Pro DeLong Contra Say

Robert Waldmann

In this post below , I note that I find Brad DeLong’s analogy between the bubble, the housing bubble and the stimulus plan unconvincing. The post is poorly written, but the point is that in fresh water models rapid technological progress is very different from a demand shock, and, so, if one added irrationality, the irrational perception of rapid technological progress is different from a demand shock.

In comments to that post, Jack worries that economists seem to be able to make equally plausible arguments for every conclusion (and so none must be very plausible). I am not able to reassure him. In particular I will now make an equally plausible argument for the position that we know for sure that demand shocks cause increased production.

I’m pretty sure this argument has been made. In any case the name to google (scholar) is Jeff Miron (whose name is on the Cato open letter against the stimulus).I can make the argument with one word


Exposition after the jump.

Every year there is a shift in consumer demand. Demand is high during the dread Christmas shopping season. In particular demand for big ticket consumer durables is very high in each and every 4th quarter compared to the other 3. I argue that this is a predictable demand shift analogous to the stimulus.

So what happens to production ? Hmm somehow macroeconomists have managed to forget. Macro economists generally use deseasonalized data. Seasonal patterns are not explained in most models and so they are removed by a-theoretic methods.
Thus the models aren’t taken seriously (that is this gets on Thomas Sargent’s nerves). Worse, much worse, data are deseasonalized by government statistical agencies, so academic economists can avoid knowing what exactly was done.

I can try to guess what should happen during the Christmas shopping season and confront my guesses with reality. I promise you, my guesses are reliably wrong.

First, I guessed that high consumption crowds out inventory investment. In plain(er) English, firms make goods during the non rush seasons and store them for the Christmas rush. I was *wrong*. In fact inventories track sales. As the Christmas season gets going sales and inventories increase. Inventory investment implies that the variation of production is greater than the variation of sales (this is true of the business cycle as well as of the seasonal cycle).

Second, I’d guess that firms charge high prices (then very low prices in post Christmas sales). This means that real wages would be low during the season and firms would be glad to hire more workers and increase output (this is a very New Keynesian type explanation). This too is false. Retail prices do not rise during the Christmas rush (in fact if anything they decline).

Hmm how about a labor supply shift ? No people are desperately busy shopping. If anything, willingness to do anything else (like work overtime) is reduced.

So what the hell is going on ?

Flailing helplessly, I guess that there is typically excess supply of labor and goods and that prices are not marginal cost times a markup but rather average cost times a markup and that production is at a flat point of the average cost curve (this last should be true if there is monopolistic competition).

Imperfect competition in product markets is absolutely not enough to explain the pattern. Imperfect competition and small costs of changing prices certainly isn’t (they change prices just not in the direction that they should).

In any case, high consumer demand during the Christmas season doesn’t seem to crowd anything out. I’d say that unless and until someone either shows it does crowd soemthing out or explains why the stimulus is different people should stop arguing that government spending crowds out private spending one for one.