Central banks should pursue simpler, less costly and more effective policies. If policymakers want higher spending, they should transfer cash directly to households.
Some advocates of helicopter drops have argued that central banks should “permanently” finance larger budget deficits, but to some extent that misses the point. If there was a consensus in favour of fiscal stimulus we wouldn’t face the current weakness in demand. We need new monetary policy tools precisely because the fiscal authorities can’t get their act together, and because existing monetary measures have limited effect. In the Eurozone, what many would regard as effective counter-cyclical fiscal policy is not even legal. Forget about monetary and fiscal coordination.
We need to define legal policies for the ECB which are likely to work.
In our Foreign Affairs article, Mark Blyth and I also suggest that monetary policy has many advantages over fiscal policy – the speed of decision-making and immediacy of impact are non-trivial. An additional problem with coordinating fiscal and monetary policy is that freedom from electoral and political pressures is part of the case for having independent central banks in the first place.
The real problem with monetary policy is that its tools are not fit for purpose. This simple observation seems to missed by many theorists. Open-market operations were designed in the 19th century, with the purpose of developing government bond markets and providing liquidity to banks. They are not designed to directly affect spending.
Interest rates and bond purchases work indirectly, through asset prices and financial intermediation. When policy rates are close to zero, and credit growth hampered by demand and supply, they barely work at all. For this reason, talk of nominal GDP targets, or commitments to raise the inflation target, are a distraction. Not because central banks don’t want to – but because they can’t. So the real case for advocating cash transfers is the immediate and direct impact on spending, in the same way that lower mortgage rates boost household cash flow in the UK’s predominantly variable rate mortgage market, or mortgage refinancing does in the US.
The administrative challenges of implementing cash transfers are considerable. Since our article, John Muellbauer has argued that the ECB could use social security numbers or the electoral register. More recently, I have suggested that the ECB could do TLTROs for all. This has the advantage that the ECB would just be extending an existing facility: it has already eased collateral requirements, extended maturities, and attached targeted criteria to its lending facilities. Why not open them up to individuals, at infinite maturities and zero rates, up to a maximum of, say, €1500 per borrower. There is precedent for central banks offering credit to individuals, and infinite maturity loans with no interest don’t have credit risk, so there is no need for collateral or government indemnity.
Others have objected that cash transfers – even if implemented by central banks – are really “fiscal” not “monetary” policy. This is a complex question because the distinction is not clear – and now is not the occasion to address it. From a practical, European, standpoint let’s just accept the legal distinction. A zero interest rate, infinite maturity, TLTRO to individuals is certainly not fiscal policy, as defined by Eurozone law – because no government is involved. The ECB is not incurring credit risk, so there is no blurred lines regarding its balance sheet, and if implemented in order to meet its inflation mandate it is legal.
One of the most articulate opponents to helicopter drops without a price level target, is David Beckworth, who writes this superb blog. If Beckworth is correct, the ECB has a problem. It clearly cannot commit to keep future inflation above 2%.
Beckworth seems to have two concerns: 1) in order to work, cash transfers or any equivalent, have to be “permanent” and 2) unless the ECB allows inflation to go above target, any effect will be offset by higher interest rates.
Beckworth’s arguments are largely technical, but his concerns about inflation resonate more generally. At least he has a clear model in mind: the recovery caused by the payments will reduce the output gap, cause higher inflation and then be offset by higher rates. In brief, this strikes me as oddly circular: the policy will fail, because it works. Why would the ECB raise rates to prevent its own policy from working? I am equally sceptical of a mechanical relationship between increased spending, a shrinking output gap and higher inflation. David Reifschneider’s hypothesis that supply may well be endogenous is even more plausible when a combination of labour market deregulation and technological innovation is occurring. A strong recovery in demand in Europe could well bring unanticipated improvements productivity.
Interestingly, only economists focus on the issue of whether or not helicopter drops are “permanent”. Most people, rightly I think, don’t really understand what it means to ask “is a cash transfer ‘permanent’?” This is the language of fiscal policy. Tax cuts can be temporary, but a payment from a central bank? It’s not clear what “permanent” means in this context.
The literature focuses specifically on whether the increase in the monetary base will be reversed at some point in the future (or perhaps more accurately, will it be permanently higher than it would otherwise have been). Beckworth documents this point at length. He could also have cited Milton Friedman in Paper Promises, when he describes the fictitious helicopter dropping money, he calls it as a “permanent” increase in the stock of money.
There are obvious practical flaws in this way of thinking. For example, it is not clear how we would identify when an increase in reserves was “permanent” relative to what reserves would otherwise have done. In order to do so we have to know what the demand for reserves is doing.
But what is really odd about this fixation on the durability of the change in reserves, is it assumes that the quantity of reserves is everything, and ignores how and why the reserves are created.
To make this explicit, let’s look at a Constant Reserves Multiplier (CRM). Consider the following: let’s assume the Fed decides to target a constant stock of reserves, but completely change the way it is creating them. It shrinks the monetary base, by 5% of GDP, through reversing some QE: i.e. by selling some of its treasury holdings. At the same time, it makes transfer payments to the household sector also equivalent to 5% of GDP, increasing reserves. The net effect on reserves is zero.
The issue of permanence is no longer relevant, because the stock of reserves is unchanged.[1] There is nothing to reverse. A mechanical “money-printing creates inflation” relationship cannot be relevant for the same reason, because no money has been printed. (I think this also addresses Tony Yates’s main concerns.)
The interesting question is: would demand rise? The empirical literature is compelling, and the CRM is likely to be strongly positive. How do we know this? The effect of a cash transfer from the Fed to the household sector is almost identical to receiving a tax rebate check, with the added benefit that it causes the budget deficit to fall, not rise (Tax revenues will rise and there has been no increase in government borrowing).
All the evidence on US tax rebates shows substantial impact on consumer spending. The effects of QE are contentious, but even the most fervent believers would expect little impact on demand from a withdrawal of QE equivalent to 5% of GDP. Cash transfers are much more likely to have the effects of a “regime change” that Beckworth alludes to and Christina Romer has written about.
I should be clear that none of this should be taken as implying that expectations or asset prices do not matter. Far from it. On balance, the asset price response, certainly in the Eurozone, is likely to be a positive stimulus to demand. It is conceivable that a sharp rise in government bond yields could follow a significant cash transfer to households. This would logically reflect an increase in future real interest rate expectations. But there are some important caveats. There are very elevated cyclical risk premia embedded in many Eurozone asset prices. That is why the Eurozone’s cost of equity is high. It is also likely that the remaining credit spread of Italian, Spanish and peripheral bonds over German bunds is also cyclical. Cash transfers would likely trigger a rapid rise in equity markets, because earnings are currently cyclically depressed, so the asset price effect of cash transfers would likely be way more powerful than any impact of “small” amounts of QE.
If markets believe that the threat of a deflationary spiral is mitigated by a recovery in demand, profits and tax revenues, many risk premia will compress. In simple terms, shifting expectations will cause financial conditions to ease, even if the unobservable risk free rate rises. The impact on asset prices from such a shift in policy gears in the Eurozone would likely dwarf any negative bond price effects.
Those who worry that the increase in reserves caused by cash transfers to households will cause inflation or create major central bank balance sheet problems down the road, no longer need to oppose this policy. Nor do those who worry about – or want – the monetary base expansion to be permanent. All the evidence suggests that boosting household disposable incomes has way more effect on demand than QE. So there is actually no need to print any new money.
[1] An adjustment could even be made to account for the small loss of net interest income.
Leave a Reply