Nominal GDP Targeting Through Unconventional Monetary And Fiscal Policy

by: Brad DeLong

Duncan Black complains because he thinks I have not been critical enough of nominal GDP targeting via unconventional monetary policy alone:

Eschaton: Why Don't They Lend Me $30 Billion On The Security Of My Cats?: If we're going to actually move to more "unconventional" monetary policy, can we please recognize that the reason to do so is largely because conventional monetary policy - acting through the banking system - isn't working? We should understand that it isn't working because it almost destroyed the world a few years ago and is about to do so again because, you know, nothing changed and the overpaid assholes who almost destroyed the world then are still in charge. If we're going to give out dodgy loans, how about giving dodgy loans to people who might do something with the money other than visiting the Great Casino?

Point taken.


I will report to the reeducation camp tomorrow, and if Duncan will send me pictures of his cats I will draw up plans for the DBCFRLF alongside the JDDFRLF…[1]

I have been saying that coordinated fiscal and monetary policy--jen-U-ine helicopter drops or simple government-print-and-buy-useful-stuff--is the superior way to accomplish nominal GDP targeting, and that doing so via monetary policy alone runs risks.

But I have not been saying so loudly enough.

Look: targeting the nominal GDP path via monetary policy alone in a liquidity trap is a bet that private-sector financiers will:

  • be confident that the policy will not be reversed when the economy emerges from its liquidity trap,

  • be confident that the policy will succeed and that they should start spending now in anticipation of the faster nominal GDP growth that the policy will produce, plus

  • a little bit of taking risk onto the Federal Reserve's balance sheet and so freeing up private financier risk-bearing capacity to expand their loan portfolio.

Mostly, that is, the policy is a policy that succeeds if it is generally expected to succeed and fails if it is generally expected to fail. It thus has the confidence fairy nature.

To the extent that the policy does not have the confidence fairy nature, it is because it changes asset supplies here and now and thus private financiers' incentives to lend and businesses' incentives to produce. It does so because the policy involves swapping one asset for another asset that is not the same.

Right now because we are in a liquidity trap short-term Treasury bills and cash are effectively, for the moment, the same asset: they are both short-term zero-yield safe nominal government liabilities. Very few believe that the Federal Reserve's buying Treasury bills for cash and saying: "See! We are doing something! Nominal GDP growth will be faster! You should raise your expectations of real growth and inflation and act accordingly!" would actually do anything. By contrast, if the Federal Reserve buys long-term Treasury or agency or private debt the assets it is buying carry an expectational term premium, duration risk, and (perhaps) default risk: they are not identical to the assets that they are selling. Because the private sector's asset holdings change, private-sector financiers and businesses have incentives to change their behavior even if they don't buy the appearance of the confidence fairy at all--and the fact that they will change their behavior even if they don't believe is a reason for people to believe.

The superiority of unconventional monetary policy thus works off of the fact that the assets the government is buying are different than the assets it is selling--and thus the more different the assets it buys from the assets it sells, the greater the non-confidence-fairy bang from the policy.

What asset is most different from cash?

With a helicopter drop, the Federal Reserve sells cash and it buys… nothing at all. Cash and nothing are pretty different assets. Add cash to private-sector portfolios and take nothing away, and portfolios have shifted in meaningful ways and people will change what they do.

With print-money-and-buy-useful-things, the government sells cash and buys… roads, bridges, research into public health, flu shots, killer robots--all kinds of things that are very very different indeed from cash.

Thus, as Milton Friedman's teacher Jacob Viner knew well back in 1933, coordinated fiscal and monetary expansion via printing money and buying useful stuff (or handing it out via helicopter drops) is a policy that really does not have the confidence fairy nature. Because it does not require confidence to start working, it will (probably) work much more rapidly and certainly.

[1] DBCFRLF: The Duncan Black's Cats' Federal Reserve Loan Facility. JDDFRLF: The Jamie Dimon's Dog Federal Reserve Loan Facility.