Scott Sumner on Negative Interest Rate Policy
I am grateful to Scott Sumner for permission to mirror his post “Miles Kimball on Negative Interest Rates” as a guest post here. This will gives you an idea of what Scott thinks of one of my main emphases. Here is Scott:
David Beckworth did a very interesting podcast with Miles Kimball. You probably know that Miles is an economics professor at Michigan and blogs under the name “Supply Side Liberal” (a label not far from my own views.)
Here are some good points that Miles emphasized:
1. If the Fed had been able to do negative interest back in 2008, the average interest rate over the past 8 years would probably have been higher than what actually occurred. Lower in 2008-09, but then higher ever since, as the economy would have recovered more quickly. He did not mention the eurozone, but it’s a good example of a central bank that raised rates at the wrong time (in 2011) and as a result will end up with much lower rates than the US, on average, for the decade of the “teens”. Frustrated eurozone savers should blame German hawks.
2. He suggested that if the Fed had been able to do negative interest rates back in 2008-09, the financial crisis would have been milder, because part of the financial crisis was caused by the severe recession, which would itself have been much less severe if rates had been cut to negative 4% in 2008.
3. Central banks should not engage in interest rate smoothing. He did not mention this, but one of the worst examples occurred in 2008, when it took 8 months to cut rates from 2% (April 2008) to 0.25% (December 2008.) The Fed needs to be much more aggressive in moving rates when the business cycle is impacted by a dramatic a shock.
Although I suggested negative IOR early in 2009, I was behind the curve on Miles’s broader proposal (coauthor Ruchir Agarwal), which calls for negative interest on all of the monetary base, not just bank deposits at the Fed. To do this, Miles recommends a flexible exchange rate between currency and electronic reserves, with the reserves serving as the medium of account. Currency would gradually depreciate when rates are negative. Initially I was very skeptical because of the confusion caused by currency no longer being the medium of account. I still slightly prefer my own approach, but I now am more positively inclined to Miles’s proposal and view it as better than current Fed policy.
Miles argued that the depreciation of cash against reserves would probably be mild, just a few percentage points. Then when the recession ended and interest rates rose back above zero, cash could gradually appreciate until brought into par with bank reserves. He suggested that the gap would be small enough that many retailers would accept cash at par value. As an analogy, retailers often accept credit cards at par, even though they lose a few percent on the credit card fees.
If cash was still accepted at par, would that mean that it did not earn negative interest, and hence you would not have evaded the zero bound? No, because Miles proposes that the official exchange rate apply to cash transactions at banks. This would prevent anyone from hoarding large quantities of cash as an end run around the negative interest rates on bank deposits. So that’s a pretty ingenious idea, which I had not considered. Still, I think my 2009 reply to Mankiw on negative IOR holds up pretty well, even if I did not go far enough (in retrospect.)
Why is negative interest still not my preferred solution? Because I don’t think the zero bound is quite the problem that Miles assumes it is, which may reflect differing perspectives on macro. Listening to the podcast my sense was that he looked at macro from a more conventional perspective than I do. At the risk of slightly misstating his argument, he sees the key problem during recessions as the failure of interest rates to get low enough to generate the sort of investment needed to equilibrate the jobs market. That’s a bit too Keynesian for me (although he regards his views as somewhat monetarist.)
In my view interest rates are an epiphenomenon. The key problem is not a shortfall of investment, it’s a shortfall of NGDP growth relative to nominal hourly wage growth. I call that my “musical chairs model” although the term ‘model’ may create confusion, as it’s not really a “model” in the sense used by most economists. In my view, the key macro problem is the lack of one market, specifically the lack of a NGDP futures market that is so heavily subsidized that it provides minute by minute forecasts of future expected NGDP. If the Fed would create this sort of futures/prediction market (which it could easily do), then the price of NGDP futures would replace interest rates as the key macro indicator and instrument of monetary policy. Recessions occur when the Fed lets NGDP futures prices fall (or shadow NGDP futures if we lack this market). Since there is no zero bound on NGDP futures prices, we don’t need negative interest rates. However, in place of negative rates the central bank may need to buy an awful lot of assets. You could say there is a zero lower bound on eligible assets not yet bought by the central bank. Which is why we need to set an NGDPLT path high enough so that the central bank doesn’t end up owning the entire economy.
To conclude, although Miles’s negative interest proposal is not my first preference, put me down as someone who regards it as better than current policy.
PS. I was struck by how many areas we have similar views. For instance he thought blogging was really important because what mattered in the long run was not so much the number of publications you have, but whether you’ve been able to influence the younger generation economists (grad students and junior faculty).
PSS. I will gradually catch-up on the podcasts, and then do another post on the 2nd half of the Brookings conference on negative IOR.
You might also be interested in these other posts on my blog that involve Scott: