The Financial Times has an article that discusses how central banks could give away money during a recession, in order to stimulate the economy:
In the past, central banks set the price of money using interest rates. In the future, it seems, they will be giving it away. . . . One problem with this common sense idea is its simplicity, which rarely appeals to economists charged with taking important decisions.
I like uncomplicated ideas, but I don’t see how this idea would be “simple”. Who would get the money? You might argue for a neutral policy of everyone getting a check for $5000. But even that raises questions. Does everyone include children, or is it $5000 per household? Those decisions have distributional consequences, and have traditionally been made by elected representatives.
We are already seeing an erosion of democratic accountability. In the EU, unelected bureaucrats make many important decisions. In the US, a President can now spend billions of dollars on a project even after Congress has rejected his request to approve the expenditure. Do we really want unelected bankers in the private sector (i.e. regional Fed bank presidents) making important decisions on how to allocate government spending?
Furthermore, if the Fed gives away the newly created money, how will the monetary injection be withdrawn if and when it is no longer needed? Under conventional monetary policy, central banks buy assets with newly created money. Those assets are then sold when the QE needs to be withdrawn from circulation. If the new money is given away, then the funds to withdraw the money will have to come from the Treasury, which will then have to raise the money via distortionary taxes on the economy.
Alternatively, think of this in an opportunity cost sense. Consider all the wonderful things we could do if we didn’t have to worry about budget constraints. We could do a radical tax reform to make the system far more efficient. Opposition would be “bought off” by making sure almost everyone gets a cut. (Thus imagine switching from making health insurance tax deductible to a simple health insurance tax credit big enough to benefit almost everyone.) We could build lots of infrastructure, or provide subsidies to low wage workers. By giving the money away in a neutral fashion, the central bank would be depriving the fiscal authorities of the funds required for initiatives with a much greater value to the country.
Note that this is not a partisan position that I am taking. Neither political party in the US favors giving everyone a check. The GOP typical favors cuts in tax rates, whereas the Democrats typically favor targeted programs aimed at specific problems. You can certainly argue that giving money away is better than one of these two alternatives, but it’s pretty hard to argue it’s better than both. A Republican supporting this proposal might assume the money giveaway will not lead to higher future taxes, while a Democratic supporter might assume that it won’t trigger future cuts in social spending. But can either side be confident of their assumption? I don’t see how. (I suspect it would lead to both higher future taxes and lower future spending.)
The most intriguing and practically viable idea of all is emerging from the least likely of sources, the ECB: so-called targeted long-term refinancing operations. In straightforward terms, this is a policy of dual interest rates which involves giving money to both borrowers and savers.
This seems even worse than a simple money giveaway. We already subsidize debt in America via deposit insurance and tax breaks. This sort of debt subsidy program would distort the economy by favoring activities that generate debt over activities financed by equity (or non-debt financed consumption.)
There is a much better alternative:
First, countries need to decide the largest central bank balance sheet that is appropriate (as a share of GDP.) Then set a NGDP level target path at a high enough growth rate so that the central bank balance sheet doesn’t go beyond that threshold. For the US, a 4% NGDPLT should be sufficient.
Second, have central banks buy as many assets as are required to hit their NGDP target.
If something seems too good to be true, it usually is. Having the central bank give money to everyone during a recession is a bad idea.
READER COMMENTS
Quite Likely
Aug 2 2019 at 4:48pm
The trouble is that demand stimulus via asset price inflation is the least efficient possible way to stimulate the economy. Having the Fed buy tons of assets is just shoveling money into the pockets of people who are not changing their spending much based on getting an extra marginal dollar. A fraction as much money creation can go much farther if it’s put directly into the hands of people who will immediately spend it, whether that’s individuals getting checks, government entities getting larger budgets, or people in the bottom 90% getting tax cuts.
Don Geddis
Aug 2 2019 at 5:44pm
Fiat currency is “costless”. So what do you mean by “least efficient”? You seem to be implicit assuming some kind of metric such as “dollars of QE per additional dollars of NGDP”. But even if you could calculate such a thing, why is that ratio of any real interest? “Dollars of QE” are free, so who cares what the ratio is?
Fed OMOs are not “shoveling money into the pockets of people”, since the Fed purchases assets at current market prices. The people who sell the assets don’t see their net worth change at all. You try to argue that they are “not changing their spending” — but why would they “change” their spending, since their net worth is unchanged? You talk about “getting an extra marginal dollar”, but again OMOs do not change people’s net worths.
It seems likely, from your wording, that you are imagining some kind of Austrian macro theory where monetary stimulus works via Cantillon effects. If so, that’s your problem. The primary monetary policy transmission mechanism is expectations, and the primary concrete mechanism is the Hot Potato Effect (from larger cash balances). Cantillon effects are trivial and mostly irrelevant.
Matthias Görgens
Aug 2 2019 at 8:51pm
It seems easy for people to imagine some naive Cantillon effects, but hard for them to square them with expectations.
If Cantillon effects are playing a role, it’s on the future net sellers of the assets the central banks is announcing they are going to buy.
That’s people who are currently owning government bonds when a new unexpected QE program is announced. But mostly it’s the government itself which is the largest net seller of government bonds.
(To see my point, imagine a central bank was barred from buying its government’s bonds. Government bond yields would probably be higher without that implicit ‘put’ below bond prices.)
ChrisA
Aug 4 2019 at 6:02am
I just don’t get this resistance to retirement of government debt via QE. What if, for whatever reason, the Government was running a surplus, would it be wrong to buy up Government bonds to retire debt because that would drive up bond prices? Why is it so much better to have high interest rates and high debt levels? It seems to me that opposers to QE are just desperate to find a problem with it.
Scott Sumner
Aug 2 2019 at 6:41pm
I’m certainly not recommending demand stimulus via asset price inflation. Look at the impact of the monetary injections of the 1960s and 1970s on bond prices—they plunged in response to the purchases.
Michael Sandifer
Aug 3 2019 at 1:02pm
Scott,
Assuming sufficient central bank credibility, could a central bank simply credit accounts without making purchases and then deal with overshooting simply by promising to create less money in future months or years?
Scott Sumner
Aug 3 2019 at 5:31pm
That might or might not work, it depends on how much money was created. But it’s a policy that a central bank should never do. Our elected representative should do fiscal policy.
Michael Sandifer
Aug 4 2019 at 5:22am
My assumption is that current central banks would lack the credibility, or ability for that matter, to make something like this work. This is particularly true of any efforts to cut money creation in future years, given the changing makeup of boards.
A computer program could pull it off though.
Simon
Aug 3 2019 at 3:23pm
I’m curious as to how you can square the threshold for central bank balance sheets with “central banks buy as many assets as are required to hit their NGDP target”. (ie what happens when they are about to hit the threshold but are below their NGDP target?)
Is it that for Japan, that threshold is much higher than 100% of GDP, or am I not understanding something.
Ivan Tcholakov
Aug 5 2019 at 4:59pm
According to your sugeestion I can imagine that the Congress is going to decide what the limit of the central bank’s balancesheet should be. I am sceptical that automatic rule of NGDP target would work well, there are situations when the market simply doesn’t demand more money. So, discretionary power is necessary in order not to provoke high inflation. So, let us say Congress sets the limit, the central bank maintains it. Next the Congress will make it higher, and again, and again…
The problem is that politicians need easy money to “bribe” the voters with pleasant promises. The problem is that the Congress can borrow money. Remove this power and then think about the next steps. What about them. For example, stop the war on cash, increase its share. Don’t even think about negative nominal interest rates. Make the fractional reserves of the trade banks higher, so they to be more stable. When the central interest rate is higher you can use the Tailor’s rule, when the interest rule is low – helicopter cash is the most neutral option. When you need to reduce the amount of money in the markets, there are two ways – the government reduces temporarily its spending and increases it own reserve of money, or, the government gives certain amount of tax-collected money to the central bank fo deletion. These technicalities somehow could be arranged fairly, I am sure about that. But I repeat, sine qua non is that your government should finance its activities only through tax-collected money, withot any borrowing, as it was in the 19-th sentury in USA and Great Britain (with the exception of war-times).
Benjamin Cole
Aug 6 2019 at 12:14am
Well, count me in with the money-dropping lulu’s (which include, at times, Ben Bernanke, Adair Turner, Ray Dalio and indirectly, Michael Woodford).
We wear tin-foil ghats proudly!
The simplest way to do a “fair” money drop is this:
Cut payroll (Social Security-Medicare) taxes. Offset lost revenue by the Federal Reserve printing money and injecting it into the Social Security trust fund.
Okay, now people who spend money have more to spend. Not just employees—-remember, employers pay half of such onerous taxes. (Scott Sumner: Since businesses pay SS taxes, is it really a consumption tax?).
But the payroll tax cut, by definition, only goes to productive people or businesses.
Sooner or later, the tin-foil hats will outnumber the orthodox. One more recession, and a few more Martin Feldsteins who enter nirvana….
Mike Sproul
Aug 7 2019 at 12:17am
Backing theory perspective: If the Fed issues 10% more money but gets no new assets, then each dollar will lose 10% of its value. The 10% money injection will leave real cash balances unaffected. Assuming the real money supply was 10% too low to begin with, this money injection will provide no relief from the (assumed) 10% money shortage.
On the other hand, if the Fed issues 10% more money and gets 10% more backing, then the dollar will hold its value, real balances will rise 10%, and the (assumed) money shortage will be relieved; liquidity will be restored, and the economy will get a real stimulus. This is also true if the Treasury raises taxes to repurchase the money.
There’s no reason the injection should be limited to 4%, or to anything based on NGDP. As long as new money is adequately backed by new assets, the new money will relieve any money shortage with no risk of inflation. So what if the new money piles up in bank vaults? The resource cost of that is negligible, especially compared with the the gigantic cost of a recession caused by too little money.
Comments are closed.