More than meets the eye

From Has the U.S. Finance Industry Become Less Efficient? (PDF, 28 pages) by Thomas Philippon:

One important point is that the model does not predict an income effect, i.e., just because a country becomes richer does not mean that it should spend a higher fraction of its income on financial services. Of course, income per capita is not a stationary variable, while the share of finance is bounded by definition. We can nonetheless look at long run changes in income and ask whether the share of finance tends to increase more when income growth is faster. Figure 4 shows that there is no mechanical relationship between the growth of income per capita and the share of finance. This rules out any theory that predicts that the finance share should mechanically increase with income.

And the Figure 4:

Source: Thomas Philippon

I'm gonna want to say something about this graph, six months or a year from now maybe. Meanwhile I'm not sure what I want to say about it. Except this. Follow along with me from the end of the Roaring '20s and the start of the Great Depression:

Up at the very top of the graph, higher than anything else, we have 1930.

Drop straight down to the bottom, and a tad to the right, and we have 1940.

Go all the way over to the right and a tad up, and we have 1950.

Jump to the center of the graph and find 1960.

A little bit right from there, 1970.

Back to the center, 1980.

Up, to 1990.

Up again, to 2000. And we are back almost to where we started.

Since the 1970s, the trend has been up and to the left, toward the place we were at in the 1930s. And they say nobody saw it coming.