Stabilization policy with ignorance

Eric Lonergan’s good discussion of fiscal rules has triggered an intemperate debate about MMT. For me, though, it poses a question: from which fact(s) should thinking about macro policy begin?

For old-style Keynesians, the fact was that labour markets could not be relied upon to clear. For new classicals, it was that people were rational maximizers. I want to suggest a different fact – that recessions are unpredictable.

Prakash Loungani has shown that both private and official forecasters “miss the magnitude of the recession by a wide margin until the year is almost over”. In February 2008 for example – after the recession had begun – the Bank of England attached only a slight probability to GDP falling when in fact it fell more than 6% in the following 12 months*.

Exactly why this should be is a question for another day: I suspect it’s because recessions have micro-level origins which are magnified by network (pdf) effects between heterogenous firms.

Whatever the reason, this Big Fact has a devastating implication. It warns us that there are severe limits upon what demand management can achieve, whether it be through fiscal or monetary means. This is simply because there’s a lag between policy and outcome. Fiscal or monetary policy can boost demand once it has fallen. But it cannot reliably prevent all recessions. This point applies to conventional as well as unconventional policy. By the time the Bank of England knows there is a sufficiently deep recession to require (say) a helicopter drop, the recession will be well under way.

How can we respond to this?

We could follow Robert Lucas and say it’s a non-problem because the welfare costs (pdf) of economic fluctuations are trivial. Most of you, though, would resist this, and I think (pdf) rightly so.

Or policy-makers could reject conventional macroeconomic forecasting and place more weight upon the yield curve (pdf) to tell us the probability of recession. This, though, is inviting us to be smacked in the face by Goodhart’s Law.

This leaves a third response: we need better automatic stabilizers.

The obvious possibility here is more progressive taxation and a higher welfare safety net. Another possibility is for Shiller-style macro markets to enable those most exposed to the risk of recession to buy insurance: this requires state support for such markets. A third possibility is one suggested by Eric: a sovereign wealth fund. Insofar as this holds overseas assets, it would help diversify local UK economic risks, and would rise as sterling falls.

It’s in this context that a Job Guarantee becomes attractive (pdf). It would provide schemes which expand to create jobs in recessions, and contract as private sector jobs are created in the subsequent upturn. It’s another automatic stabilizer.

Now, there are issues with a JG, not least of which are administrative ones: how do we identify what work needs doing? How can we ensure a JG doesn’t deteriorate into demeaning workfare under rightist governments? And so on. Such problems are a case for thinking about a JG now, and for rolling it out in good times, so that difficulties can be addressed before the scheme needs to be scaled up.

We should regard a JG not as a faddish idea of some sect – although that it how it is sometimes presented – but rather as one of a suite of possible policy responses to one of the fundamental facts of the human condition: our inability to reliably foresee the future.

* It’s unclear that heterodox economists foresaw the 2008-09 recession; permanent warnings of the risk of recession are of limited use for the purposes of macroeconomic stabilization.

Comments

[e]

Great post. All Labour factions should be leaning in to the space this debate is developing. Job guarantees in conjunction with lifelong learning, renewal of FE/technical colleges, automatic stabilisers that don't damage our future prospects.

[Jim]

Suggesting the UK should have a Sovereign Wealth Fund is a bit like telling someone who is in debt and whose spending exceeds their income by some considerable margin that what they need is a savings account with lots of money in it.

[Ed Zimmer]

"there are issues with a JG, not least of which are administrative ones"

Administrative cost and complexity is a significant deterrent to a workable Job Guarantee program. The April 2018 Levy Economics Institute proposal allocates almost 50% of program costs to administration. Just out of interest, I took a pass at a less bureaucratic approach (TENonline.org). Its features:

1) Bottom-up administration. Local government knows what work the community can best benefit from — and this JG program offers a free labor pool for them to make best use of. Rather than "jobs" being defined by State or Federal (or Non-Profit) bureaucrats for people to fill, "work" is made available by the local government for people who need (and wish) work to choose from. Being local, the character and requirements of the work offered can evolve via interplay between the community and those in the local government. Some of that work may be helping local government employees do a better or more complete job, but other can be work the community perceives as needed yet is currently un- or haphazardly-done, for example, providing better care for the young or elderly in the community.

2) Maximum automation. Program activity is via a National website, divided by State into Local government pages. Each Local government page displays the JG work available locally with communication tools that facilitate interviewing for the work with the applicable manager and payment for the work if hired. The National website is integrated with the existing consumer banking system so that payments are automatically credited to workers' JG debit cards for spending as soon as the day following the work. Inflation fears are inapplicable as those choosing JG work will be spending, by necessity, with local businesses coincident with receipt of payments.

3) Minimum overhead. No program costs are incurred by Local government as their participation is from self-interest and constitutes only an added duty for existing employees. No program costs are incurred by State government as their function is limited to deciding which Local governments are eligible for participation. Program costs incurred by the National government involve setting up and maintaining the National website and integrating it with the consumer banking system. This can be done internally or by contracting it out. I favor contracting it out through competitive bidding with preference given to a small entrepreneurial firm that can demonstrate the required level of security with an ability to work innovatively with Local governments.

4) Orderly growth. Local governments choose whether and when to participate in the program. The most needy and daring Local governments will be the first to participate. This gives time to fine-tune the website display formats and communication tools to best fit the needs of the Local users as more Local governments choose to participate.

[Ralph Musgrave]

I like Ed Zimmer’s above claim that “Local government knows what work the community can best benefit from — and this JG program offers a free labor pool for them to make best use of.” Just one slight problem there: Bill Mitchell, one of the most ardent advocates of JG actually did a survey of local governments to see how many jobs they might be able to create for low skill JG people, and the answer was “precious few”.

Lefties may not be comfortable with this, but the fact is that it’s the PRIVATE sector which is good at employing the less skilled, not the public sector. Plus the empirical evidence is that private sector JG type jobs result in better subsequent employment records than training or public sector JG jobs. See:

https://ideas.repec.org/p/iza/izadps/dp606.html

That is one reason why in this work I advocate JG work in both sectors:

http://kspjournals.org/index.php/JEPE/article/view/1237

[Christopher H.]

Yes, we should do all of the above. A JG and also a social wealth fund with a basic universal dividend:

https://www.peoplespolicyproject.org/projects/social-wealth-fund/

[Scurra]

I'm all in favour of a JG. My problem isn't with the idea, it's with the obvious fact that this is a spreadsheet solution that treats people as being interchangable "assets" like machinery or capital.
Which is fine, but one would need to be honest about that. Say up front "we're not interested in you as an individual, merely as your contribution to our balance sheet" and at least the cynicism would be honest.

I write this as someone who has a disability that makes certain types of work almost impossible for me. There is a serious risk that a JG would simply say "here's your new job, be grateful for it", with no consideration for anything beyond that.
In fact, what has happened is that I have become acutely aware that *everybody* has these problems, it's just that many people find it possible to compromise (for whatever reason.)

And that's why JG schemes cost, and why they should cost. It's not just about a Job Guarantee, it's about the *right* Job Guarantee.

And, of course, this is the same problem that the health service, the eduction system, the criminal justice system etc. etc. all have, to varying degrees. They are all forced to produce spreadsheet solutions because accepting that individuals need individual solutions is an immense and terrifying prospect.

[Blissex]

«that recessions are unpredictable. Prakash Loungani has shown that both private and official forecasters “miss the magnitude of the recession by a wide margin until the year is almost over”.»

As JK Galbraith wrote extensively in "The Great Crash 1929" almost nobody wants to predict a recession. In particular it is hard to imagine a sell-side Economist wanting to talk down "confiodenc"prosperity".
Perhaps there are buy-side people who can predict recessions, but of course they are going to trade on it, so there is little incentive to go public.

As to policy, the pious hope that any government-related entity might publish a forecast of recession is quite ridiculous, and so is the expectation that they would forecast something else than a short duration and shallow depth for one that has already manifested.

[Hugh Ferguson]

The link https://www.iadb.org/intal/intalcdi/PE/2013/12532.pdf seems to be broken. Could you correct it? Thanks

[Hugh Ferguson]

Likewise http://fperri.net/PAPERS/ESWCKMP_latest.pdf rather than https://www.fperri.net/PAPERS/ESWCKMP_latest.pdf?

Not giving you a hard time, just had trouble downloading them. It's hard to pass up a free paper!

[acorn]

Think first link on network amplification should direct to:
https://publications.iadb.org/bitstream/handle/11319/4198/Banking%20Crises%20and%20Financial%20Integration.pdf?sequence=1&isAllowed=y

[chris]

Sorry about that Hugh. The links should work now.

[Luc Hansen]

I love your asterisk note. As an Uber driver in my retirement, I was explaining exactly this to an interested fare, who had noted my (hobby) econ quals in my profile, only yesterday. I was also asked if we were headed for recession, which gave me the opportunity to trot out the old standby of OTOH...OTOH, after which the fare correctly noted, "You mean you don't know."

[B.L. Zebub]

@Chris

You might enjoy this post by Peter Cooper (although he prefers the "heterodox" label to describe himself, I think a mix of MMTer/Marxist, or Marxist/MMTer wouldn't be far off the mark):

http://heteconomist.com/the-income-expenditure-model-with-a-job-guarantee/

Unlike the overwhelming majority of internet "experts" and comment-thread pundits, Peter is educated, intellectually honest, and reasonable.

[Peach]

I am not that knowledgable about MMT. How would a JBG differ from standard welfare benefits as automatic stabilisers?

I am in favour of a job guarantee because it would lower the social costs of unemployment, reduce hysteresis, increase the bargaining power of labour, and slowly limit the stigma attached to unemployment, but as ever I think the devil is in the detail.

[Peach]

@Ed Zimmer What are you talking about? "Inflation fears are inapplicable as those choosing JG work will be spending, by necessity, with local businesses coincident with receipt of payments."

Your comment above makes no sense. JG payments presumably wouldn't be inflationary because they kick in during downturns which are deflationary.

I like the idea that local governments know how best to utilise local people but have you ever met local government employees? Roads to nowhere are the likely outcome.

[Blissex]

«How would a JBG differ from standard welfare benefits as automatic stabilisers?»

Presumably the laws about minimum wages tax credits etc. would apply, so it would be much higher than the JSA.

[Alex]

Ed Zimmer's comment is similar to the Future Jobs Fund, a Gordon Brown-era programme that amounted to a small-scale JG trial and that did very well indeed in impact evaluation, before being cancelled by the incoming coalition.

[Robert Mitchell]

Shiller's idea is the best. Have the Fed construct an External Finance Premium index and sell shares based on it. Then you can buy panic insurance. The Fed is going to print money to end a panic anyway. Selling insurance beforehand makes it explicit and decreases uncertainty.