Skip to content

Helicopter money and some challenges

In a recent FT column Martin Sandbu writes about the “timidity of monetary policy” following the Great Financial Crisis, accusing central banks of not doing enough in order to boost the recovery from Great Recession. One might be surprised to hear this, since we have seen central bankers quickly cut short-term interest rates until hitting the zero lower bound (ZLB), and even after that, experimenting with new(ish) policy tools such as negative interest rates, quantitative easing (QE) and forward guidance in order to depress longer yields, too. Besides, it is generally understood that in a liquidity trap (when interest rates hit the ZLB) monetary policy loses most of its traction on the real economy anyway: the market clearing equilibrium real interest rates can go so negative (e.g. due to a rise in deleveraging pressure) that with the zero bound on nominal interest rates and low inflation, the central bank is unable the engineer a correspondingly low actual real interest rate. The result is excess saving desire and chronically weak aggregate demand, while monetary policy is out of firepower to stimulate the economy – this is a liquidity trap. So it might seem unjust to blame them for not doing enough. However, Sandbu argues that both QE and forward guidance could have been pursued more aggressively on the one hand, and that other potential policy strategies were left untried on the other hand.

While I would dispute that the former would have mattered much, I agree that the on the latter front monetary policymakers indeed left some options unused. One of these include monetary strategies which try to engineer higher inflation expectations, and thereby deliver a lower real interest rate, an argument put forward by Mike Woodford (but also endorsed by Ben Bernanke). Such policy strategies, like nominal GDP or price level path targeting, and unlike the current inflation rate targeting regimes, automatically make up for missed inflation (and growth) targets in the past by temporarily targeting higher growth rates now: this history dependence can deliver the desired rise in inflation expectations, something I have also argued for before, and something which characterizes optimal monetary policy in most standard theoretical macro models. Another of these alternative policy options mentioned by Sandbu is helicopter money. This has recently been discussed quite a lot (see Buiter, Bernanke, Turner, Galí, Woodford, Deutsche Bank), as the ultimate radical weapon which central banks have at their disposal to boost demand: just print fresh money and wire it to people’s bank accounts – surely they would spend at least some of it. In this post I would like to explore in a bit more detail just why and under what conditions helicopter money is more effective than a combined standard monetary and fiscal loosening. We will see that it might be trickier to engineer than it first seems.

Helicopter money – how it is supposed to work

What is exactly meant by helicopter money? It can be its most straightforward version, where the central bank just creates new digital money and transfers them to each household, as a gift. Or, more likely, it can take the form of a fiscal spending, directly paid for by newly printed money (the central bank buying government bonds and destroying them immediately, or just crediting the government’s account immediately). In both cases, on a consolidated government level, new spending is financed by a liability of the state which is never to be repaid (or which can be repaid with itself): money – instead of a liability of the state which needs to be repaid in the future: bonds. Helicopter money is basically a fiscal stimulus financed by a permanent increase in the monetary base, the key word being permanent.

It is precisely its permanent nature that is supposed to distinguish helicopter money from “plain Quantitative Easing” (coordinated with a fiscal stimulus). QE also involves the central bank buying government bonds with freshly printed money, but those bonds are expected to be repaid in the future, which means the government will need to raise taxes or cut back on spending at some point in the future. (Sidenote: the primary aim of asset purchases under QE was not to support fiscal spending, but to lower interest rates also at the longer ends of the yield curve.) Rational households foresee this, and will save most of the windfall from the fiscal stimulus instead of spending it: this is what we call Ricardian equivalence. Of course, Ricardian equivalence is a poor description of reality, as many households are credit constrained, so called “hand-to-mouth” agents who cannot smooth out most temporary income changes and whose consumption therefore reacts much more to changes in their current income (they have large marginal propensity to consume, or MPC) – as opposed to permanent income consumer, consumption-smoothing “Ricardian” households with a low MPC. In addition, fiscal policy has much larger multipliers on output in a liquidity trap, since the usual crowding out effects of extra government borrowing through pushing up interest rates are non-existent at the ZLB (where interest rates are already higher than they should be, and where we have excess saving anyway): investment and the consumption of interest-sensitive Ricardian households will therefore not be crowded out. That is why even a temporary fiscal stimulus is likely to have a large effect: 1) the presence of hand-to-mouth households and the 2) absence of crowding out make fiscal policy very effective at the ZLB. The only channel through which helicopter money promises to be more than a temporary fiscal stimulus, is by getting around the Ricardian equivalence problem and affecting the spending of Ricardian households, too: they react to permanent  income changes, so the fiscal stimulus needs to be permanent, i.e. the increase in the monetary base financing the stimulus must never be reverted, no new taxes must be raised in the future. Otherwise we are not talking about helicopter money, but just about a plain vanilla fiscal stimulus.

The permanently higher monetary base turns into permanently higher money supply as it is spent in the economy (ending up as deposits, without having to rely on more lending from the banking sector), and this might translate into a permanently higher price level, delivering badly needed increase rise in inflation expectations, which might also be of help in getting out of the liquidity trap according to the Woodford-type argument.

Apart from the necessity of its permanent nature, another thing about helicopter money is that it involves the central bank creating new liabilities (money) without acquiring a corresponding amount of assets (like government bonds) so it suffers losses on its equity: its net worth can potentially turn negative. Many commentators note that negative equity should not concern central banks since they are not like ordinary firms. Their liabilities are denominated in government paper which they can produce at will, so they will always be able to pay their debts: in essence, paper or fiat money is promised to be redeemable by itself (and not by gold or FX), so they say the “central bank balance sheet is an accounting fiction”: the only balance sheet which matters is that of the consolidated government, and helicopter money just causes a shift in the mix of government liabilities from bonds to money.

Fiscal backing of central bank balance sheets

Now that we understand the mechanisms of helicopter money, let us see how problems might arise. According to Christopher Sims, when monetary policy is not set jointly with fiscal policy, and the central bank aims to control the price level on the long run, its balance sheet does matter. Unless fiscal authorities recapitalise a central bank with negative equity, monetary policy might lose control over the price level. This is simply because it might run out of assets to sell with which to contract the monetary base in case the need arises at some point in the future (negative equity is equivalent to having too many liabilities relative to assets). In fact, this is a good way of making sure that the initial increase in the monetary base does indeed stay permanent: it can only be reversed to the extent that the central bank has enough assets to sell. Alternatively, the central bank can raise the interest rates which it pays on its liabilities (reserves) to try to fight inflation pressures, but without enough sellable assets these interest payments can only be paid for by printing new money and further increasing the monetary base in a feedback loop, ending up being inflationary rather than deflationary. Of course, this seems like an extreme case: for a central bank to run out of all of its assets, its equity must go into negative territory to a similar extent which implies a huge initial helicopter money operation which might be much more than enough, making this constraint not really binding. However, this still puts some limit on the degree to which the ultimate weapon of helicopter money can be wielded without jeopardising future price stability. And I would argue that this limit is not a very large one: while the lower bound on equity can be the negative of all the assets the central bank currently has, potentially making up a third of GDP, it has to be understood that this firepower of money printing is for eternity, i.e. if it is used up now, nothing remains for the future. Viewed from this perspective, the constraint can seem a much stricter one.

The presence of seigniorage revenue can loosen this constraint a little bit to the extent, that moderately negative net worth can be “worked off” by a series of positive seigniorage profits, but this is a very limited remedy, and might not be enough to replenish a dramatically negative equity.

One might also say, that higher inflation should not be a worry, since it is precisely what we want in the liquidity trap: in fact, we implement helicopter money for the very reason of raising prices. This is indeed true while in the liquidity trap. But the above concerns are not about a one-off increase in the price level, or temporarily higher inflation, which are indeed desirable in a liquidity trap. They are about completely losing control over inflation forever, where monetary policy loses its tools to combat inflationary pressures even when the liquidity trap is far behind us.

Of course, if fiscal authorities stand ready to back up the central bank’s balance sheet and recapitalise its negative equity e.g. by transferring some interest-bearing assets to it, then the above problems go away, and the central bank’s balance sheet indeed does not matter. In many advanced economies this is the institutional setup. But this amounts to the treasury committing to repay the losses of the central bank (incurred during the helicopter money operation, when new money was printed against central bank equity) by having to raise more taxes in the future. And this destroys the very point of helicopter money, which is its permanent nature: if helicopter money today will be offset in the future by higher taxes, i.e. if the increase in the monetary base is not truly permanent, then Ricardian households are unlikely to react much to it. In this case the whole exercise becomes just a temporary fiscal stimulus, supported by QE and loose monetary policy. Not that it would be ineffective (see arguments above), but it would not be what we defined as helicopter money. In fact, this policy mix was pursued in the immediate aftermath of the financial crisis, but was discontinued rather prematurely. And not because of timid monetary policymakers but because of the timidity of fiscal stimulus.

The bottom line is that if macroeconomic policy wants to maintain future control over price stability, either monetary policy needs to show some self-restraint in implementing helicopter money, or fiscal backing of its balance sheet is essential. But either way, a truly permanent helicopter money might be hard to engineer without limits.

Fiscal Theory of the Price Level (FTPL)

The above discussion relates nicely to the fiscal theory of the price level, which states that the value of paper money (the price level) is ultimately a function of fiscal policy, determined by the present value of future primary budget surpluses. Looking at the per-period budget constraint of the consolidated government, fiscal spending can be financed by either raising tax revenue, issuing more debt or by printing more money (seigniorage revenue). All this can be compressed into a single intertemporal budget constraint saying that the present value of future primary surpluses (taxes less spending) needs to equal the real value of today’s outstanding liabilities. If future surpluses are decreased (e.g. by implementing a permanent fiscal stimulus, like helicopter money, not to be reversed later by higher taxes, and not recapitalising the negative equity of the central bank), then the only way outright default can be avoided is that the real value of outstanding nominal debt is eroded through a higher price level. In essence, the stimulus is paid for by an “inflation tax”, seigniorage. Of course if fiscal policy chooses the path of future primary surpluses such that it stabilizes the debt level, then monetary policy is free to control the price level (active monetary + passive fiscal policy): a setup we have in normal times. However, if the treasury does not stabilise debt by corresponding future primary surpluses, then monetary policy has to accommodate it by printing more money, letting go of inflation control and ensuring that nominal debt is never defaulted on (passive monetary and active fiscal policy). In this sense the price level is ultimately determined by fiscal policy decisions.

It is here that we also see that helicopter money is at its core fiscal policy: a permanent fiscal stimulus equivalent to choosing a lower present value of future primary surpluses, which needs to be accommodated by higher prices today. If inflation becomes a concern in the future, it will only be through raising new taxes that policymakers will be able to contract their outstanding liabilities and rein in prices (e.g. by recapitalising the central bank), defeating the very permanency of the initial stimulus. This illustrates the limits to and the difficulties of engineering a genuinely permanent helicopter drop of money.

A counterargument

Notwithstanding the above limitations of helicopter money, it can still prove to be a very potent tool, especially in a liquidity trap. In addition, one can argue that the helicopter money operation together with lowering the present value of primary surpluses leads to a one-off AND permanent change in the level of prices, but in itself has no implication about the future rate of inflation. While the above arguments demonstrate how inflation control can be impaired by sufficiently negative central bank equity, they also demonstrate that at the end of the day it can be restored if fiscal policy decides so (by increasing primary surpluses). Of course, this would question whether the initial stimulus was permanent in the first place, an essential ingredient of true helicopter money. But what does it mean to be “permanent”? Combating inflationary shocks requires a change in the present value of primary surpluses, irrespectively of their initial level. I.e even if policymakers are eventually forced to tighten policy, they would have had to do so even in the alternative scenario where no helicopter money had occurred in the past. The only difference is that the tightening now occurs against a benchmark with a lower path for primary surpluses. In other words, the effect of helicopter money on the level of future primary surpluses, as well as on the price level,  was indeed permanent, even if maintaining control over future inflation might require variations in the opposite direction later on.

In any case, the transmission and effect of helicopter money depends crucially on how economic actors perceive its permanency and how their expectations about the future path of primary surpluses is changed. We have seen that just because spending was financed by new money, it does not necessarily guarantee its irreversibility, so this in itself might not be enough to convince households outside of a perfect foresight equilibrium.

Published inArticle