Winner of the New Statesman SPERI Prize in Political Economy 2016

Monday, 27 May 2013

Debating Helicopter Money (while the lunatics continue to run the asylum)

Vox has published an excellent summary account of a debate between Adair Turner and Michael Woodford (skilfully moderated by Lucrezia Reichlin) on helicopter money. My line on helicopter money has been that it is formally equivalent to fiscal expansion coupled with an increase in the central bank’s inflation target (or whatever nominal target it uses), and so adds nothing new to current policy discussions.  I think the Woodford/Turner debate confirms that basic point, but as this may not be obvious from the discussion (it was a debate), let me try to make the argument here.

Two Equivalences and two Red Herrings

Turner calls his proposal ‘Outright Money Financing’ (OMF), so let’s use that term to avoid confusion with other versions of helicopter money. Under OMF, the central bank would decide to permanently print a certain amount of extra money, which the government would spend in a way of its choosing. The alternative that Woodford proposes is that the government spends more money by issuing debt, but the central bank buys that debt by printing more money (Quantitative Easing), gives any interest it receives straight back to the government, and promises to ‘never’ sell the debt. (If it reaches maturity, it uses the proceeds to buy more.) Let’s call this Indirect Money Financing (IMF). If everyone realises what is going on in each case, and policymakers stick to their plans, the two policies have the same impact.  That is the first equivalence, which both Woodford and Turner seem to be happy with.

In both cases, more base money has been permanently created. This will have implications for the inflation or NGDP level that the central bank attempts to achieve. If you believe that in the long run there is a stable relationship between the amount of base money in the economy and the price level, then permanently printing more base money must raise inflation for a time at some point. In other words, the central bank cannot independently control both base money and inflation. This leads to a second equivalence: we can either talk about long run levels of base money or average future levels of inflation: one is implied by the other. [1]

The point of either OMF or IMF is to combine short run fiscal stimulus with raising the long run level of prices. Sometimes advocates of helicopter money suggest that there is no reason for long run prices to be higher. However, as long as there is a link between how fiscal deficits are financed and the long run price level, to keep average inflation constant requires that money financing is temporary, so in that sense it is even closer to current Quantitative Easing. Much of the discussion below is relevant to that case too.

Now the red herrings. First, although Woodford argues in terms of nominal GDP targets rather than inflation targets, the distinction is irrelevant to this debate, if both targets are perfectly credible (see more below). Under both schemes the central bank has some form of nominal target, and its money creation has to be consistent with this. Second, this particular debate has nothing to do with the form of fiscal expansion: under Turner’s proposal the central bank decides the aggregate amount of OMF, but the government directs where the helicopter distributes its money, which could be over schools and hospitals, the population as a whole, or just tax payers.

How can the two proposals differ?

So if the two policies can be formally equivalent, where are the differences? The first point to make is that if the government and central bank were a single entity, there would be no difference at all. Under IMF the government would be selling debt to itself. So any difference has to involve the fact that the central bank is an independent actor.

The promise to raise future inflation to stimulate the economy today suffers from a well known time inconsistency problem. The promise works if it is believed, but when the future comes there is an incentive to go back on the promise. So how likely is it that the central bank will take that incentive? Using the second equivalence noted above, let’s talk about the central bank going back on its promise to make money creation permanent. One argument for OMF is that it may be easier for the central bank to do this if it can just sell some of its government debt. On the other hand, if the only way of taking money out of the system is by persuading the government to raise taxes (or cut spending), that seems less likely. Thus OMT is a commitment device for the central bank not to renege on future inflation targets which sophisticated agents may recognise. This argument seems a little tenuous to me, as it assumes that OMF takes place to such an extent that the central bank in effect loses the ability to raise short term interest rates sufficiently to control inflation. I cannot see any central bank willingly undertaking this amount of OMF.

On the fiscal side, agents may base their assessment of future tax liabilities on the published government debt numbers, and fail to account for the additional future revenue the government will receive from the central bank as it passes the interest on its debt back to the government. Here agents are sophisticated enough to base their spending plans on an assessment of future tax liabilities, but naive about how these liabilities are calculated. Possible I suppose, but then the government just needs to start publishing figures for debt held by the private sector excluding the central bank.

A naive government may feel constrained by its fiscal targets and so feel it is unable to undertake bond financed fiscal stimulus, but may be prepared to contemplate money financed fiscal stimulus. That seems quite plausible, until you note that under the second equivalence above, any switch from bond to money financing that was not reversed should be coupled with a temporary increase in the central bank’s inflation target.  A temporary period of OMF that was later undone would not raise average future inflation, but it would equally do nothing to change long run levels of debt either. In that case surely everyone would just start counting OMF as (soon to be) debt.

Turner argues that OMF would discipline governments more than IMF, because central banks would determine the amount of OMF. This argument also seems strained. For example in the UK, where the government sets the inflation target and we have Quantitative Easing, the problem is that the government is borrowing too little, not too much. Woodford worries that OMF blurs the lines between who take fiscal and monetary policy decisions, which is why he prefers IMF. However if the government decides how OMF is spent, and retains control over aggregate borrowing, it is difficult to see the force of this argument.

So I think OMF and IMF are pretty well equivalent. As I’m in favour of IMF (fiscal expansion and more future inflation), then this implies I am in favour of OMF.  If my earlier posts have appeared critical, I think that is because some proponents of helicopter money seemed to deny the equivalence to IMF. However, if pretending that helicopter money is monetary rather than fiscal policy could convince some policymakers to change course, maybe the end justifies the means. Unfortunately Eurozone consolidation continues unchecked, the UK government brushes aside advice from the IMF to relax austerity, and the US recovery is dampened as sequester bites. The intellectual case against austerity is overwhelming (beside the Krugman piece I mentioned in my last post, see also Martin Wolf here), more and more academics are suggesting we need higher inflation (to take just five: Mankiw, Rogoff, Krugman, Ball, Crafts), yet only Japan has been moved to change course. The lunatics and asylum jibe is of course unfair, but just how long will it take policymakers to start addressing the problems of today, rather than the half imagined problems of the past.

[1] Some comments on earlier posts have disputed this point. There is not space to discuss this here, as we need to consider not only the detailed mechanics of monetary policy but also the nature of money itself.  Obviously if the long run price level is really independent of how deficits are financed, then much of the discussion here becomes unnecessary.


  1. As you say, exact equivalence depends on the details of independence, mandate, credibility and commitment. Within that space, I see another possible variation on the helicopter drop in which it is the central bank that is given the power to perform the drop (eg a citizens' dividend) as well as the mandate to deliver an aggressive nominal target. In that circumstance the CB would have the ability to counteract any amount of fiscal austerity from the side of the treasury, no matter how severe.

    The problem we have is that the CB can't get traction to achieve its nominal target, in which case they obviously haven't been given the proper tools to do the job. The way I see it we can either revoke cash so they can use negative rates (preferable), but failing that, hand them the power of the helicopter drop. If we are going to give them a mandate they need the tools to achieve them, no matter how stupid or pigheaded the (principal) fiscal authority.

    " In other words, the central bank cannot independently control both base money and inflation."

    I'm sure you are becoming bored of me pointing out that for CBs that pay IOR (ie most of them including the BoE), that statement is exactly false (maybe footnote 1 was directed at me). Anyways, I can't see why I'd be wrong, so I'll keep pointing it out.

  2. I can only make sense of the "money financing" concept if the public debt happens to exactly equal the money base. In reality, public debt is larger than the base, so $1 of so-called permanent money creation "finances" more than $1 of spending (i.e. debt/GDP falls). Limiting deficit spending to the amount of money creation is excessively conservative.

    (The above should not be read as applying to the current situation, where the base size is decoupled from money demand. It's pre-2008 language).

  3. Don't distributional concerns break the equivalence? Marginal propensity to spend seems like it would be much higher with cash distributed via OMF rather than IMF.

  4. "if the government and central bank were a single entity, there would be no difference at all"

    Exactly right. If you consolidate their balance sheets, the two approaches it comes to the same thing. (Except in terms of who controls where the money's dropped.)

    If you then consolidate the GSEs as well, making them part of "government," you start to see a pretty useful depiction of govt vs. private balance sheet changes.

    I poked at this here:

    “Public” Debt and Safe Assets: A View from Space

    A true helicopter drop, of course, would be a physical currency one by Treasury (or Treasury-ordered bank-deposit credits, or mailed Treasury Visa gift cards, or...), with no additional issuance of Treasury debt, purchased by "the public," the Fed, or anyone else. MMT World...

  5. I would also like to know why K is wrong in regard to CBs paying IOR.

  6. Agree that from a practical pov IMF and OMF are the same.
    Starting from there, however it is clear that if you decide to go that way your IMF will be the best option and by far. It is simply much better to manage and uses relatively more things that have at least partially be tested in the marketplace like QE stuff.
    OMF has likely to go in one huge step at the time (say write off 30 or 40% of GDP debt, now or in the future, but anyway the amount or perc of GDP will be clear). And we donot know how markets react to OMF certainly not if it is a huge single event (which it likely is).
    IMF can go in steps that can be corrected if necessary and markets donot have to be informed (not at this stage at least).
    Markets might suspect something (like they do now), but that is different from a single day newsevent that the printing press is on for huge amounts. IMF will go much more gradual also in market behaviour while OMF as it will be a one time event. Hard to see that printing 30-40% of GDP will not give a marketreaction.

    Another remark: Macros are extremely poor in packaging such a thing, make a product of it. It is as far from a ready product to be sold to politicians or voters and to be used in the real world as can be.

  7. Isn't the assumption of a unified banking system whose liabilities are "money" wearing a little thin? It's more realistic to think of two banking circuits corresponding to the distinction made by the Vickers Report: a smaller retail banking system that lends to households and non-financial firms (the "real economy"), and a larger investment banking circuit that lends to itself in a merry-go-round and to nonbank financial intermediaries. Expanding the monetary base of current universal banks primarily expands the second, financial circuit, with little effect on the real economy. There is some leakage, a positive multiplier into retail banking, but it's much less than 1.
    This is of course an updated, institutionalized version of classic Keynesian LM theory. If it's right, it makes a big difference whether your monetary expansion starts in the retail sector, by helicopter drops to firms and households, or pushing on the investment baking string in the City.

  8. Most Central Banks have some regulations that prevent them from targeting the helicopter drop where it will have the greatest effect on domestic unemployment. Helicopter drop is a blunt instrument. Fiscal stimulus can be fine tuned, but is often corrupted by political interests. An "infrastructure bank" would be a great addition, because it could prioritize projects for fiscal stimulus based on unemployment and Return on investment.

    - jonny bakho

  9. Great post. I am wonderful looking such a blog which is like a dot com domain. You are great sir.

    I really appreciate of your contribution in making such a excellent blog. I am so much impressed. I now must be active and keep posting. Thank you.

    I would like to ask one question that how many post, backlinks and visitors re required to increase the page rank of blogspot blog? I will be waiting for your reply.

  10. I am finding it interesting and also these are the terms which exists reasonably tanner mainstain makes sense for many things.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time.