The discussion of helicopter money may have moved from the periphery of the policy debate to its centre. Unfortunately, analytical confusion seems correlated with interest levels.
Keynes once remarked that when economists are confronted by novel ideas their initial objections frequently cede into the assertion that there is nothing new.
Amidst the considerable noise and confusion around helicopter money there are some genuine policy innovations and some old policies in new clothes.
The two innovations are:
1) Transfers to the private sector from the central bank financed by changes in base money. These can take multiple forms – direct transfers to households, and transfers by stealth through the banking system, which are already underway in Japan and Europe.
2) Money-financed budget deficits, which attempt to alter beliefs about variables in the future – mainly beliefs about future inflation (points 2 & 3 highlighted here by Bernanke).
The old policy in not-so-well disguised new clothes is simply money-financed government spending. The monetisation of deficits neither deserves nor needs a new name. That this old policy could generate nominal demand in any circumstance is probably beyond doubt – its merits are likely highly conditional on institutional and political structure, a critical feature ignored by the literature, although recently raised by Adair Turner.
I want to focus on why (1) is novel, and is urgently needed, and why (2) is an unhelpful distraction – an artifice of model-building which threatens to derail a project of social value, not least because it is illegal in the Eurozone – the part of the global economy most urgently in need of cyclical stimulus.
The tendency of economists to conflate something new and unfamiliar with something distinct but familiar is doubtless a particular case of a more universal human impulse. It may explain why we use metaphors.
It can also be decidedly unhelpful. For example, if we decide that the central bank is merely an office of the treasury, base money is a government debt, no different to a bond or a bill, and households are synonymous with ‘taxpayers’ then we will quickly conclude that a policy of money-financed transfers from central banks is ‘simply’ fiscal policy in disguise.
It is not unreasonable to respond to such objections with ‘who cares, as long as it works’. As an exasperated Simon Wren-Lewis puts it, helicopter money is what it is. I am inclined to deliver a more robust rebuttal, because I fear analytical complacency and model-based confusion are obstacles to much-needed reform.
The analytical confusion is straightforward: the central bank is not the national treasury; base money is different to government bonds, and a check (or perpetual loan) from an independent central bank is not a tax cut.
None of these distinctions are trivial in practice or in law. Economists seem forever determined to deny the existence of money, or to conflate it with ‘debt’. At least in this regard, the Eurozone is a blessing: only the ECB creates base money. The Greek treasury can issue bills and bonds, but cannot create base money – there is no clearer case for a distinction. Some will argue that this is peculiar to the Eurozone. Why then do we have ‘independent’ central banks? The Eurozone is an extreme; independence is a spectrum. There are good reasons why we want central banks to issue money, and governments to issue bonds. They’re different.
A rare voice of clarity in this discussion, Martin Sandbu at the FT, puts it succinctly: “It is more helpful to call policies involving the government budget fiscal policy and policies involving central bank money monetary policy. That avoids collapsing important distinctions […]”
Central banks are not treasuries in other ways, too. Central banks make decisions that effect the economy instantaneously, after a conference call. Fiscal policy doesn’t work like that.
A perpetual loan or transfer from a central bank to a household is not a tax cut, or that wonderful artefact of US economic marketing ‘a tax rebate’. A perpetual loan from the ECB can’t be reversed, it can’t be ‘temporary’. It is qualitatively different to tax policy. Which of course is not to say that if in the future there is too much base money the central bank cannot take multiple actions to reduce its quantity or alter its effects. This reinforces the difference: a ‘fiscal’ tightening takes a very different form to a monetary tightening.
So my first conclusion is straightforward: money-financed transfers to the private sector from central banks are a policy innovation. This is not ‘just’ fiscal policy. It is very different – and likely far more effective – than any fiscal policies currently on the table.
Now what about the other type of ‘helicopter money’, which involves shocking our beliefs about the future – where words and commitments almost miraculously raise demand?
The tedious theoretical games being played by some economists, which masquerade as policy insights, are confusing at best. The prevalent reference to ‘permanence’ seems determined to haunt us, so it is best addressed head on. The worst of economic theory resembles tautology premised on empirical falsehoods. Consider the relevant example: if I assume changes in spending occur due to the inter-temporal substitution of consumption by households optimising over multi-period, or infinite, time-horizons, it will follow by logical necessity that the effectiveness of policy hinges on expectations about conditions in future periods. The central relevance of ‘permanent’ changes to base money or inflation expectations is not a conclusion derived from robust empirically-grounded macro theory – it’s an assumption.
Worse still it is an empirically false assumption. When the answer to most questions pertaining to our long-term economic futures is ‘don’t know’, households have evolved to be rationally myopic, forming habits, and using sensible rules-of-thumb. Modelling their responses to receiving a check in the post from a central bank could be a daunting task – except we know what they do. Those on lower incomes tend to spend a large share, some save the windfall, others repay debt. Forecasting the increase in demand is imprecise, but far less so than with negative interest rates or QE (we don’t even know the sign of the former). Oh, and households in receipt of a check from the central bank don’t ask, ‘before I spend/save/repay debt with this windfall, can you remind me if the associated change in monetary base is permanent or temporary?’
Now our models care if the change in the monetary base is temporary or permanent, or if inflation expectations change – of course they do, that’s how we designed them. But households ‘know’ that the future path of base money is highly contingent, and if the central bank in some distant boom decides to raise reserve requirements, sell assets, raise loan-to-values, or introduce and tax tiered reserves, none of this is particularly relevant to their current decision-making. For similar reasons, cancelling debt, ‘raising’ inflation targets, announcing that QE is ‘permanent’ is likely to be greeted with a shrug – after all, no one has any more money to spend.
Cyclical weakness in China, and persistent political risk in the Eurozone, is a reminder that we need a contingency plan should the global economy slow further. Fiscal policy has been neutered in the Eurozone and is of questionable efficacy in the rest of the developed world for a host of political and institutional reasons, negative interest rates are likely counter-productive and create financial instability, so too does monetary policy reliant on asset price targeting and increasing leverage. There is scope for policy innovation, contingency planning, and better policies for the long run. As it stands, granting central banks the power to make money-financed transfers to the private sector – as close as possible an approximation to Friedman’s helicopter drop – seems the best on offer.
Thanks to you for this clear exposition and to Martin Sandhu for alerting FT readers to it.
Thank you for the insightful article, Eric.
I’d be curious to hear your take on the recent article “Helicopter money: The illusion of a free lunch” by Borio et al. (24 May 2016, http://voxeu.org/article/helicopter-money-illusion-free-lunch), as it provides what I’d consider a more cogent critique of HM than the sadly still all-too-frequent inflation scaremongering.
*Personally, I see three major objections, but your expertise goes far beyond the bits and pieces I’ve read – hence my interest in hearing your take.
1. The authors’ argument is premised on the idea that the boost to demand comes from a lower interest debt service burden. However, this seems erroneous, since most proponents underscore an important boost from consumption, as at least non-Ricardian consumers (the majority) believe to have higher net worth. There is an element of irony here insofar as they criticize standard models for omitting real life complexities, yet they themselves omit this important but ‘messy’ channel.
2. Suppose, however, we grant them the above. More importantly, the article fails to acknowledge that the very reason for employing HM is that governments are unwilling to employ debt-financed fiscal stimulus – whether out of (irrational) fear of high debt or political motives. Put differently, a proponent of HM can acknowledge that the effects of HM won’t exceed those of debt- or tax-financed fiscal expansion and be perfectly happy with that outcome.
3. Three further, related considerations. First, many HM proponents already acknowledge that in the long run¬ there is no difference between money and non-money financed fiscal expansion. For instance, it is widely acknowledged that the government might have to recapitalize the CB, at which HM does indeed turn into conventional non-money finance fiscal stimulus. The reason why this might be necessary is simple: When it becomes clear after the recession that there is now too much money in the economy, the central bank mops it up by selling bonds. In other words, monetary policy acts as normal. If the central bank runs out of assets to do this, it gets recapitalised. Second, as footnote (5) acknowledges, there are schemes to recoup the cost of paying interests on reserves through a levy on banks. The authors claim that this amounts to tax-financed deficit spending, but again, that’s a feature not a bug. Moreover, it is plausible to suppose that MPS(banks) >MPS(avg.taxpayer), such that the redistributive facet of this policy would amplify the demand-boosting effect of HM. A similar, third point is buried in footnote (4), viz., that borrowing at the overnight rate (HM) as opposed to a longer-term rate (standard) does provide interest rate savings.
Borio incinerates a straw man. Helicopter money as advocated by myself, Simon Wren-Lewis and Mark Blyth does not involve any commitments about future monetary policy – so Borio’s argument is not even relevant.
Thanks, Eric. However, I’m not sure whether this is sufficient to address the argument by Borio et al. You’re entirely right, of course, that neither you SWL or MB have advocated explicitly committing to a future policy course (in this particular context of HM). However, the argument by Borio is that such a commitment is implicit in HM. It arises because with the injection of base money, the question arises how to remunerate the resulting excess reserves*. If the interest paid on these excess reserves is zero, then this implicitly commits you to a future policy rate of zero. If not, the remuneration cost is ultimately charged to the fiscal authority, leading to HM being comparable in effect to debt-financed fiscal stimulus. Or so the argument goes at least (I’ve mentioned a few counterarguments above…).
*See also Toby Nangle’s recent piece: http://www.columbiathreadneedle.co.uk/media/10023117/en_viewpoint_the_impact_of_helicopter_money.pdf
Could you get Scott Fulwiller or someone else to comment on this?
Does the price inflation target apply to the lender of last resort function of the central bank to the commercial banks?
Thanks!
Scott and I usually agree on substantive points – but he would probably put it differently! Tweet him …
I tweeted him once. I did not get an answer. Could you tweet him?
What is your answer to the question?
Thanks!
Hi Eric
My take:
http://monetaryrealism.com/helicopter-recall-fiscal-repairs-needed/
quite different (no surprise)
Why are you so attached to labelling ‘fiscal’?
Because I think it is meaningful to the explanation of what HM comprises. And that is relevant to the marketability of the proposal.
Conventional OMO requires fiscal as antecedent.
QE requires fiscal as antecedent.
HM = the limit of QE as the time lag between fiscal and monetary goes to zero.
That limit is not a black hole that vaporizes the essential fiscal prerequisite.
Moreover, the fiscal policy must be spelled out and effectively delegated to the CB for it to pull the trigger on timing. The CB does not have the autonomous responsibility or authority to design the required authority – because it is fiscal. In that sense, HM is more fiscally bound and dependent than even OMO or QE.
It is not appropriate to abandon sensible policy differentiation in the case of HM. Obscuring the fiscal content and fiscal risk (either inadvertently or as a message strategy) will not be effective in persuading policy makers who need to understand what it is they’re being asked to approve.
Monetary policy has fiscal effects – that is not new, nor does this make it fiscal policy. There is a reason why we have given CBs control of the money supply.
I cannot help but notice that you do not define monetary and fiscal policy – you are not alone, but this isn’t really good enough.
As regards tactics, it is true that labels matter. Fortunately, in the Eurozone the legal definition is clearer than most most economists: fiscal policy is directed by national treasuries, and involves borrowing; monetary policy involves money and is directed by the CB.
“There is a reason why we have given CBs control of the money supply.”
Are you sure CB’s control the money supply?
Good spot, Fed Up! I am referring to base money in this context. In extremis, though, the CB can determine deposit growth too. If the CB keeps shrinking the monetary base, eventually the stock of loans will shrink (either through defaults or repayments), and if the CB credits households with money it can grow the stock of deposits.
I’ll come back on this
Meanwhile, I found this very interesting:
https://monetaryfreedom-billwoolsey.blogspot.com/2016/05/the-paper-gold-fallacy.html?m=1
Just read – nearly right about a lot of things. Crucially wrong about most.
But can the CB control the monetary base?
Let’s assume no entity wants to hold demand deposits, only currency. However, assume all entities need to borrow from the commercial bank(s). That is as soon as someone gets a loan/demand deposit, it is converted to currency. Also, assume all of the commercial bank(s) is/are solvent.
What happens?
The preminent fiscal policy identifier is spending or transferring money – without a financial asset received in exchange. I made that pretty clear in my post. It is essentially the same qualification as the MMT net financial asset description.
That is why HM as typically proposed is fiscal policy from the get go. It includes spending or transfers.
Conversely, monetary policy includes financial asset exchange – e.g. bonds for bank reserves.
That is why HM as generally proposed is not monetary policy. It results in negative central bank equity instead.
Fiscal policy also includes financing with taxes or bonds.
The central bank can convert fiscal bond financing to money financing by buying government bonds and creating bank reserves – as in the cases of OMO and QE.
HM could be transformed to a proper fiscal/monetary nexus if Treasury issued bonds to finance spending/transfers and the CB simultaneously bought the bonds (or at least the same quantity of bonds).
The absence of bonds in the typical HM proposal doesn’t negate it’s essential fiscal character. Rather it amount to an error in balance sheet construction. The central bank should not create a negative equity position by undertaking systemic fiscal policy implementation for its own account and balance sheet consequence.
(A zero interest perpetual loan has a value of zero, which results in the same negative capital situation.)
A transformed HM policy as I suggest combines distinct fiscal and monetary policy, logically sequential but operationally simultaneoous, with Treasury responsible for fiscal and CB responsible for monetary.
The overarching policy hierarchy should be clear – the CB is effectively subordinate to Treasury. Monetary base expansion is a function of prior or simultaneous Treasury bond financing – including OMO, QE, and HM (with my proposed revision). Treasury also has a claim on CB profit as the fiscal effect of monetary policy.
HM is a nuanced case. No point in trying to squeeze an understanding of its fiscal characteristics into general definitions that haven’t anticipated such a strategy.
Monetary policy “involves money” is too general. So does bond financing involve money.
My post has a number of examples that should inform definitions.
HM is a monetary operation with a required fiscal policy authorization and a required coincident fiscal policy action. The monetary component cannot be a standalone action because of these characteristics. The very high degree of fiscal policy dependence makes the policy fundamentally fiscal in a way that OMO and QE are not. By contrast OMO and QE are independent monetary policy to the degree they require no more than prior fiscal action, including bond issuance, and a delegation of independent policy authority.
The monetary component of HM requires a fiscal policy directive for the intended distribution of the spending/transfer of money to the private sector. That makes the policy predominantly fiscal under the usual proposal.
My proposal clarifies the fiscal/monetary nexus, consistent with the logical order already found with OMO and QE. The monetary timing is compressed to be coincident with fiscal, but fiscal is recognized as a Treasury transaction, as under OMO and QE, with the CB pulling the trigger on timing under a preauthorized fiscally delegated authority for the timing of implementation.
In the case of HM, there should be no question that the monetary policy that allows the CB to pull the trigger on timing requires a fiscal policy delegation for the identification of designated recipients.
The key difference from either OMO or QE is the association of fiscal spending with immediate rather than lagged monetary financing.
Ironically, it is that immediacy that makes the policy combination predominantly fiscal.
Under the typical HM authority, there is an unproductive partial transfer of the core Treasury negative equity position to the CB balance sheet. Not only is this not necesary. It is not advisable. There is no consolidated fiscal effect from CB held bonds – but such bonds are advisable for purposes of insuring against all contingencies for potentially beneficial reserve drains in the future.
There is absolutely no downside to such a reserve of bonds, whereas the CB negative equity contrivance is an expedient intellectual compromise around a subject that should be immunized from intellectual compromise.
‘Spending or transferring money without receiving a financial asset in exchange’ – are you sure that’s the definition of ‘fiscal policy’? (Last chance to amend!)
Consider: a) the scope of fiscal policy under that definition, b) the potential scope of monetary policy under that definition. In both cases, way too wide.
I said preeminent identifier.
That’s not the same thing as a definition.
It’s the most important qualification, as in:
“It is essentially the same qualification as the MMT net financial asset description.”
Then:
“Fiscal policy also includes financing with taxes or bonds.”
And:
“HM is a nuanced case. No point in trying to squeeze an understanding of its fiscal characteristics into general definitions that haven’t anticipated such a strategy.”
And:
“My post has a number of examples that should INFORM definitions.”
I think this is a reasonable approach.
Your insistence on top down rigidity is unnecessary in my view.
I think I will prevail in a contest of complete definitional development – but it isn’t necessary.
And I don’t think it is quite fair to suggest in a non-specific way that I need to think things through more – you don’t actually know what I’ve already thought through.
There can be exceptions to the order of general qualifications. For example TARP in the US was definitely not fiscal, notwithstanding Treasury execution of that policy. There’s an explanation as to how that happened. But there is no credible argument for why HM needs to funded by a CB cheque or why it should exclude bonds.
There is an element of common sense in all this, analogous to the judge’s observation:
“I know pornography when I see it.”
HM by negative CB equity just ain’t gonna happen.
I meant funded by a CB cheque in the sense of a direct negative equity result instead of bond buying.
P.S.
I’ve been punching all of this into a phone while on vacation – a dreadful physical disadvantage.
But temporary.
On guard, sir.
🙂
One can assign whatever meaning one wants to words in order to engage in meaningful linguistic communication. But we are trying to achieve something more substantive. The fiscal/monetary distinction is not obvious or this discussion would not be occurring.
There is a theoretical, abstract point – can we in the abstract identify qualitatively distinct sets of policies? There is an institutional question: do we ascribe distinct policies to separate institutions? And there is a legal question: what policies are ascribed to which institutions under law? The matter is further complicated by the reality that laws, and the institutional division of labour, vary considerably by geography and by point in history. For example, in the past the Federal Reserve had far greater policy freedom than it does today.
At the abstract level, policies aimed at changing the stock of base money are clearly distinct from any other policies (regardless of the institution in control of base money). Tax and spend policies are also distinct (irrespective of the means of finance).
Whether changes in the stock of base money are adopted via an exchange of assets or not seems tangential. Whether it does or not, changes in the availability of base money alway have implications for the CBs balance sheet (as is clear with QE). This too is tangential.
One of the points that Simon Wren-Lewis, Mark Blyth and I have repeatedly made is that the characteristics of cash transfers from the CB to households that people usually deem to be fiscal, are in fact characteristics of all monetary policy. Whatever means the CB chooses to make changes to the stock of base money has distributive, fiscal, and balance sheet implications. HM is not unique in this regard.
In summary, it remains clear that monetary policy – as its name suggests – pertains to changes in the supply of base money. The means by which these changes are implemented is historically and geographically contingent.
As regards the law, it is clear that helicopter money in the Eurozone can be implemented as monetary policy. I believe it is not legal in the US, and in the UK it would be implemented in concert with the Treasury.
As regards the institutional division of labour – this is clearly whatever we decide it should be. Hence the diversity of institutional arrangements.
Whether changes in the stock of base money are adopted via an exchange of assets or not seems tangential.
Pardon me, if I barge in.
I don’t mean to speak for JKH, but an initial monetisation of current output, whether through an executive order by the sovereign (fiscal policy) or by commercial bank lending to firms, is in a different category than any subsequent act of swapping financial assets.
Monetising output is logically prior and has a 1:1 effect on GDP, per definition. Asset swaps OTOH, whether two existing assets are swapped or one is new, have 0 effect on GDP in the first order. I remember you saying as much in your post on QE, IIRC.
To my mind, that is as binary a distinction as one is likely to come across in economics and feels like a natural line along which to distinguish between fiscal and monetary policy. It also fits well with institutional setups in almost all jurisdictions I’m acquainted with.
And the fact that one would like to merge the technocratic efficiency of independent monetary policy with the economic potency of fiscal policy does not change the inherent distinction. It’s an attempt to circumvent the often sad state of democratic non-decision making by redefining executive acts as acts of administration. Understandable but intelectually dishonest and not helpful in the long run as what is really needed is that we fix politics, not circumvent it.
Thanks Oliver. I think it’s tangential to the definition of monetary policy, whether or not an asset is exchanged for base money or not. That is entirely different to saying that assets swapped under QE may have no effect. We have to stand back and ask what monetary policy is trying to achieve. It is currently a means to constrain or increase aggregate demand via changes in the availability of base money. The supply of base money has been used by central banks to affect demand through three channels (all of which have fiscal effects): the price of credit (interest rates), duration (ie forward guidance/LTROs), and the degree of credit risk (collateral & directed lending). Helicopter drops are an extreme on each of these axes. None of this confuses monetary policy and fiscal policy, which are distinct in the abstract, institutionally, and legally.
I think I missed the point of the post when I focussed blindly on the one sentence I quoted. None of what I said actually covers HM.
Maybe I can simplify things like this:
1) When money is spent by government to buy non financial assets that is considered fiscal policy, whether it is financed by bonds or taxes. The monetary authority is involved in the sense that it acts as banker on behalf of government.
2) When new money is created to buy existing financial assets that is considered an administrative act of the monetary authorities, aka monetary policy.
3) There is a third category where money of the central bank is either a) not taken away from non-government agents (tax cut) or b) explicitly given to non-government agents (helicopter money). This does not feed directly into the income flow as 1) but rather changes the financial net worth of the agents involved vs. what they would otherwise have expected. The hope being of course that it will feed into the income stream as a second order effect.
I contend that 1) is as I described in my comment above while 2) is exactly as you describe in your reply and both 3)a) and 3)b) are distinct and must clearly be considered fiscal rather than monetary policy. Graphically, HM is the mirror image of a tax cut.
Nothing new to add, but a couple of points I want to emphasize:
I think of a continuum:
OMO –> QE –> HM
All 3 of these are distinct combinations of fiscal policy and monetary policy.
The first is prior fiscal policy that produces the bonds that monetary policy wants to buy in order to increase the monetary base over time.
The second is the expansion of the first in an “unconventional” way.
The third is the bone of contention.
I think of the third as the limiting case of the second – as the time lag between fiscal and monetary policy implementation goes to zero.
In other words it is the compression of a temporal sequence of fiscal and monetary policy to a sequence that has the same logical order but effectively becomes simultaneous in the timing of the implementation of those two differentiated policy constituents.
That interpretation of logical consistency forms the basis for how I put it all together from there – with consistently interconnected operational and accounting features.
I think one area where I have not done the best job of explaining is this:
When I say that HM “is fiscal” I mean that it is not exclusively monetary.
I.e. it is a combination of fiscal and monetary, as holds along the continuum I described.
And that leads me to suggest the operational and accounting arrangements that I do.
And that informs definitions of fiscal and monetary that fit with OMO, QE, and HM as I have proposed.
One thing I know is that with time I can definitely improve the explanation of my own thoughts on this subject.
Thanks for the discussion.
Much appreciated.
[…] and holding onto what they’ve got, the Chancellor’s only real option this autumn may be helicopter money. The image is of a government helicopter circling overhead and dropping money onto the population […]
There is no conceptual difference between #1 and money-financed government transfer. That’s still fiscal policy.
Practically, it is a matter of who should decide the who gets the transfe – the executive and legislative branches of elected government, or central bankers.
Thanks for the comment Jason. I think to further discuss this issue, you have to define ‘fiscal policy’ and ‘monetary policy’. Only if we are clear what we both mean by these terms can we decide which is which.