If you look at rankings, we were 18th from the top in per-pupil spending in academic 1975; by our nadir (1995), we were 41st. Translated into average class size, we were 49th—Utah managed to get under us. And the performance of California schoolchildren on a series of standardized tests also declined, all pretty much in step with this decline in public expenditures.

But there's another way to look at the level of state and local support for education. The graph above shows spending per pupil in terms of the available tax base, i.e., per thousands of dollars of real personal income per capita. Here you see a different pattern that has a lot more continuity. In fact, we never, during this entire period, spent as much per pupil per available resource as the rest of the country. What's going on here? This curve is nearly flat, so even if we've always been a little chintzy compared to the rest of the country, why did spending per pupil in the previous graph suddenly take a tumble?

One possibility might be that we just don't tax ourselves very hard in general. That's not true, it turns out. It hasn't been, and still isn't, throughout this 30-year period. If you sum up all the expenditures made by state and local governments (below), you'll see that before Proposition 13 we were spending a little more than the rest of the country, and after Proposition 13 we spent less. Proposition 13 hammered us for a year or two, but we had a five-billion-dollar surplus at the time that cushioned the blow, which was one reason why people voted yes on 13 in the first place. They felt that Jerry Brown was sitting on their money. We've been catching up with the other states since then, and now we're about even. So our decline in per-pupil spending is not the result of not taxing ourselves enough, or of not spending enough public money in general

Another possibility came to me one day when there was a big public event at Beckman Auditorium—I think it was a puppet show—and hundreds of school buses converged on the campus. Caltech was just taken over by six- to nine-year-olds, and I thought, "My God, there's a lot of kids in California!" And so I thought, well, maybe we can't spend as much per pupil because we have so many more of them. Maybe the state has a very young age structure, and there are just a lot more public-school children in California as a percentage of the population. But if you can see the difference between the two lines in the plot above, your eyes are pretty good, because there isn't much. With certain exceptions, like Florida, it turns out that there aren't major differences in the age structures of America's states. All right then, I wondered, does California send more of its students to private schools? No. It turns out we don't.