Here is a long and very interesting post by Paul Krugman, also referencing a recent talk by Larry Summers. There is also this older Krugman post, and here is Gavyn Davies, and also Ryan Avent. And Scott Sumner. Do read and listen to these, there is much in there to ponder. I do very much agree with the claim that lower rates of return make recovery more difficult and for the longer haul as well. And I am happy to welcome these thinkers, or in the case of Krugman re-welcome, to stagnationist ideas.
I cannot, however, agree with the central arguments about negative real interest rates, and the necessity for negative natural rates of interest (there are a variety of interlocking claims here, so do read them for yourself. I am not sure any brief summary can quite reproduce the arguments, which are also not fully clear).
As I frame the data, we have had negative real rates on government securities, but positive rates on many other investments in the U.S. The difference reflects a very high real risk premium, which of course we would like to lower, and the differences also reflect some degree of investment segmentation. The positive rates on these other investments are evidenced by recent broad stock market gains, observed rates of productivity growth (low but clearly positive), high internal corporate hurdle rates, and so on. The “average vs. marginal” distinction is an important one, but still I don’t see how it can be used to push us away from seeing relevant real rates of return as positive. Nor do I think monopoly is widespread enough for that assumption to be a game-changer. Even Apple competes with Samsung and others in its major product lines.
Given the multiplicity of real rates in the American economy, I get nervous when I read about the real rate or the natural rate. (Don’t forget Sraffa  and also Arnold Kling discusses the different issue of varying rates across people. Interfluidity questions whether the idea of a natural rate makes sense at all.) I also get nervous when I do not see serious talk about the embedded risk premium in the observed structure of market rates. I grow more nervous yet when the average vs. marginal question is not spelled out more explicitly.
In my view very negative real rates of return would not be a “natural rate” giving rise to full employment through a better equilibration of planned savings and investment. Given a pretty flat employment to population ratio, very negative real rates of return across the economy as a whole would have to mean negative economic growth and other attendant difficulties.
And no, I don’t think that output shrinkage associated with the persistently negative real interest rate would be expansionary through liquidity trap mechanisms; for one thing the negative wealth effect and the higher risk premium likely would offset the positive velocity effect on currency balances. The velocity effect on currency balances, from inflation, just isn’t that strong. At persistent negative rates of return we are much more likely to see an interdependence of AS and AD and some kind of cascading collapse of both. Or maybe it is simply better to say the framework has broken down than to try to squeeze one’s own predictions out of that set up.
Furthermore if you think destruction will help you ought then think that capital obsolescence will pull us out of Hansen’s long-term stagnation within five to ten years. On top of all that, I worry about the apparent “out of equilibrium” assumptions embedded in a model that has both a) negative real rates of return on investment and b) those investments being made in the first place, given that storage costs don’t seem to be enormously high.
I don’t mean this in a rude or polemic way, but the arguments we have been reading do not yet make sense.
Here is a claim I do find possible, although it is not one I am pushing. That would be a neo-Wicksellian argument that rates of return on capital are positive but low, and investors need low and indeed very negative borrowing rates to reflate the economy, given how high the risk premium is. I don’t read Krugman as promoting that view (note his citation of Samuelson’s OLG model for instance), although I think that is what the argument will have to boil down to. Otherwise it ends up being a call for output destruction, which, while I do understand how in some models at some margins that can help, I don’t think at current margins is going to be anything other than an unmitigated disaster. Literally.
I see it this way. If you are postulating a stagnation across the longer run, ultimately it will have to boil down to supply side deficiencies. The simple way to explain the mediocre recovery is to tack on slow growth assumptions to the underlying demand deficiencies. But that would constitute a big concession to real business cycle theory and it would put Thiel-Mandel-Gordon-Cowen stagnationist views in the driver’s seat, all the more so over time. The look back to Alvin Hansen is an effort to work in some (very much needed) stagnationist ideas, while at the same time doubling down on a demand-side perspective.
That just isn’t going to work.