Clarification of ‘Helicopter’ Money Column
I recently wrote a column on helicopter money for BloombergView that got a lot of feedback. My goal in this post is to clarify the arguments that I made there.
My column considered a $100 billion increase in government spending. With any increase in government spending, there are two distinct issues:
Many readers were passionate that the government would achieve desirable outcomes (defined in a variety of ways) if the answer to (1) was, “make transfers to low-income households.”. I’m happy to take this claim as given in what follows - that was not the main point of my column.
Instead, my column mainly argued that two different ways of financing the $100 billion increase in spending:
were equivalent in terms of their budgetary implications. Let’s add a third possibility (which is completely illegal in the US):
iii. the Fed prints $100 billion and hands it out to low-income households.
I’ll argue that this third option is also equivalent.
Here’s the argument. Let me start, as I did in my column, with the debt in financing option 1 taking the form of floating coupon consols. A consol is a bond with a face value of $1 never matures. It pays an annual coupon equal to the interest rate on excess reserves paid to banks during that year. (Notice that the size of this coupon is completely under the control of the Fed.)
The question is: how much money can the Treasury raise by selling such consol bonds to banks? It’s easy to see that a consol bond with a face value of $1 can’t sell for more than $1. Otherwise, why would a bank (who can earn interest on excess reserves) buy such a bond?
What’s more subtle is the other direction: can the bond sell for less than $1? Here, I made an implicit assumption which clearly struck some readers as strong: multiple banks who currently have excess reserves believe that they will never face a binding reserve requirement in the future. If that’s true, these banks will all be willing exchange their reserves for the consol if it costs less than $1. Their competition will bid up the consol’s price to $1.
These two arbitrage arguments mean that the Treasury can raise $100 billion by promising to pay interest on reserves (whatever the Fed chooses it to be in the future) on that $100 billion each year.
At this point, I hope that I’ve convinced you that, under current conditions, money is not a cheaper form of financing for the government than bonds. It is then easy to show that the three financing options are the same.
First, recall that I’m assuming that Treasury transfers the $100 billion to low-income households. Under all three financing options, they will put that transfer into banks (assuming interest is not too negative). Bank reserves rise by $100 billion.
Under financing option 1, banks reduce their reserves with the Fed by $100 billion in order to buy the debt issue. With the increase in $100 billion in reserves noted above (that comes from the transfer to low income households), bank reserves are unaffected by the new government spending initiative. However, the banks now own the Treasury debt issue. As a result, the Treasury owes banks an annual payment equal to $100 billion times the interest on reserves in that year.
Under financing options 2 and 3, the Treasury need not make any payments to anyone. However, the transfer from the Treasury to individuals means that bank reserves rise by $100 billion. The Fed owes banks an extra annual payment equal to $100 billion times the interest on reserves in that year. It will simply pass that cost on to Treasury by reducing its remittances by the same amount.
The three financing options are exactly identical in terms of their resource implications. Under all three, the transfer recipients have $100 billion more. Under all three, each year, the Treasury is essentially paying the banking sector $100 billion times the interest on reserves (whatever the Fed chooses that to be).
In the real world, of course, Treasury doesn’t use floating rate consols. Financing option 1 is about using fixed coupon (Treasury) debt. Financing options 2 and 3 are about using floating coupon debt (bank reserves) I don’t believe that swapping between the two kinds of debt would have large macroeconomic consequences. More importantly, I don’t believe it’s what the fans of helicopter money have in mind.
Basically, money and bonds are equivalent forms of finance in a world in which banks perceive themselves to be flush with liquidity for any time horizon of interest. We generally think money is a cheaper form of government finance because the government is the monopoly provider of monetary liquidity, and the government keeps the economy liquidity-starved. THat’s not the world we live in now and I doubt that we’ll return there soon.
But, more neutrally, if you think money is a cheaper form of finance, it’s because you think that banks assign some probability to the possibility that they will face a binding reserve requirement in the future. That probability, appropriately discounted, would make the second form of finance somewhat cheaper for the government. And it’s that wedge that makes helicopter money, as opposed to bond finance, attractive.
N. Kocherlakota
Rochester, NY, March 25, 2016
Please address media enquiries and non-academic speaking requests to Monique Patenaude (monique.patenaude@rochester.edu and 585-276-3693).