Wednesday, June 13, 2007

Teacher's Pet

That would not be Jonah Goldberg, whose recent article recommending the end of public schools made Ezra Klein point out earlier that many of Goldberg's preferred alternatives don't do even as well as the public schools in Washington, D.C.. And those public schools are indeed not doing well as Goldberg states:

HERE'S A GOOD question for you: Why have public schools at all?

OK, cue the marching music. We need public schools because blah blah blah and yada yada yada. We could say blah is common culture and yada is the government's interest in promoting the general welfare. Or that children are the future. And a mind is a terrible thing to waste. Because we can't leave any child behind.

The problem with all these bromides is that they leave out the simple fact that one of the surest ways to leave a kid "behind" is to hand him over to the government. Americans want universal education, just as they want universally safe food. But nobody believes that the government should run 90% of the restaurants, farms and supermarkets. Why should it run 90% of the schools — particularly when it gets terrible results?

Consider Washington, home of the nation's most devoted government-lovers and, ironically, the city with arguably the worst public schools in the country. Out of the 100 largest school districts, according to the Washington Post, D.C. ranks third in spending for each pupil — $12,979 — but last in spending on instruction. Fifty-six cents out of every dollar goes to administrators who, it's no secret, do a miserable job administrating, even though D.C. schools have been in a state of "reform" for nearly 40 years.

In a blistering series, the Post has documented how badly the bureaucrats have run public education. More than half of the District of Columbia's kids spend their days in "persistently dangerous" schools, with an average of nine violent incidents a day in a system with 135 schools. "Principals reporting dangerous conditions or urgently needed repairs in their buildings wait, on average, 379 days … for the problems to be fixed," according to the Post. But hey, at least the kids are getting a lousy education. A mere 19 schools managed to get "proficient" scores or better for a majority of students on the district's Comprehensive Assessment Test.

It's possible that bad bureaucrats are causing some of these problems. But Goldberg fails to point out another reason, and that is the poverty level of Washington, D.C.. Poor families are unlikely to send their children to private schools. Poor children come to school saddled with more problems than children from more affluent families. To gather a large number of poor children into one public classroom is not going to create an easy educational experience for the students or the teachers. Comparing the achievement levels of these children to those of children going to private schools fails to standardize for the income differences. It also ignores the fact that private schools can refuse students they don't want to have but public schools have no such luxury. This means that school comparisons of the sort Goldberg wants to use suffer from selection bias. (To give an example of this bias from the field of higher education, Harvard is good at least partly because its incoming students are good. )

Goldberg's piece is an opinion column. Perhaps it's acceptable in that context to pick the worst possible example as a snapshot of how public schools in general are doing. But it's still a little bit odd to summarize the theories explaining why public schools exist as "bla, bla, bla" and "yada, yada, yada" , though of course it makes writing the piece much easier. Likewise, to state that "the simple fact that one of the surest ways to leave a kid "behind" is to hand him over to the government" is quicker to type than any evidence for this argument. It also has the additional bonus of hinting that parents who have a child in a public school are "handing the child over", as if they'd never see the poor mite again. The horror!

And what about the general argument Goldberg makes that schools shouldn't be run by the public sector? He states that the government can't provide services and should provide money instead. A comparison between the Veterans' Administration system of health care (government operated and funded) and the Medicare system (only government funded) might suggest the opposite. But in reality the government provides certain services well and the private markets provide other services well. Which system to use depends not on some ideological statements but on the empirical facts in each particular case, though it's also worthwhile to point out that the countries with the best educational results rely mostly on a publicly funded and operated system, though it might not look like the American public school system.

Time to address Goldberg's "bla, bla, bla" abbreviation for the theories which try to explain why education is so often carried out by the public sector or at least by the not-for-profit sector, rather than by profit-making firms. There are at least four reasons for this.

The first one has to do with the fact that an educated nation provides better living conditions for everybody, not just pride and pleasure to the child and his or her parents. For a thought experiment, imagine a United States in which the majority of the people could not afford an education for their children. What would this do in a generation or so? The country would become a Banana Republic, with the wealthier few living in gated communities and the poor masses trying to get in. Labor would be cheap, true, but without many skills or abilities to acquire them.

In short, the benefits of education fall on a wider group than the students and their parents. But markets can't really enforce payments from that wider community of beneficiaries. Governments can, because they have the power to tax.

This explanation is sometimes used as a justification for publicly funded education. But public funding of education can also be justified by the desire to provide all children with equal opportunities for education. Its absence would mean that the children of the poor start the race several hundred yards behind other children, and might never be able to catch up, however hard they raced.

Now, Goldberg isn't arguing against the public funding of education, only its public provision. One might argue that public provision gives the taxpayers more control over the quality of the education they subsidize. If there is a public interest in the basic education of all children it is pretty unlikely that the government would ever limit its role to just writing blank checks. Some level of control and supervision would be required and the step from that to public provision of the services is a short one.

What makes it even shorter are the characteristics of the "output" of the school system. This output is difficult to measure. It depends not only on the teachers and the teaching tools but also on the child's talents, efforts and family participation, and in most cases years go by before the final effects of schooling are visible.

Whenever the product of some industry (education, health care) has this problem of verifiability (a particular type of informational asymmetry), we find that the industry tends to consist of a large number of not-for-profit firms. Perhaps the reason is that for-profit firms can't attract enough customers when selling products that are hard to verify. Perhaps not-for-profits and the public sector are better suited when the trust consumers have in the product is crucial. Whatever the reason, it is unlikely that a for-profit industry of educational firms would ever take over the schools in this country, never mind that such industries exist in the restaurant, bar and supermarket industries. If Goldberg wanted to offer an alternative to the current public schools a better example might be the U.S. health care industry.

Sigh. I now know why it's more fun to write "yada, yada, yada."
Cross-posted on the TAPPED blog.