Zero Lower Bounds

These are the usual confused thoughts mostly stimulated by this excellent post by Simon Wren-Lewis

You really should click the link and read that post. I will try to summarise it.

First professor Wren-Lewis argues that monetary policy would not have offset less contractionary fiscal policy in the past 6 years, because short term safe nominal interest rates were stuck at the zero lower bound. Second, he argues that “For whatever reason (resistance to nominal wage cuts being the most obvious), inflation ceases to be a good indicator of underutilised resources when inflation starts off low and we have a major negative demand shock. ” So the not at all absolute zero lower bound on changes in nominal wages matters too.

The problem with the argument that nominal interest rates can’t be negative is that they are in Denmark and Switzerland. The logic was that interest rates on bank notes can’t be negative so negative interest rates on other assets can only equal the cost of holding wealth as cash in a safe. It turns out that these costs make negative interest rates possible. I don’t think this matters all that much. Cash in a box arbitrage still implies that there is a lower bound which is no higher than -1% per year (and might be lower but it exists). It also does not imply that looser fiscal policy would have been offset. If central bankers believe there is a zero lower bound and they are stuck at it(as the Federal Reserve board and the Bank of England did) will not raise interest rates above zero following any increase in aggregate demand. I don’t think recent market rates in Switzerland and Denmark refute the whole liquidity trap argument, but I do think that Krugman, Wren-Lewis and others (including your humble correspondent) have to consider the issue a lot more.

There is another, much less valid alleged zero lower bound. It is argued by, for example Jeremy Stein that safe interest rates of at least 2% are needed for banks to make profits without taking risky gambles. It is assumed that such gambles can’t be prevented by prudential regulation, so we have to pay them 2% protection money or else something might happen to the nice economy we have here. But the argument is that the cost of maturity transformation and such is 2% per year and it is just assumed that banks can’t (or won’t) charge depositers. That the worst deal a bank can offer me is the convenience of a checking account for zero interst. This is nonsense. I can’t get a deal that good from either of my two banks (including SunTrust which promised me exactly that deal, but has charged me fees). Banks can charge for the convenience. Italian banks have always done this. The zero lower bound on checking accounts is like the zero lower bound on nominal wage changes. It isn’t a no arbitrage condition, it is a norm. I can agree that bankers won’t drive off clients by charging fees because providing zero interest zero fee checking is temporarily unprofitable. But if it is agreed that they are willing to accept a little red ink, why is it assumed that they can’t be kept from making reckless gambles to avoid it ?

Finally the “resistance to nominal wage cuts”. This is very important. It is clearly not absolute. In US data there is a mass of wage changes of exactly zero, but there are also wage reductions. During the great depression, there were sharp reductions of nominal wages.
I think it is reasonable to assume that there are two key unemployment rates. The NAIRU and a higher rate needed before nominal wages are cut. Given differences in different local labor markets, downward nominal wage rigidity matters even when average wages are growing. The (not absolute) ZLB can bind for some occupations in some places even if average wages are increasing. Together these imply a region with a downward sloping long run Phillips curve. I am trying to summarize Akerlof, Dickens and Perry 1996 (warning pdf).

This would imply that a very low inflation target might cause permanently high unemployment (not high enough to break the downward nominal wage rigidity). It is clear that 2% inflation is consistent with low unemployment. It possible that a monetary authority which accepts inflation anywhere from 0% to 2% allows a permanently high unemployment.