Supply Side Recession Stories

Having started the final external editing pass of my manuscript on recessions, I am now confronted with a possibility I dreaded: having the text overtaken by events. I split my text into two parts, and the first part mainly covered post-Keynesian theory, which is decidedly focused on demand-side explanations of recessions. Meanwhile, the coronavirus provides us with an example of a recession driven by the supply side...


To be fair to myself, I had some text that did cover such possibilities, and I just added text that noted that the pandemic was developing at the end of writing. However, it is somewhat awkward that I have a book coming about the causes of recessions, and very little coverage of the causes of a recession that could possibly hit almost simultaneously with the publication of the book.

(Note: I am taking a detached view of the effects of the coronavirus spread herein. I have a natural pessimistic tendency, and so I could very easily run off with scare stories. There are already other people panicking, so I am not adding much value if I hop onto that train as well.)

From a theoretical point of view, the slowdown in China (the magnitude of which is yet unknown) appears somewhat similar to the story of a Real Business Cycle (RBC) model: growth just shuts down because of a "random shock" that hits production ("the supply side of the economy"). And since most dynamic stochastic general equilibrium (DSGE) models have a RBC core, they behave similarly. The reason why I did not write much about such a possibility is that I pushed RBC models to the second volume, which is only partly written. To what extent the coronavirus causes a recession, the natural place to discuss it is in that context.

The problem with the RBC story is straightforward: the mechanism in the model is a drop in productivity. That is, less output from the same number of hours worked. However, this is not what has happened in China: output has fallen because an external factor has put a hard constraint on hours worked (workers are under quarantine). There is absolutely no way to pretend that the drop in working hours was the result of optimising decisions of workers with regards to their lifetime consumption plans: they are either not being allowed to work, or they wish to skip work because of fears of being infected.

The effects outside of China very much depend upon the growth rate of infections -- a topic in which I have no expertise. However, even if we are optimistic about the spread of the virus, the possibility exists of a disruption to production courtesy of the lack of required inputs, given the central importance of China in global supply chains. This mechanism could be thought of as a drop in productivity. Nevertheless, we are confronted with the reality that any aggregated macro model offers almost no useful tools for analysing this scenario, nor any useful forecasting capabilities. (Update: the previous sentence caused some consternation online. I tweaked it to be more clear, although it is still equivalent to my original wording. There is a big leap from discussing RBC models to dismissing all aggregated models, and it was not covered here. Since I just wrote a 58,000 word manuscript where variations of this text appeared multiple times, it seemed obvious to me when I wrote it.)

The other thing to keep in mind is that the raw effects of the virus are only providing the initial impetus for reduced production. Unless the death toll is catastrophic, production would be expected to bounce back.  The concern is that any slowdown would trigger further problems. For example, a reduction in risk-taking could lead to to the failure of weak borrowers, causing ripple effects. At that point, we are returning to the demand-side recession narrative.

(c) Brian Romanchuk 2020