Nick Gillespie drew attention yesterday to an op-ed Charles Murray wrote on school choice. Murray’s thesis was that the dominance of family environment and genetics in determining student achievement is such as to allow little room for schools to affect academic outcomes. That said, Murray goes on to argue for school choice anyway, on the grounds that families differ in their educational preferences, and the best way to match families to schools is to allow the former to choose the latter. This, he says, “should be the beginning and the end of the argument for school choice.”

Certainly Murray’s point about the value of choice is true, so far as it goes. But it doesn’t go nearly far enough. First, there are other compelling non-academic arguments for school choice (e.g., they minimize social conflict by allowing families to get the sort of education they want for their own kids without imposing it on everybody elses, as happens of necessity when there is a single official government organ of education.) Second, there is very good reason to believe that true market education would lead to higher student achievement.

Murray cites the pathbreaking work of James Coleman, who revealed that home-related factors explain more of the observed variation in student achievement than does choice of school, to argue that schools can’t have much effect on achievement. This is a non sequitur. While Murray’s inference is consistent with Coleman’s evidence, it does not necessarily follow from it. It is possible that since 90 percent of U.S. students are enrolled in government monopoly schools, and since those schools operate on similar lines not just within states but between states, variation in schools’ contributions to student learning have been artificially curtailed.

Furthermore, even in a highly competitive and free education marketplace, variation in student achievement between schools wouldn’t necessarily be very large, since the very best schools would be emulated by many of their competitors, and the very worst schools would go out of business.

But, and this is the point that Murray did not address, the mean level of student achievement in the competitive marketplace could well be much higher than the mean level in our current monopoly system despite the fact that, within each of these systems, school-to-school variation might be low.

From our previous exchanges on this topic, it seems Murray is skeptical of that possibility–skeptical that markets could lead to a substantial increase in mean academic achievement above the mean we observe the existing school monopoly. I offered some counter-evidence in those earlier exchanges, but here’s another reason to expect a higher market mean: the abandonment of rigid age-based grading.

Age-based grading is arbitrary and pedagogically counterproductive. Market education systems, such as the Asian after-school tutoring industry, often group students based on their performance in each subject, promoting them to the next level of the curriculum as soon as they have mastered the current one. This allows students to progress at their own pace. Empirical studies have shown that performance based grouping helps students at all levels of performance to learn more quickly, and it would almost certainly contribute to a boost in the overall average even if nothing else changed.

But the existing research on performance based grouping fails, I think, to capture the full extent of the difference that it can make when allowed to operate unfettered. It is not at all uncommon to see kids who take an avid interest in some pursuit leap well ahead of the typical adult expectations of what they can achieve. Music is a good example, as are strategy games like chess and go (a.k.a. weiqi or baduk). There are 11 and 12 year olds who play these games well above the level that most adult players ever reach.

Furthermore, it does not seem that we can fully attribute the stellar achievement of these youngsters to stratospheric IQs. The research on chess and IQ is surprisingly sparse, but what there is does not point to a strong linear relationship between the two. My (admittedly incomplete) reading of it suggests that there might be an IQ floor below which high-level strategy game performance is unlikely, but that above this floor assiduous practice and access to high level players (and/​or books on the game) seem more important. The latter are the kinds of things that performance based grouping allows across academic subjects. There’s no reason that kids with an affinity for math couldn’t be learning calculus in middle school or early high-school, for example.

So there is much more value in market education reform than simply letting families choose the flavor of curriculum they prefer.