I haven’t gotten into twitter (lack of time) but occasionally see twitter threads linked to in blogs. Miles Kimball recently linked to an interesting twitter discussion of electronic money. I’d like to offer a few comments, but first a brief explanation.
Some economists (not me) think the zero lower bound prevents central banks from enacting monetary stimulus. Others think it makes monetary stimulus more difficult, uncertain, or controversial. I think it’s only a problem because central banks foolishly use interest rate targets as both their policy instrument and their signaling device.
One alternative is to eliminate the zero lower bound by paying negative interest on base money. In the 1930s Silvio Gesell proposed doing that through “stamped money.” In 2009 I proposed a more limited plan aimed only at bank reserves (including vault cash.) There are technological problems that make negative interest on currency held by the public much more difficult to implement. On the other hand in a few decades we’ll probably switch our monetary system from currency to electronic money, where negative interest will be easy to implement. No more zero bound. Miles Kimbell has suggested that we speed up that process, and do today what we are likely to at some point anyway. (Although for some reason he contemplates paper currency continuing to exist, albeit de-monetized.)
Here’s the first tweet of interest from Miles Kimball:
It is not only better, but less radical to eliminate the ZLB by e-$. Lift Inflation Expectations to Boost Growth?
Here Kimball suggests that negative interest rates on base money are less controversial than higher inflation expectations. I can think of 4 possible ways of addressing the zero bound problem:
1. Raise the inflation target so that nominal rates never fall to zero.
2. Boost the monetary base as needed to hit the existing inflation target, perhaps buying lots of unconventional assets if the Fed runs out of T-securities to buy.
3. Reduce the demand for base money by a Robert Hall-style regime where the interest rate on base money is reduced to a level expected to produce on-target inflation.
4. Keep the existing 2% inflation target, but shift to level targeting. That might (or might not) allow us to avoid the zero bound. In other words, it’s not clear the extent to which we are at the zero bound because our inflation target is too low, and the extent we are there because the Fed allowed deflation in 2009, and didn’t attempt to make up the lost ground. Even better, shift to NGDPLT.
The supply approach might require the Fed to accumulate a large stock of financial assets, raising concerns about risk/socialism, etc. Or it might not. It’s quite possible that the Fed could hit its existing inflation/employment target without printing more money, if the central bank added credibility through a futures targeting regime. We simply don’t know.
Kimball’s approach reduces base demand with a controversial policy of e-money, which raises issues of “big brother” watching everything you do. However I notice the younger generation doesn’t seem greatly concerned about the loss of privacy coming along with the electronic surveillance state, so maybe this isn’t a big issue.
A possible compromise would pay negative interest on bank reserves (including vault cash) but not currency held by the public. In theory, that doesn’t really “solve the problem” of the zero lower bound, as I envision currency continuing to be a media of account (unlike Miles.) But as a practical matter it would, as vast private currency holdings are fairly costly. If the public had to suddenly absorb an extra $2 trillion in currency (nearly $20,000 per family of three) it would be highly inflationary. Unfortunately even this move is much too controversial for the Fed. They refuse to even go to a zero interest rate on bank reserves, which was standard policy for the first 95 years of the Fed’s existence. QE is less controversial.
Of course I favor option 4, shifting to level targeting (and also shifting from prices to NGDP.) This might not be enough, in which case I’m pretty indifferent between a slightly higher NGDP target (say 5% vs. the current de facto 4%), or a large Fed balance sheet. If the balance sheet got so large that the Fed had to buy assets other than Treasury securities or their equivalent (Treasury-backed MBSs) then I’d favor a negative IOR or a higher NGDP target. But again, I think it is extremely unlikely that this scenario would occur in the US. It’s probably not even worth thinking about.
I believe much of the blogosphere debate is off target. We shouldn’t be looking for new ideas that might “solve problems.” The real problem is that the Fed hopes we end up in a position that is different from where they expect us to end up. If they’d been targeting the forecast since 2008 then none of this would have happened. The zero bound is a symptom of a failed policy, not a constraint that prevents the Fed from enacting a policy that they expect will produce the results that they hope to achieve. For instance, consider these Kimball tweets:
Australia is not a good example. I don’t think Australia was ever near the Zero Lower Bound. . . .
e-money is only directly important when you hit the Zero Lower Bound.
Australia did not hit the zero bound precisely because they followed a policy that (de facto) mimicked the path of NGDPLT. So it’s a great example. If you have sound monetary policy you probably won’t hit the zero bound in the first place, unless you are a slow-growing economy with an ultra-low inflation target (i.e. Japan.) Then the zero bound might even become the norm. But why would a country want to do that?
BTW, I do agree with Kimball that it would be bad idea to raise and lower the inflation target depending on the state of the economy. The beauty of NGDP targeting is that it is actually a target that a central bank would be comfortable sticking with in good times or bad. As we are seeing in Britain, an inflation target gets ignored when it signals the need for the sort of (contractionary) demand-side policies that seem obviously crazy to the policy elite.
PS. Kimball also tweeted the following:
Not much inflation in 30′s. In any case Fed controlled monetary policy, not Roosevelt.
Actually Roosevelt did control monetary policy, using the lever of changes in the price of gold. This forced the Fed to be much more expansionary than they wished to be. Between March 1933 and March 1934 the WPI rose by over 20%, and broader price indices rose by 5% to 10%. That’s pretty impressive inflation in an economy with 25% unemployment and risk-free interest rates stuck at the zero bound.
Today we have a 100% fiat money regime and (supposedly) can’t even manage a measly 2.5% inflation. Talk about technological regress! The interwar generation would have laughed at our incompetence: ”You have unbacked paper money and can’t figure out how to debase it? Are you kidding me?”
The following video shows how I view fiat money central banks that don’t know how to debase their own currencies:
Love the rabbit hunting sequence.