, but it was close enough—a federally-funded, locally-administered intervention by the state to offer jobs to workers who otherwise wouldn't have them—and I think it illustrates what can go wrong in practice. Congress allocated $23 million to it in 1966, but by 1969 only $3m had been spent, and that most of that went to a highway that was already planned. Extra jobs created: zero.

The failure of that pilot study is analysed in detail by Pressman and Wildavsky's Implementation, or to give it its full title, Implementation: How great expectations in Washington are dashed in Oakland; or, Why it's amazing that Federal programs work at all, this being a saga of the Economic Development Administration as told by two sympathetic observers who seek to build morals on a foundation of ruined hopes.

In reality, at the point of spending the money and hiring the people, the local administrators of this job guarantee couldn't agree on what it was supposed to do. Some wanted to insist on hiring local workers, no matter their skill level; others wanted to hire anyone who could do the work, no matter where they lived. But Pressman and Wildavsky's main finding is that there were just too many people involved, too many decision-makers, too many interest groups to persuade and align.

Would any of those problems be solved if a jobs guarantee were put in place in 2019? Possibly, but I doubt it. One can imagine finding ways around all those blockages for a particular investment project, but to have a hope of succeeding in its goals, a job guarantee scheme would need to figure them out for every project, in every town, across the whole country.

To be clear, none of this implies that fiscal policy is futile. Just that it's probably better to keep investment and transfers separate. For public infrastructure and other projects that we need, it's best to focus on getting them done quickly and efficiently; when we want to affect the whole labour force, to spread the benefits of globalisation more fairly, it seems like a Universal Basic Income would be more likely to work.

Back to the future

There's a huge appetite for bold public policy at the moment, and for a larger public imagination. That's wonderful, particularly while the ideas coming out of the sensible centre are numbingly boring.

Still, history suggests that a job guarantee is the wrong goal to stretch for. And that the social-democratic agenda will be more effective if it's ultimately paid for in some way, in the sense that activists persuade people that the objectives are important enough that we should be willing to make real sacrifices to get them.

If we pay more attention to the ruined hopes of the 1960s, then I think schemes like the Green New Deal will have a real chance at succeeding.


Cover photo: an abandoned Futuro House near Royse City, TX. Photo by Steve Rainwater on Flickr.

1. Net public debt actually became negative a couple of years before the crisis. The government decided to keep gross debt at a stable positive level, to ensure that bond traders' skills didn't atrophy.

2. To be pedantic, you can also change the supply side of the economy through deregulation and labour market reform. But OECD economies already did that 30 years ago, and you can't do it twice. Plus, those effects show up over years, not months.

3. I think. Judging by Stephanie Kelton's description of IS-LM 🤔, maybe not though.

4. On The Breunigs podcast, Matt Breunig made two more interesting points about a job guarantee, which don't fit as neatly in the dilemma framework. First, the kinds of work which spring to mind for this type of scheme tend to be highly cyclical, which is self-defeating. For instance, we need more childcare workers, but mainly during booms, which is exactly when people are less likely to want and need a job guarantee. Second, the talk about job guarantees has a suspicious bias towards physical labour that produces physical stuff, when more than 80% of a modern economy is in services. If we're serious about equipping these workers to succeed in a 21st century workplace, then a workfare programme should consist of something like requiring people to upload 3 video reviews of NBA games to YouTube per week.