Monday, October 12, 2015

People's QE Has Been Tried Before and Failed

I am seeing more and more people get excited about "People's QE", the brainchild of UK labor party leader Jeremy Corbyn. For example, Roger Farmer sees it as similar in spirit to his own preferred approach, Ambrose Evans-Pritchard says "it is exactly what the world may soon need" and Matthew C. Klein argues "the core idea is sound and has an impressive intellectual pedigree." With endorsements from such thoughtful people, this QE must be something special. So what exactly is it?

The People's QE is a program where the Bank of England (BoE) would engage in large scale asset purchases of debt used to finance investment spending in infrastructure. The issuers of the debt would not be the central government, but local governments and other agencies in the UK that fund investment spending. One advantage of this approach, according to its advocates, is that it would be politically easier to implement since it would not explicitly create bigger budget deficits (even though implicitly it would be doing so). More importantly, supporters argue this form of QE would send the newly created money directly to people and institutions that actually spend the money. More bang for your  buck! So what could go wrong?

A lot, actually. For this approach is nothing more than a monetization of debt--a helicopter drop. It is widely recognized that helicopter drops will have no effect on aggregated demand if the monetary injections are perceived as temporary. And monetary injections will always be perceived as temporary without a credible commitment from the government to reflate the economy. In practical terms, this means a helicopter drop needs to be accompanied by a higher inflation target or a price (or NGDP) level target high enough to create some reflation. Otherwise, the helicopter drop will be all for naught. But don't take my word for it, ask Paul Krugman or the list of notable economists found here.

To  make this understanding concrete, let's imagine Jeremy Corbyn becomes Prime Minister and the BoE engages in People's QE with its current 2% inflation target. Investment spending would initially begin to grow, but quickly run up against the inflation target and force the BoE to either pull back on or start sterilizing its purchases. The UK economy might eke out a few more basis points of aggregate demand growth, but not the kind needed for full employment or envisioned by advocates of the People's QE. Contrary to claims of its advocates, then, QE would not pack more of a punch simply because the money went directly to infrastructure spending. No matter who spent the money, the economy would run up against the 2% inflation constraint. This is what I have called the Penske Problem

The hugely under appreciated point here is credibility. As Paul Krugman noted in his 1998 paper, the government has to credibly commit to a permanent expansion of its liabilities, but that is very hard  to do because the resulting reflation will be seen as irresponsible by the public. Another way of saying this is that the BoE 2% inflation target is not just a central bank commitment, but a government commitment to low inflation. Until you can credibly commit to changing that via a higher inflation target or a level target a helicopter drop will not matter.

This is not just a theoretical story. People's QE has been tried before and failed miserably. Between 2001 and 2006 Japan conducted the original QE program while running large deficits, as can be seen in the figures below. 

Japan, in other words, was engaged in helicopter drops just like the UK would be under the People's QE. As the figures above show, the monetary base increase was eventually reversed. And what did this do to aggregate demand growth in Japan? Not much:

This notion that doing helicopter drops without any monetary regime change will make a meaningful difference is an economic zombie that needs to be laid to rest. If you want robust aggregate demand growth you have to reanchor your economy to either a higher inflation target or some kind of level target such as a NGDP level target. Otherwise, an economy will be spinning its wheels.


  1. Aren't all OMO's non permanent in reality? Issuing MB through OMO's implies non permanence of money expansions because the bonds the central bank acquires only have a finite life.

    1. Yes, in a mechanical sense. The assets the Fed holds will eventually mature. But the Fed has effectively kept its pre-2008 purchases growing on a permanent growth path by reinvesting. This can be seen in this figure.

    2. I think you provided the wrong figure.

      If OMO's are impermanent then they are surely less credible than one way transfers from the central bank to the public without the purchase of bonds.

    3. CMA, that is the figure I wanted to show. It reveals that the asset side of the Fed's sheet has grown despite the fact that individual assets do mature. It illustrates that Fed can keep its liability expansion permanent by reinvesting asset side of balance sheet as it grows.

      Also, keep in mind that even a pure transfer is not permanent if there is a credible commitment to price stability.

    4. If expansions under a credible commitment to price stability arent permanent then neither are expansions under a credible commitment to expand ngdp. NGDP can expand without monetary growth due to a decline for money demand. Does this make sense?

  2. Helicopter drops imply permanence, end of story. The entire point of the helicopter drop analogy is that of unconditionality, people are getting more money for nothing, without having to give something back in return. If the monetary injection is in the form of a loan that must be repaid, or an asset swat, with commitments to swap the assets back again later, it ceases to bare any resemblance to Freidman's analogy, period. This is why the Japanese experiment didn't work, because there was a commitment that this would be temporary right from the start, they didn't even try any notion of permanence, so I don't regard it as a helicopter drop to begin with.

    "Investment spending would initially begin to grow, but quickly run up against the inflation target and force the BoE to either pull back on or start sterilizing its purchases"

    Why must it be so quick? How do you know there isn't a considerable amount of slack in the economy, where you can pump a huge amount of demand and add a lot of employment before exceeding the inflation target? Hitting the current inflation target would be unambiguously a good thing in any case, because disinflation is certainly not helping.

    1. "Investment spending would initially begin to grow, but quickly run up against the inflation target and force the BoE to either pull back on or start sterilizing its purchases"

      This is like saying :Banker QE would intially begin to grow, but quickly run up against the inflation target and force the BOE to ..."

      That's the whole point. In the U.S. they tapered QE before hitting the inflation target.

    2. My comment exactly, and better said, thank you.

      From this outpost watching markets, it appears people low in the distribution could absorb a shocking amount of cash, spend it all to their much needed welfare, and multiply it as it is spent again and again, all without any inflation effect because the labor market is simply flat on its back.

    3. Not true. There is no such thing as a "permanent fiscal/monetary transfer". Any money which the government (or the central bank) gives away can always be sucked back in via various methods:
      a. The central bank raises interest rates, as in DB's post.
      b. The government issues a great deal of bonds.
      c. The government raises taxes considerably.

  3. For a given , normalized interest rate regime extending into the future , higher growth rates will be achieved with lower debt burdens. Private debt burdens at 50% of gdp will be more conducive to growth than burdens of 100% or 150%. So , peoples QE doesn't need to have any robust immediate effect in order to be beneficial. As long as it is sized , targeted , and timed so that debt burdens are reduced , it could do a world of good. You could even mandate , as Keen suggests , that receipts ( or some fraction thereof ) must first be applied to debt reduction.

    The Penske analogy is bullocks. No conventional or unconventional demand management policies to date can claim to have done anything more than move future demand forward ( or bring it from other countries ) , and commitment to permanence hasn't been required for that to occur , as evidenced by the 2008 rebate checks in the U.S. and by the massive household stimulus in Australia.

    Debt overhang reduction using people's QE might be the first example of a lasting demand stimulus ever achieved - let's get on with it and find out.


    1. Sure, if it is credible. But David's point is that People's QE does not seem credible.

    2. He doesn't explain why People's QE isn't credible whereas QE was credible.

      I take it his point was that QE wasn't credible either which is why it didn't work very well.

  4. The counterfactual chart is more or less the same as the factual chart for the UK.

    The analysis requires effective control of inflation. If the central bank can't even manage positive inflation it won't be set on undoing the effects of PQE.

    I'm uncomfortable with an analysis that is more or less oblivious to actual spending. If PQE prevents individual catastrophic loss of consumption or prevents skill loss or other hysteresis that can make a difference. This is basically an argument for a rewording. While PQE may share some problems with helicopter drops it is not exactly the same thing and the process by which the future liability or loss of credibility offsets spending is not the same - specifically individuals would not have the option of saving their windfall.

  5. You have left-leaning people who endorse NGDP path targeting but posts like this will make them question proponents like Beckworth.

    People's QE is like Bankers' QE except the money is going to the economy via the government instead of via investors. Japan is a bad example.

    Corbyn's point was that if the UK hits the next recession at the zero lower bound they'll need to do QE again.

    But QE is unpopular and didn't seem to work very well. So they suggest People's QE as an alternative.

    "Japan, in other words, was engaged in helicopter drops just like the UK would be under the People's QE."

    As you say they didn't permanently increase the monetary base. If they did with People's QE it would work just as if they did with Bankers' QE.

    A helicopter drop is just giving people money and not diverting it through the government. It's a money-financed tax-cut and if it were done by permanently increasing the monetary base it would work as well.

  6. How would Beckworth do a QE? He seems to suggest it won't work well when combined with a "credible commitment to price stability." Is an NGDP path target the only way? Raising the inflation target to 4 percent?

    But after a recession hits there's the question of how long it takes to get back to "price stability" and the inflation target.

    If they hadn't done QE it would have taken longer. If Bernanke hadn't tapered perhaps the output gap would have closed more quickly. What if they had targeted long-term rates outright instead of saying they'll buy such and such of an amount each month?

    If I were to make a recession brief, I would target the NGDP path, emphasizing the permanence of the monetary stimulus and do all three: a bankers' QE, a peoples' QE and a helicopter drop - giving money directly to the people.

    1. Peter, I'm not opposed to it if tied to an appropriate rule. Here is my preferred way of doing it that is consistent with criticism of post.

  7. If QE financed FICA tax cuts, that would not be inflationary. Business costs would actually go down.

    The Fed would buy US Treasuries and place them into the Social Security and Medicare trust funds.