There ain’t no such thing as “free” helicopter money
10 November 2016
SUGGESTED
Economic Theory
Economic Theory
11 November 2016
Economic Theory
25 June 2024
It is sometimes claimed that helicopter money (HM) is “free” money but it is not. In this posting I would like to explain why.
Let’s start by defining our terms. HM is often described as the central bank “printing” money and giving it away, e.g., as if via a helicopter drop. This description is a little misleading, however. In most proposals, HM would involve the central bank issuing and giving away not physical cash, but a digital or credit equivalent. For example, it might issue every citizen with the digital equivalent of $1,000, which might be delivered as payments directly to individuals’ bank accounts; alternatively, it might send everyone checks in the post. Or it might digitally “print” the money and give it away to the government to spend.
Now to say that something is “free” is to suggest that a valuable asset with a positive exchange value can be conjured up out of nothing, i.e., that that asset has no tangible opportunity cost. If helicopter money really is “free” in this sense, then we have effectively found the monetary Holy Grail: helicopter money would become the policy instrument of choice, to be preferred above all others.
It is therefore important to get to the bottom of this question.
Now helicopter is obviously “free” as far as its recipients are concerned: they get something for nothing. In the case of HM, however, the opportunity cost is lost seigniorage, the foregone profits from the issue of base money.
To appreciate this point, consider that when a central bank engages in a HM operation it acquires a notional bond ‘asset’ – think of this as a zero-coupon perpetual bond – that promises no future payments. However, the central bank could have issued the same amount of additional base money to acquire a conventional bond that did promise future payments. These future bond payments are the foregone seigniorage profits from HM.
One might object that these foregone seigniorage profits do not matter to the central bank because it hands over its profits to the Treasury. This claim is true but irrelevant. Those foregone seigniorage profits would still matter because they would be lost to the Treasury. Consequently, HM has an opportunity cost and is therefore not free.
This argument about HM not being free also holds true even if interest rates are zero or negative. Suppose that we have zero interest rates and these lead to bonds having zero coupons. In this case, the foregone opportunity is for the central bank to have acquired a zero-coupon bond that repays its face value on its maturity date. That terminal payment is foregone seigniorage. In the case of bonds with negative coupons, that same terminal payment will still be foregone seigniorage, albeit now offset somewhat by the negative coupon payments. However, it will still be the case that the present value of the terminal payment will outweigh the (negative) present value of the coupon payments – otherwise the bond will have no positive price – and so those forgone payments will still have a positive present value, i.e., there is still foregone seigniorage.
That HM is not free can also be seen by comparing the impact of debt monetisation (DM) and HM on the consolidated government balance sheet, considering the Fed as part of the government. Consider two cases:
Case 1 (DM) is where the government runs a deficit that it finances by selling a coupon-paying bond to the central bank, which “prints money” to purchase it. To facilitate the comparison, let us assume that this bond is a perpetual one.
Case 2 (HM) is where the government finances a deficit by directly “printing money”. This case is equivalent to the government “selling” the central bank a zero-coupon perpetual bond that has a market value of zero.
The only difference is that in Case 1 the government makes coupon payments to the central bank, whereas in Case 2 it does not. Yet from the perspective of the consolidated government balance sheet this difference is merely a difference in transfers from one branch of the government to another. In Case 1, the government makes the coupon payments to the central bank and the central bank remits its seigniorage profits to the government. In Case 2, the government makes no coupon payments to the central bank and the central bank remits its profits to the government, but those profits are less in Case 2 by exactly the amount of the coupon payments in Case 1. Consequently, both operations have an identical impact on the consolidated government balance sheet.
Now no-one would argue that Case 1 involved “free” money or a zero opportunity cost, but since the two cases are functionally indistinguishable, one cannot argue that the former has an opportunity cost but the latter does not.
John Kay makes a different but complementary explanation for why HM is not a free lunch:
For every credit there is a corresponding debit; for every financial asset there is a corresponding liability. The double entry principle is [an] immutable as the first law of thermodynamics, to which it bears a certain resemblance.
You cannot create financial assets out of nothing, just as you cannot create energy out of nothing. There is no financial free lunch, just as there is no perpetual motion. The laws of thermodynamics render perpetual motion machines impossible, though it may be hard to identify the defect in any particular device. And similarly, schemes which purport to fund government spending or tax cuts at no cost to present or future taxpayers are necessarily fatally flawed, even if it requires a little effort to work out the flaw.
Actually, it is not very difficult to identify the flaw in the idea of stimulating the economy by dropping ‘helicopter money’. The confusion begins because fiat money – the coin and paper currency that we use in everyday transactions – is manifestly an asset; and yet the promise – to pay the bearer one pound on demand – appears meaningless. If you turn up at the Bank of England to demand one pound, the teller will smile benignly and exchange it for an identical coin.
But notes and coins are a government liability nonetheless. You can use them to discharge your liabilities to the government – to pay your taxes … (his italics)
However, an objection should be considered. This is the argument that HM is “free” because it produces some net benefit. An example is Bossone (2016), who dismisses the HM-is-not-a-free lunch arguments of Borio et alia (2016) in these terms: “HM is a “free lunch” in the simple sense that, if it works and succeeds in closing the output gap, people won’t have to repay it through higher taxes or undesired (above optimal) inflation.”
I don’t buy this argument. The problem with it is apparent from the word “if”: it presupposes that HM will have the effects hoped-for by their proponents, and we cannot be sure it will. Counterexample: John Law’s Mississippi scheme, which produced a short-term boom, but ruinous longer-term effects. Law’s scheme might have looked like a free lunch while it was in play, but not with the benefit of hindsight when its true costs became apparent. Put more broadly, by Bossone’s argument, any macroeconomic policy can be justified by its proponents as providing a free lunch provided they merely assert that it would produce some net benefit.
I would argue that this argument is too broad, as it could be used to justify any policy whatever as providing an alleged free good, however pointless or destructive it might be. By this logic, it would make sense to build bridges to nowhere because they boost NGDP or to manufacture financial bubbles that might seem a good idea at the time but ultimately turn out to be destructive.
More specifically, this argument confuses a (tangible) free lunch with a questionable net benefit, and these are not the same. The one is tangible, indeed quantifiable; the other involves a speculative projection and may not even exist.
The difference between them is also apparent from an elementary set comparison. We have three sets: Set A, the set of true (or certain) net benefits, Set B, the set of free lunches, and Set C, the set of claimed net benefits. Taking the meaning of the former to be self-evident, then Set B is a subset (though not necessarily a proper subset) of Set A, but Set C is not. Set C merely overlaps with Set A because some claimed net benefits will actually be true and others will not be true. To confuse a net benefit and a free lunch is thus to make a category mistake.
To repeat: HM is not a free lunch because it has a tangible opportunity cost – the cost of the seigniorage foregone.[1]
HM might look free, but it ain’t.
—
[1] For the record, I am not suggesting that this seigniorage opportunity cost is the only cost of HM. HM is likely to be very damaging due to the distorted incentives it creates, such as encouraging policy makers to make decisions based on the false perception that HM is “free” and the dangers of unconstrained issue of money. These are additional reasons to oppose HM, but for present purposes I merely wish to establish that HM is not “free” because it does have a tangible opportunity cost.
Let’s start by defining our terms. HM is often described as the central bank “printing” money and giving it away, e.g., as if via a helicopter drop. This description is a little misleading, however. In most proposals, HM would involve the central bank issuing and giving away not physical cash, but a digital or credit equivalent. For example, it might issue every citizen with the digital equivalent of $1,000, which might be delivered as payments directly to individuals’ bank accounts; alternatively, it might send everyone checks in the post. Or it might digitally “print” the money and give it away to the government to spend.
Now to say that something is “free” is to suggest that a valuable asset with a positive exchange value can be conjured up out of nothing, i.e., that that asset has no tangible opportunity cost. If helicopter money really is “free” in this sense, then we have effectively found the monetary Holy Grail: helicopter money would become the policy instrument of choice, to be preferred above all others.
It is therefore important to get to the bottom of this question.
Now helicopter is obviously “free” as far as its recipients are concerned: they get something for nothing. In the case of HM, however, the opportunity cost is lost seigniorage, the foregone profits from the issue of base money.
To appreciate this point, consider that when a central bank engages in a HM operation it acquires a notional bond ‘asset’ – think of this as a zero-coupon perpetual bond – that promises no future payments. However, the central bank could have issued the same amount of additional base money to acquire a conventional bond that did promise future payments. These future bond payments are the foregone seigniorage profits from HM.
One might object that these foregone seigniorage profits do not matter to the central bank because it hands over its profits to the Treasury. This claim is true but irrelevant. Those foregone seigniorage profits would still matter because they would be lost to the Treasury. Consequently, HM has an opportunity cost and is therefore not free.
This argument about HM not being free also holds true even if interest rates are zero or negative. Suppose that we have zero interest rates and these lead to bonds having zero coupons. In this case, the foregone opportunity is for the central bank to have acquired a zero-coupon bond that repays its face value on its maturity date. That terminal payment is foregone seigniorage. In the case of bonds with negative coupons, that same terminal payment will still be foregone seigniorage, albeit now offset somewhat by the negative coupon payments. However, it will still be the case that the present value of the terminal payment will outweigh the (negative) present value of the coupon payments – otherwise the bond will have no positive price – and so those forgone payments will still have a positive present value, i.e., there is still foregone seigniorage.
That HM is not free can also be seen by comparing the impact of debt monetisation (DM) and HM on the consolidated government balance sheet, considering the Fed as part of the government. Consider two cases:
Case 1 (DM) is where the government runs a deficit that it finances by selling a coupon-paying bond to the central bank, which “prints money” to purchase it. To facilitate the comparison, let us assume that this bond is a perpetual one.
Case 2 (HM) is where the government finances a deficit by directly “printing money”. This case is equivalent to the government “selling” the central bank a zero-coupon perpetual bond that has a market value of zero.
The only difference is that in Case 1 the government makes coupon payments to the central bank, whereas in Case 2 it does not. Yet from the perspective of the consolidated government balance sheet this difference is merely a difference in transfers from one branch of the government to another. In Case 1, the government makes the coupon payments to the central bank and the central bank remits its seigniorage profits to the government. In Case 2, the government makes no coupon payments to the central bank and the central bank remits its profits to the government, but those profits are less in Case 2 by exactly the amount of the coupon payments in Case 1. Consequently, both operations have an identical impact on the consolidated government balance sheet.
Now no-one would argue that Case 1 involved “free” money or a zero opportunity cost, but since the two cases are functionally indistinguishable, one cannot argue that the former has an opportunity cost but the latter does not.
John Kay makes a different but complementary explanation for why HM is not a free lunch:
For every credit there is a corresponding debit; for every financial asset there is a corresponding liability. The double entry principle is [an] immutable as the first law of thermodynamics, to which it bears a certain resemblance.
You cannot create financial assets out of nothing, just as you cannot create energy out of nothing. There is no financial free lunch, just as there is no perpetual motion. The laws of thermodynamics render perpetual motion machines impossible, though it may be hard to identify the defect in any particular device. And similarly, schemes which purport to fund government spending or tax cuts at no cost to present or future taxpayers are necessarily fatally flawed, even if it requires a little effort to work out the flaw.
Actually, it is not very difficult to identify the flaw in the idea of stimulating the economy by dropping ‘helicopter money’. The confusion begins because fiat money – the coin and paper currency that we use in everyday transactions – is manifestly an asset; and yet the promise – to pay the bearer one pound on demand – appears meaningless. If you turn up at the Bank of England to demand one pound, the teller will smile benignly and exchange it for an identical coin.
But notes and coins are a government liability nonetheless. You can use them to discharge your liabilities to the government – to pay your taxes … (his italics)
However, an objection should be considered. This is the argument that HM is “free” because it produces some net benefit. An example is Bossone (2016), who dismisses the HM-is-not-a-free lunch arguments of Borio et alia (2016) in these terms: “HM is a “free lunch” in the simple sense that, if it works and succeeds in closing the output gap, people won’t have to repay it through higher taxes or undesired (above optimal) inflation.”
I don’t buy this argument. The problem with it is apparent from the word “if”: it presupposes that HM will have the effects hoped-for by their proponents, and we cannot be sure it will. Counterexample: John Law’s Mississippi scheme, which produced a short-term boom, but ruinous longer-term effects. Law’s scheme might have looked like a free lunch while it was in play, but not with the benefit of hindsight when its true costs became apparent. Put more broadly, by Bossone’s argument, any macroeconomic policy can be justified by its proponents as providing a free lunch provided they merely assert that it would produce some net benefit.
I would argue that this argument is too broad, as it could be used to justify any policy whatever as providing an alleged free good, however pointless or destructive it might be. By this logic, it would make sense to build bridges to nowhere because they boost NGDP or to manufacture financial bubbles that might seem a good idea at the time but ultimately turn out to be destructive.
More specifically, this argument confuses a (tangible) free lunch with a questionable net benefit, and these are not the same. The one is tangible, indeed quantifiable; the other involves a speculative projection and may not even exist.
The difference between them is also apparent from an elementary set comparison. We have three sets: Set A, the set of true (or certain) net benefits, Set B, the set of free lunches, and Set C, the set of claimed net benefits. Taking the meaning of the former to be self-evident, then Set B is a subset (though not necessarily a proper subset) of Set A, but Set C is not. Set C merely overlaps with Set A because some claimed net benefits will actually be true and others will not be true. To confuse a net benefit and a free lunch is thus to make a category mistake.
To repeat: HM is not a free lunch because it has a tangible opportunity cost – the cost of the seigniorage foregone.[1]
HM might look free, but it ain’t.
—
[1] For the record, I am not suggesting that this seigniorage opportunity cost is the only cost of HM. HM is likely to be very damaging due to the distorted incentives it creates, such as encouraging policy makers to make decisions based on the false perception that HM is “free” and the dangers of unconstrained issue of money. These are additional reasons to oppose HM, but for present purposes I merely wish to establish that HM is not “free” because it does have a tangible opportunity cost.
SHARE
