We have shown that in the model with capital, the presence of productive assets carrying a positive marginal product does not eliminate the possibility of a secular stagnation. The key assumption is that capital has a strictly positive rate of depreciation. In the absence of depreciation, capital can serve as a perfect storage technology which places a zero bound on the real interest rate. It is straightforward to introduce other type of assets, such as land used for production, and maintain a secular stagnation equilibrium. For these extensions, however, it is important to ensure that the asset cannot operate as a perfect storage technology as this may put a zero bound on the real interest rate.
Let me recapitulate the basic problem. Secular stagnation models are supposed to exhibit persistent negative real rates of return, but how is this compatible with economic growth and positive investment? Just hold onto stuff if need be and of course the goverment can help you do this with safe assets, if need be. The earlier models had no capital, which ruled out this possibility. The new model assumes storage costs for capital are fairly high, or alternatively the depreciation rate for capital is high. Since you can’t sit on your wealth, you might as well invest it at negative real rates of return.
But at the margin, storage costs for goods (and some capital) are not that high. My cupboard is full of beans and cumin seed, but I eat the stuff only slowly. In the meantime it is hardly a burden, nor is it risky since I know it will be tasty once I make the right brew. Art has negative storage costs (for the marginal buyer it is fun to look at), although its risk admittedly makes this a more complicated example. Advances in logistics, and the success of Amazon, show that storage costs are getting lower all the time.
Secular stagnation might be a good model for Liberia and Venezuela and Mad Max, but not for the United States today or other growing economies with forward momentum. But a credible stagnation model for America needs to recognize that rates of return will be lower than usual but not negative in real terms. And there won’t be a long-run shortfall of demand because eventually market prices will adjust so that demand meets the supply we have. That is a supply-side stagnation model of the sort promoted by myself, Robert Gordon, Peter Thiel, Michael Mandel, and others. In the secular stagnation model as it is now being discussed by Keynesian macroeconomists, you end up twisting yourself in knots to force that real rate of return into permanently negative territory. Of course if you allow the real rate of return to be positive albeit low, the economy is not stuck in a perpetual liquidity trap as people move out of cash into investment assets. The demand-side stagnation mechanisms fade away into irrelevance once prices have some time to adjust.
Izabella Kaminska comments here. Josh Hendrickson has a very good blog post on the model here. I’ve already cited Stephen Williamson here, he notes the model is really about a credit friction and would be remedied with a greater supply of safe assets for savings, an easy enough problem to solve, for instance try the Bush tax cuts. Here is Ryan Decker on the model, and here is Ryan arguing that investment is aggregate demand also and many of us seem to have forgotten that, a very good post.
This is an important and interesting paper, but only because it shows the model doesn’t really hold and requires such contortions. The discussion of policy results is premature and way off the mark. The authors should have included sentences like “storage costs aren’t very high, and the economy as a whole does not exhibit negative real rates of return, so these policy conclusions are not actual recommendations.”