To put a finer point on it than I would like, I think there is a tendency among structuralists and Austrians to complain that Keynesians only look at aggregate data and draw sweeping conclusions, but then the structuralists and Austrians fail to examine the disaggregated data.

I think there are a lot of things going on, but to put it in an Austrian frame, the ability to gather dispersed knowledge has increased many many with the introduction of the computer and the internet. Indeed, this is precisely why they are such a big deal. But, that’s another story.

The point I want to make here, is that we no longer face a tradeoff between simple aggregate measures and thinking about how various parts of the macroeconomy interact.

If the sector level is of interest to you then we have lots of sector data at our finger tips. If that’s not enough the Economic Census lets us dig down to a level of aggregation that is only roughly 3 firms deep.

We can’t get lower than three firms because of privacy concerns – though of course the data does exist and there are probably some circumstances under which it could be accessed.

Right now there are questions about the jobs. Is there such a thing as “the labor market” or are there simply lots of markets in which labor is bought and sold.

I think there is a labor market. Here are a few more pictures that help explain why.

So here are two very different sectors and their job opening rates. Openings as a percentage of employment overtime.

fredgraph

Even though the difference between sectors is at least as big as the difference in sectors, the key is that they move together.

We can also take a much wider view and do all sectors together

FRED Graph

Now layered over top of this is the growth rate real real retail and food service sales. So that is how much stuff consumers are buying.

FRED Graph

There is not a perfect match but I suspect some of that is due to how I am measuring the data. What I really need is year-over-year change in openings compared to year-over-year change is sales.

I can do that for overall opening but it would take longer than I have right now to do it for each individual. I hope from the above we can see that the individuals move together.

FRED Graph

Moreover, openings is a slightly right shifted version of sales. This would seem to indicate that sales moves first. Though, of course, expectations matter.

About these ads