Money, inflation and the zero bound. Krugman-Evans-Pritchard revisited.

Paul Krugman and Ambrose Evans-Pritchard have been jousting on the question of whether the central bank can create inflation at the zero bound or not.  I don’t think either of them have teased out all the subtleties, and, partly as a consequence, are talking somewhat at cross-purposes.  In particular, we need to clarify whether we are talking about a permanent or a temporary zero bound episode;  and whether we are talking about conventional open market operations or a helicopter drop.  And we also need to clarify whether we are conjecturing what happens in the real world, or just talking through the properties of a model.

Let’s talk about what happens in a standard, rational expectations, flex or sticky price macro model with standard ways of getting money into the models.

Do conventional open market operations that create more money have any effect at the zero lower bound if the ZLB incident is temporary?

No, not unless there is a corresponding expectation that at some future point in time, after the ZLB ceases to bind, that interest rates would be sufficiently lower than they would normally be, given the announced schedule for interest rates, ie sufficient to absorb the extra money and preserve money market equilibrium.  Without that expectation, the operation that created the money would have to be reversed later on.  Moreover, the creation of extra money at the ZLB is not even needed in order to lower expected future rates.  In the model, all that’s needed is to promise to lower them credibly.  At the point that they are lower, money market equilibrium will induce the central bank to create the extra money then through open market ops.

If the ZLB episode was permanent, then such operations would have no effect whatsoever.

What about helicopter drops?  In the standard framework for modelling money, both money and bonds are treated as economic liabilities of the public sector.  So helicopter drops at the ZLB don’t generate inflation, because they won’t be treated as wealth.  The typical person in the model simply looks forward to the time when the operation will have to be reversed, or made good with taxes.  And, at the ZLB, any non-pecuniary liquidity or transactional benefits from holding money are exhausted [that’s why in these models interest rates are zero].   In slightly more realistic versions of this model, where people face credit constraints, and money now may be useful despite there being higher taxes later to make up for it, the helicopter drop will work.  But in those situations so will tax reductions.  If institutional continuity is valuable, [though nothing in the model says this], such drops are are to be considered inferior to tax reductions since they achieve the same thing but at cost of institutional change.

Recently, Buiter has been arguing that we should model money as net wealth.  As conferring transactional/liquidity services (away from the zero bound) but as not redeemable like government bonds.  In such a world, helicoptering in money does confer wealth, and would generate spending and inflation, in which combination depending on what assumption was made about the stickiness of prices.  However, as I argued in previous posts here and here, while on realism grounds there may be a case for thinking about the irredeemability of money [as the new monetarists like Wright, Wallace and their collaborators would urge], as a modification of the standard model, which preserves the assumption that money simply confers liquidity and transactional benefits, it’s problematic.  Because it begs the question where those non-pecuniary benefits come from.  [One crude answer being:  from the authorities treating money as a liability].  I am left being prepared to concede that helicopter drops might generate inflation, but not that they would be beneficial, since they may lead to the steady destruction of that money altogether with all the attendant costs.

So, in some model worlds, helicopter drops work, and in others they don’t.  But it’s contentious whether they would ever be desirable.

Of course, so far we’ve only thrashed out what happens in models.  We can’t be sure what would happen in the real world.  Perhaps this caveat is implicit in the confident sounding exchange between Krugman and Evans-Pritchard.  But, even if it is, it’s worth repeating.  The monetary models of OMOs and helicopter drops we have are very rudimentary, give conflicting answers, and may be misleading on account of other assumptions made in their construction.

This leaves me here:  in the most plausible theoretical models, helicopter drops either don’t work, or work no better than conventional fiscal policy.  As a practical matter, they are to be feared as they risk being associated with monetary and fiscal disorder;  and are anyway unnecessary for the US or UK or EZ where there is ample fiscal capacity to do more of the usual kind of stimulus.  Such policies may be infeasible because of politics.  But helicopter drops are hardly likely to succeed on those grounds either.  Imagine the reaction of the Republican controlled Congress or of the Germans to a helicopter drop.