Helicopter money and QE. Avoiding the fuzzy reversibility problem.

One view of QE - encoded in, for example, Eggertson and Woodford's 2003 paper, and similar - is as follows.

It comprises two steps, one effective, one redundant.  Step one is the creation of money to buy on the open market a short-dated government bond.  Step two is the trade of that short dated government bond for a long one.  If interest rates are at zero, but expected to be positive later on, step one may have an effect, but only if it induces, via some appropriate communication, lower interest rates out into the future.  If rates are zero and expected to be so forever, step 1 is redundant.

Step 2 is debt management, and exploits what central banks sometimes refer to as the 'portfolio balance channel'.  Mark Carney called it that today at Treasury Committee.  And he also alluded to the fact that the 'monetary channel' as he termed it was not why QE worked.  We might presume he was thinking along the lines I'm sketching here.

The fuzzy distinction between helicopter money [HM] and QE comes about due to using Step 1.  But central banks can, and did sometimes, skip step 1 and simply swap short for long dated gilts.  The Fed did this explicitly in one round of QE, calling it 'Operation Twist' after a similarly named episode in the 1960s.

If central banks had always skipped the redundant Step 1 [you can always lower future rates using forward guidance], there would never have been any doubt that this was QE rather than helicopter money.  Indeed, it would have been better not to use the term QE at all, but stick to 'debt management'.  I don't recall anyone ever screaming at debt managers that they were doing helicopter money.

Doing it this way would also get round what one person who commented privately on my blog pointed out is a bit inconvenient for central banks.  Central banks are already talking [see Jackson Hole, for example, or also words by Mark Carney and Ben Broadbent] about the fact that the balance sheet size might not shrink back to what it was before the crisis.  This might not look good next to the promise that QE would be reversed, and the religious fervour with which they have said HM should be avoided.

If the portfolio balance channel had simply been exploited with debt management, there would have been no such fuzziness to worry about in contemplating a larger central bank balance sheet after the crisis.

However, you ought to be saying 'not so fast!'  Suppose we consolidated debt management inside the central bank [it was in the UK until 1997].  And then allowed the central bank to do the issuing of the short debt which was used to get the funds to buy the long debt.  In the standard monetary version of the New Keynesian model, this is satisfyingly distinct from helicopter money.  Short term debt is not money.  But in the real world, things might not be so clear-cut.

But if there was a chance that the real world was like the model world, you might ask - why if the Bank of England, for example, thought that Step 1 [create reserves to buy short debt] was redundant, [what Mark Carney today implied] did they do it at all?  In his defence, he was inheriting a precedent set by the MPC under Mervyn King.  In early communications about QE, he and his MPC took a different view of the monetary channel.  Although subsequently the emphasis was slowly dropped.  And implementing 'debt management' in an orderly way, and on a sufficient scale, might have been impossible without this being done by the Debt Management Office.

[PS voice recognition software gets better the more you use it.  RSI sufferers take heart:  Even if you can't do real work, you can still play.]