Bryan weighs in on an issue of seasonal employment. This issue has produced much controversy among a few wonks. See the post by Mark Thoma.
Almost 40 years ago, I was a seasonal teenage worker, in a factory in South St. Louis County. As a result, the labor force on the payroll increased. However, the labor force in the factory did not. Summer was when the regular workers were more likely to take vacations. In some sense, I replaced a different worker every week.
To the extent that my experience was normal for teenagers, I do not think it represents an argument either for or against Keynesian economics. Neither aggregate demand nor aggregate supply were involved in having a teenager make up for the bulge in vacationing by other workers during the summer.
In Ed Leamer's Macroeconomic Patterns and Stories, his chapter on GDP has an appendix called "The Seasonal in GDP is Very Large." GDP grows very well in the fourth quarter (the Christmas effect), collapses in the first quarter, grows very well again in the second quarter, and then grows modestly in the third quarter.
Do these large seasonal factors tell us something important about aggregate demand? Aggregate supply? The welfare effects of short-term fluctuations in GDP? I don't think so, but I confess that I have not thought about it terribly hard.
In any event, this seasonality has become less pronounced over the years. I suspect that as we have become a Garett Jones economy, we have moved some of the Christmas seasonality offshore. This leads me to another point that Leamer makes, in an appendix to the following chapter. He thinks that it may be misleading to think of net exports as a "sector" of the economy, rather than looking at consumption net of imported consumer goods, investment net of imported capital goods, and government spending net of imported goods and services used by government. Leamer writes,
Suppose for the sake of argument that US manufacturers make machinery to be sold to other businesses, but make no consumer items at all. (This is not so far off, by the way.) Then all of the consumer spending C would come from imports M. In that case, changes in the spending of consumers C would have not effect at all on US GDP...you might conclude that tax reductions that stimulate consumer spending C would increase US GDP when all they do is to increase the demand for foreign goods M, keeping C-M constant.
The parenthetical remark is Leamer's, not mine.
I would add a cautionary note. A lot of consumer spending is on services, which are unlikely to be imported. Still, he makes a fair point.