Returns to Schooling in the U.S.: 1940 to 2000

Two weeks ago, I played MC at a day long conference at the Federal Reserve Bank of Minneapolis. Among the talks was a look at returns to schooling by Lance Lochner, economist at the University of Western Ontario.

The returns to college for those that got their B.A. was around a real (inflation-adjusted) 20% for white men, up from about 15% in 1980. A college sheepskin still beats the real return of 7% on stocks by a considerable margin. The returns on compelting a college education have come down a bit since 1990 for African-American men, from around the 22% level to 20%.

But what really surprised me was the return on competing high school--slightly above 50%. Yes, you read that right. That's a real return. Now, most economists are suspicious of any return figure that high. But Lochner made a convincing case that the challenges so far have managed at most to shave a few percentage points off. For African-American men the figure is slightly higher, in the 56% range.

Here's the puzzle. The long run returns on a college degree rose modestly, but there has been a continuous and sizeable rise in college attendance. In essence, the market worked. Yet when you look at high school completion rates, they are down slightly since 1970s while the rate of return on graduating went up (eyeballing his chart, from about a 30% rate of return since the early '70s to around 50%).

So, according to Lochner, "returns to high school are substantially higher than college and have increased more rapidly in recent decades. High school graduation rates actually declined in recent decades."

What's going on here?