Schools and Competition

Daniel Mitchell at Cato says school choice “is better than government-imposed monopolies” and also that “[t]he evidence about the school-choice systems in Sweden, Chile, and the Netherlands is particularly impressive.”

I think the buyer needs to beware when he hears libertarian touting school choice concepts. Choice can add a lot of value to education, or it can be destructive. The details actually matter a great deal. Bentley MacLeod and Miguel Urquiola did a paper, “Anti-Lemons: School Reputation and Educational Quality” which sheds important light on this issue:

Friedman (1962) argued that a free market in which schools compete based upon their reputation would lead to an efficient supply of educational services. This paper explores this issue by building a tractable model in which rational individuals go to school and accumulate skill valued in a perfectly competitive labor market. To this it adds one ingredient: school reputation in the spirit of Holmstrom (1982). The first result is that if schools cannot select students based upon their ability, then a free market is indeed efficient and encourages entry by high productivity schools. However, if schools are allowed to select on ability, then competition leads to stratification by parental income, increased transmission of income inequality, and reduced student effort—in some cases lowering the accumulation of skill. The model accounts for several (sometimes puzzling) findings in the educational literature, and implies that national standardized testing can play a key role in enhancing learning.

Kevin Carey did a good post last week, for example, about Trinity Washington University here in DC. Trinity’s not “the best” school in America, or even “the best” school in the DC area. But unlike Georgetown or George Washington University or UVA, what Trinity is doing is offering an education to mostly poor, mostly black graduates of the mostly bad DC public school system. They’re doing so at an affordable price and they’re doing it effectively. But as Carey explains and as MacLeod & Urquiola predict, the perverse incentives of college competition in the United States militate against other schools imitating Trinity:

It’s also worth noting the Trinity is consistently ranked as a “4th Tier” Master’s-granting university in the north region by U.S. News & World Report. This is the worst possible ranking and classification a college can receive. Trinity receives this ranking not despite the fact that it serves under-prepared students on a modest budget, but because it serves under-prepared students on a modest budget. If Trinity began jacking up tuition and telling African-American women from the impoverished Anacostia region of Washington, DC to find somewhere else to enroll in college, it would rise in the rankings. In this way, the incentives built into the U.S. News system run 180 degrees from the values animating colleges that do more than pay lip service to their public obligations.

By contrast, the way K-12 charter schools work in the United States is that they have to take all comers or else exclude students via lottery. Under the circumstances, the best way to improve their reputation is to improve their students’ results. That’s the virtuous kind of competition. And while the Swedish K-12 system doesn’t work exactly like US charters (main difference being that for-profit school operators are common) as I’ve pointed out before, Swedish schools are under a charter-like regulatory mandate to be unselective in their admissions and it’s more-or-less the same in the Netherlands.

Swedish for "hair care" (my photo, available under cc license)

In other words, there really is decent evidence from these countries that more school competition can improve outcomes, but the terms on which the competition is organized makes a huge amount of difference.