“Helicopter money” (HM) is a term invented by Milton Friedman to refer to the fantasy of a monetary authority (central bank) distributing money – ex nihilo, in the absence of any related asset acquisition – as a way of stimulating aggregate demand. It’s an idea that’s permeated the blogosphere in recent months, in the wake of the financial crisis, for obvious reasons.
To take a recent example, Brad DeLong writes the following in support of interpreting HM as fiscal policy executed by central banks:
“The actions of central banks have always been “fiscal policy” in a very real sense, simply because their interventions alter the present value of future government principal and interest payments.”
This is a somewhat desperate rationale for the institutional legitimacy of central bank helicopter money. All financial assets are affected by the general level of interest rates as influenced by central banks – not just government debt. Moreover, government treasury departments do not design their funding strategies based on the present value of outstanding liabilities. They manage debt strategy according to the requirements of the cash budget, including servicing of outstanding debt. They don’t do this by trading a massive portfolio of present value risk. The present value perspective, while maybe attractive to academic sensibilities, is not directly important to the real world fiscal management of government cash flow. Scenario based projections of cash flow are the more relevant focus for budget risks.
Economists in recent times have manufactured a number of notions in an attempt to frame helicopter money as the natural purview of central bank monetary policy. Mainstream economics seems determined to abandon the intended institutional division of responsibilities between fiscal and monetary policy, while presenting contorted interpretations of responsibility to justify central bank authority to implement Friedman’s helicopter strategy. In this particular case, “present value-itis” constitutes one of the flimsier rationales, invoking ephemeral discount mathematics instead of examining actual fiscal and monetary operations under intended institutional responsibility and coherent policy formulation and implementation.
Helicopter money at its core is a fiscal idea corresponding to fiscal responsibility. Crediting commercial bank money accounts or even “dropping” currency ex nihilo to selected private sector recipients – with no financial assets received in exchange – constitutes a transfer payment, which according to common sense is a form of government spending. And spending is part of fiscal policy. So helicopter money is about spending from the government budget, with financing provided by the creation of new money, instead of taxes or debt. The transfer of money from the government to the private sector, without financial assets in exchange, is a fiscal transaction. It is no less fiscal than the same amount of money used by the government to pay for public works. It is no less fiscal than would be the case if financed by bonds or by taxes.
Friedman’s idea includes “money financing” (i.e. money creation) as an essential characteristic. I won’t address the effectiveness of money versus bond financing in this essay. That question is important and pertinent to both helicopter money and quantitative easing (QE), with a choice that warrants plenty of debate in its own right, in both cases. But I consider here only the use of fiscal and monetary architecture in implementing the idea underlying helicopter money – i.e. including the implicit assumption that some method of money finance for fiscal distributions has merit and is preferred to market bonds as the ultimate form of finance in the circumstances. In that context, the substance of the HM idea requires a straightforward justification as to why a program of fiscal transfers (or other types of spending foreseen under an HM proposal) is appropriate, and then why the associated money financing is appropriate as a facilitation. Assuming such justification, I conclude that the usual HM proposal can be converted to an accelerated or co-ordinated QE process with targeted fiscal expenditure by Treasury and immediate money financing by the central bank. This is contrary to most proposals for HM that envisage the central bank as making direct fiscal expenditures on its own account.
Money financing falls naturally under the operational framework of monetary policy. But fantasy helicopter drops are more than money financing. They are also government expenditures. It happens that the instrument of spending coincides with the instrument of financing. But spending and financing are not the same thing. They are no more the same thing than a bank loan and the bank deposit that it creates. They are no more the same thing than macroeconomic investment and the equivalent quantity of macroeconomic saving that it creates.
Central banks normally expand the monetary base by creating loans or acquiring securities. They do this in response to the needs of banks and the public for reserve balances and currency. In recent years they have extended this practice to the unconventional mode of quantitative easing, acquiring financial assets to supply reserve balances and bank money beyond conventional requirements.
This does not mean that any imaginable means of expanding the monetary base falls within the exclusive purview of monetary policy. In particular, it does not mean that a fiscal operation that is somehow jiggered to effect an immediate monetary base effect suddenly falls within the purview of monetary policy. But that is what the helicopter proponents want to argue.
The true role of accounting is fundamental to understanding these distinctions. Accounting reflects operations and operations reflect policy. If a central bank undertakes a fiscal expenditure operation, it will deliberately blow a hole in its balance sheet and create a negative capital position and a mismatched balance sheet. This activity is obviously not legitimate as a delegated responsibility for central banks. It is a usurpation of fiscal responsibility that produces a corresponding deformation of the central bank balance sheet. Accounting reflects operations and operations reflect policy. This is policy hijacking.
By contrast, the usual government treasury financial position is that of a net liability profile resulting from deficit financing over time. This profile is naturally “unbalanced” in the context of how we usually think of balance sheet management and accounting. But this institutional exception is easily understood by comparison with the normal private sector case. The government has the power of taxation, and therefore can manage a net liability position and sustain it in a way that isn’t generally possible in the private sector. It has a claim on the future that just isn’t available to the private sector.
Empirical evidence abounds for the effective sustainability of such a net government debt profile over time. By contrast, holders of private sector equity will obviously suffer from claims on negative book equity, because they have no similar claim to the backup role of taxation. Indeed, private sector taxpayers instead are in effect the last resort equity providers for the government sector, bound by the role of taxation as the ultimate means for controlling a sustainable net debt exposure, which ironically is what allows that net debt position to grow over time. Taxation is in effect a transfer of equity value, notwithstanding the ongoing negative book equity position that is typically the case for the transferee.
The central bank is different. Its balance sheet profile is suited to the facts of its delegated responsibilities. Its presumed independence exists only as a function of that delegated authority and law. And provided the central bank operates with that law, it can operate otherwise independently from a fiscal authority that has mostly separate operational responsibilities. The issue of the appropriate or sustainable size for the fiscal authority’s net debt position (including debt held by the central bank) is separate from the delegated responsibilities of the central bank.
The central bank’s function as an institution is to manage the production of the monetary base (i.e. manage its balance sheet) through financial asset acquisition and short term interest rate control. It has a delegated authority with laws under which to do this. It has an integral balance sheet with coherent financial accounting in order to track its performance under this authority. Its capital position is the essential connection to an overarching fiscal framework, including the regular mechanism for distributing central bank profit to the government treasury function. That profit is the fiscal effect of executing monetary policy under a delegated central bank authority.
The economics profession lately likes to argue that because the balance sheets of central banks are not necessarily subject to the same harshness of effective solvency constraints as private sector balance sheets, they should be freed up to implement helicopter money. But this overlooks the intention of the institutional design. That design, which includes a coherent balance sheet with a well-defined capital position, is not the product of some naïve view of government solvency economics. We obviously know that central banks if forced/allowed could continue to operate on the basis of their liquidity production powers, whether or not capital positions are technically positive. But a coherent central bank balance sheet reflects the objective of transparent accountability in how their operations affect the government fiscal position. That effect flows through regularly from the capital position to the Treasury in the form of distributed profit, reflecting for the most part a positive net interest margin and seigniorage benefit.
Conversely, a central bank with a negative capital position (for example, as would be a consequence of HM) is operating in an asymmetric world of shrouded accountability – where net fiscal benefits have been passed on to the Treasury but net fiscal expenses (such as would occur with helicopter expenditures or transfers) have been buried in a mismatched central bank balance sheet, instead of being precluded at the outset by consistent policy, or at least recovered by recapitalization in the form of a government bond distribution from the fiscal authority. In this asymmetric accountability mode, profits from central bank operations are recognized in the government budget while losses are buried in the form of disappearing bonds and a mismatched balance sheet. This asymmetry is no model for proper accountability of central bank operations.
An HM expenditure or transfer as proposed is a net fiscal cost, because it has not been financed with taxes. This would be the case whether the money is distributed by Treasury or by the central bank. Note that this criterion does not directly depend on whether the ultimate financing consists of market bonds or central bank money. The fact that HM as proposed uses money financing reflects the usual case that the central bank issues money rather than its own bonds. But if it did issue bonds instead of money, an HM distribution would still be a fiscal cost.
Not only is an HM distribution a net fiscal cost at the point of disbursement, but it results in further net fiscal costs as a result of the future interest cost of deficit financing. And again, this deficit accumulation over time is the case whether the money is distributed by Treasury along with bonds or by the central bank with or without bonds. All fiscal deficit financing results in a future cost (unless reversed by future surpluses) in the form of either the interest paid on market debt or the interest paid on bank reserves (except for growth in currency issued). (The latter alternative holds both in the actual case of the net interest cost when bonds are held by the central bank and in the fantasy case of HM with negative central bank capital.) In today’s unusual slow growth environment, the interest rate on either debt or bank reserves conceivably might be positive, zero or negative. But zero bound adventures, however long they last, should not obscure fact that HM distributions as proposed are as much a fiscal transaction as are bond financed expenditures from Treasury.
Consider a comparison of quantitative easing and helicopter money.
The actual observable case of quantitative easing (QE) is that of a central bank strategy that operates according to a recognizable division of responsibilities for the execution of the government budget and the ultimate financing of the budget. The QE process is time sensitive in terms of the order of execution of respective fiscal and monetary responsibilities. The government first executes fiscal policy using normal financing requirements in the form of bonds. The net fiscal effect of the initial money distribution, financed by bonds, is captured at the Treasury level. The central bank then executes monetary policy responsibility (at its own option) by converting (in effect) initial fiscal bond financing to money financing. The central bank buys the bonds following the fiscal implementation, and then holds the bonds on balance sheet. The result is tracked on the central bank balance sheet and income statement, including the effect on profit and the capital position. The future net fiscal effect of the ultimate financing method is thus captured at the central bank level. Profit is remitted to the Treasury, the same as in the conventional case. The future net fiscal effect of QE can then be gauged for its impact on central bank profit and remittance to Treasury, as has been evident in the case in the Federal Reserve’s outsized distributions to the US Treasury over the period since QE inception. (Of course, the original starting net fiscal effect of the government disbursement at the time of the original expenditure and associated bond issuance is the amount of the disbursement itself.)
The imagined case of helicopter money (HM) as generally proposed is different from QE, in part due to the different timing for the fiscal and monetary stages of this sequence. The helicopter story specifies that money expenditure and money financing are effectively coincident, rather than lagged as in QE. And because of this coincidence, its proponents seem to suggest that the central bank control both the fiscal function and the monetary function. This proposal amounts to a reverse takeover of policy and operating responsibility, guided by the deemed urgency of combining the fiscal effect with immediate monetary financing. The fiscal nature of the process as we have already described is defined by the fact that a disbursement is made with no other financial asset received in exchange. The central bank is imagined to produce monetary financing ex nihilo – with no other financial asset received in exchange. This runs parallel to Treasury bond financing in the normal course of deficit finance – also with no other financial asset received in exchange. By endowing this ex nihilo disbursement capacity to central bank operations, the HM proposal requires that the central bank now run a mismatched balance sheet, with no central bank asset corresponding to the helicopter monetary base expansion. In the case of QE, the net fiscal effect of the government disbursement is the amount of the money disbursement that has been rerouted (in effect) by bond issuance. In the case of HM, the net fiscal effect of the central bank disbursement is also the amount of the money disbursement, in this case money that has been newly created by the central bank.
The salient point in the comparison is that the chaotic central bank operational and financial architecture that is suggested implicitly for HM amounts to a shell game that submerges important long term fiscal contingencies. The generally accepted argument in favor of HM requires that the money financing be “permanent” (a la the Krugman liquidity trap thesis in the case of QE). It is not even clear what this means in terms of substantive balance sheet projections. Nevertheless, this mode of “permanence” suggests some pretty wobbly central bank balance sheet accounting and management as a means of facilitation. Any central bank governor agreeing to this condition would by implication be committing not to undertake reversing actions in the future that would drain the excess bank reserves that must inevitably result from HM. (Even in the original Friedman super-fantasy of an HM currency drop, it should be acknowledged that the public has a propensity to convert excess currency holdings to more manageable bank deposit balances).
With that kind of committed inflexibility, the central bank as an institution may as well be wound up and folded into the Treasury function, because it can no longer claim to have the same order of influence over monetary base growth. It is supposed to commit by implication to a “permanent” creation of excess bank reserves, whatever that means in terms of substantive balance sheet projections and scenarios. This is opposite to the type of control normally assumed in the case of QE, where the central bank independently constructs a plan for eventual QE “exit” with a corresponding withdrawal of excess bank reserves, depending on future monetary conditions. (The Fed is certainly doing this type of contingency planning now.) And even if in the case of HM there is an “off-balance sheet” commitment to recapitalize the central bank (at the option of either Treasury or the central bank), the obviousness of the shell game is still evident. The only purpose of such a ruse can be to avoid the creation and visibility of outstanding government debt, even though the net fiscal effect of debt while held by the central bank can be easily explained as neutral – in and of itself (the bond interest received by the central bank equals interest paid back to Treasury as part of profit). Thus, there is no fiscal substance to be found in the idea of making bonds “disappear” in an HM operation any more than there is in QE. The proposed absence of bonds is a ruse designed to fool people into thinking that there is no fiscal contingency associated with money financing. But this is not the case. The consolidated fiscal position includes the remittances that the central bank makes to Treasury, and those remittances are at risk in the future if the central bank increases the interest rate it pays on bank reserves – which it will do if and when monetary conditions allow interest rate lift off from the present “liquidity trap” situation. That sort of lift off is in fact the prevailing long term objective of monetary policy almost everywhere now, given the operational and strategic complications associated with a central bank being stuck at the zero lower bound. And most importantly, central banks do not want and should not want to commit to a state of permanently saturated banking system reserves. That would be a signal of secular institutional dysfunction (banks do not need those reserve balances for efficient functioning) and a cry for help in the form of more fundamental institutional redesign. (This again is important in the Fed’s current thinking.)
In contrast to the HM notion, the actual case of QE involves an effective transformation of financing for a government budget that has already been rolled out – the money has already been spent. QE converts the financing from bonds to money. HM as imagined has the same monetary effect, but with no time lag between the fiscal expenditure and the financing with money. However, the core spending component is fundamentally the same. It is fiscal spending – typically proposed in the form of a transfer. And so there is nothing to prevent the same kind of spending that is envisaged in the helicopter parable from occurring within normal debt operations or in the same sense as QE debt/money conversion, under a compressed time horizon that amounts to the coincidence of spending and money financing as proposed under HM. Treasury and the central bank both have the operational flexibility to achieve the desired result in terms of coordinated spending and ultimate financing – i.e. the flexibility to achieve the objective under existing institutional arrangements and responsibilities, thereby synchronizing the timing of the desired fiscal effect and its ultimate monetary financing. And in doing so, Treasury can then be consistent in assuming primary responsibility for the policy that directs the kinds of fiscal payments that are proposed under HM. The policy is fiscal, with monetary accommodation as planned in execution of the policy.
For example, under a policy of such operational coordination, the central bank can buy bonds from the market in an amount equal to the fiscal expenditure at or around the time of that expenditure. In a more extreme adaptation, the law might be changed to accommodate direct monetary financing at the point of issuance – but only for this designated type of fiscal program. That’s a bigger leap, but it beats the alternative of a central bank that independently blows a hole in its balance sheet with direct fiscal expenditures, mangling the institutional alignment of responsibilities across the board in the process. That said, it is not essential to the objective of money financing that the bonds acquired be the same bonds issued at the point of fiscal expenditure. The essence of the idea is to create monetary financing in coordination with the fiscal expenditure and at the time of the expenditure. It is the quantity of bonds acquired and money created at the same time rather than the matching of bonds acquired to bonds issued that is important to the idea.
Conversely, the HM idea as proposed leads to central bank balance sheet distortion due to the capital effect of ex nihilo fiscal disbursements. As a possible “fix” for this potential effect, Treasury might recapitalize the central bank balance sheet as ongoing disbursements are made. The balance sheet capital hole created by fiscal spending could be refilled by recapitalization with bonds that serve as the nominal asset backing for the bank reserves that have been created. This effectively transfers the balance sheet mismatch and the fiscal cost resulting from the central bank disbursement back to Treasury, as is the normal case for fiscal expenditures. But this is an awkward correction to a financial management arrangement that is unnecessary and dysfunctional at the outset.
Instead of that type of fix, the problem should be avoided at the beginning. The fiscal disbursement – in this case the “HM drop” – should be made by Treasury, as is normally the case with fiscal disbursements. And bonds should be issued by Treasury as usual. The truly effective “drop” characteristic of HM is in fact a function of the money that is disbursed to recipients – whether that money is disbursed by the central bank or by Treasury. The public receiving the money is interested in the money only, and has no interest whether or not bonds are issued. And when the bonds are issued as usual, they (or other bonds, as noted above) can be vacuumed up immediately by the central bank. After all, it is not necessary in order to achieve the substantive fiscal “drop” effect that the money paid to the drop recipients be the same money that the central bank creates. This is a matter of procedural arrangement, accelerating the time sequence of normal QE, as described above, while still achieving near-simultaneous coordination of the fiscal transfer and the end profile for money finance. This central bank action of compressed-time-QE purchases of HM-linked bonds is what increases the aggregate supply of bank reserves (and broad money in the usual case of non-currency disbursement). But it doesn’t change the fact that the recipients will have already received HM “drop money” and couldn’t care less themselves whether bonds have been issued in the process (another hint as to why this is at root a fiscal exercise). Moreover, the consolidated fiscal effect of the bonds per se is zero, so long as they are held by the central bank. This once more reinforces the essential fiscal nature of the decision to spend money that is at the root of the HM idea.
To recap, given the urgent timing desired under the HM disbursement idea, proper recognition of the primary fiscal effect requires either simultaneous recapitalization if the central bank writes the check, or simultaneous QE-type bond purchase if the Treasury writes the check. At the end of the day, there is no reason for the central bank to write the check for the fiscal disbursement, since the very same result can be achieved by fully coordinated expenditure with money financing, using compressed-time QE. That obviates the need to “fix” the fiscal effect of a central bank transaction that needn’t and shouldn’t have taken place in the first place.
As is the case with QE, there may well be a fiscal benefit resulting from the HM concept when compared to the scenario of market bond financing – if and so long as the interest rate paid on bank reserves is less than the interest rate paid on alternative market bonds. And as with either market bonds or QE bonds, the gross debt to GDP ratio is a calculation that includes all bonds issued by the government, including those held by the central bank. But an intelligent interpretation of the gross debt position should always consider the effective transformation of the interest rate cost in respect of any bonds held by the central bank. And just as important, the analysis of the fiscal position should recognize balance sheet risks and contingencies over the longer term, such as potential QE exit or some similar reversing action in the case of HM.
HM as visualized suggests the illusion of a “permanent” free lunch. But future interest payments on HM created bank reserves are no less risky than future interest payments on QE created bank reserves. And while there may be a difference between the interest rate paid on bank reserves and the interest rate paid on Treasury bills (or longer term bonds), the essential nature of the interest rate risk and the fiscal consequence that results from the monetary policy for both QE and HM is fundamentally the same. This interest expense is part of the net profit determination for the central bank and its consequent remittance to the Treasury, and therefore is part of the net fiscal effect of central bank operations. While the typical taxpayer will not fully appreciate the workings of the consolidated government balance sheet or the consolidated interest cost of the government budget, he or she can probably sense in at least a vague way institutional changes (e.g. negative central bank capital) that seem benign enough when interest rates are near the zero bound, but which mute an awareness of expense contingencies down the road in the form of potentially higher interest costs on bank reserves. These prospects should be assessed and explained as a matter of thorough financial analysis and risk management, rather than hidden in the murky accounting world of the typical HM proposal.
Numerous economists believe that the helicopter money idea constitutes “out of the box” thinking, and that the time has come to implement Friedman’s wacky analogy as a central bank undertaking. But if the proposal is understood for its substantive implications, it can only be interpreted rationally as fiscal policy with immediately co-ordinated monetary policy. The usual proposal sees the disappearance of bonds from the central bank balance as a catalyst for a central bank claim of fiscal responsibility for helicopter money. But what responsible central bank governor would agree, in the absence of bonds held as assets, not to have a documented bond recapitalization contingency plan that would at least allow for the possibility of draining HM created excess bank reserves at some point in the future? And why should he or she even agree to that, when the bonds that should already be in place contribute zero net interest expense for the consolidated fiscal cost, so long as they are held by the central bank? Do economists really believe that “expectations” become more efficiently informed if, for the sake of a superficial impression of “permanence”, the capacity to deal effectively with future monetary policy contingencies is left ambiguously adrift? This amounts to irresponsible financial management, with an unnecessary and inappropriate imposition of stark inflexibility on central bank authority, where the only “benefit” is the avoidance of an apparent need to communicate that the payment and receipt of interest on government bonds held by the central bank has zero net fiscal effect so long as the bonds are held in the central bank’s portfolio.
The debate about HM has featured much discussion and tinkering as to what constitutes monetary policy and what constitutes fiscal policy. This has tended toward unproductive manipulations of institutional responsibilities, with old vocabulary in search of new meaning. To take the earlier example again, characteristics as flimsy as the effect of central bank interest rate operations on the present value of government debt cannot be taken seriously as a line of logical justification for fundamental realignment of fiscal responsibility. The proponents of HM seem driven by a desire to avoid the creation of government bonds that in fact have no net fiscal effect on their own account while held in the central bank portfolio. It is as if the entire institutional framework is uprooted simply because somebody can’t be bothered to footnote the usually favorable debt servicing cost effect of central bank holdings of bonds when analyzing the debt to GDP ratio.
Yes, the central bank has the responsibility for managing the monetary base. And yes, that along with short term interest rate control is the essence of monetary policy. But that does not mean that the central bank has the exclusive authority to expand the monetary base using any operation imaginable under the sun. The central bank operates under what is in effect a delegated authority from its ultimate fiscal master, somewhat analogous to a risk taking authority that a commercial banking trading unit may receive from a high echelon asset-liability committee. This is an “authority” that is designed with fiscal consequences in mind, such that the central bank has the operational independence to execute monetary policy under that authority and under the laws that define its institutional responsibilities.
In summary, real time coordination of fiscal and monetary policy (as desired under HM) doesn’t warrant a conflation of fiscal and monetary policy responsibilities or incoherent fiscal accounting. HM proponents come at the situation with notions about central banks somehow taking control of fiscal policy and driving the implementation with an unnecessarily destructive effect on their own balance sheets. Immediate money financing for approved fiscal actions is the substantive objective, and the operational means to achieve this objective lies in the design of coordinated fiscal and monetary plans within the existing institutional framework – not in speciously dodging the presence of government bonds on the central bank balance sheet.