Seeing that it's a Cato Institute article, I anticipate a lot of noise here. I'd like to note, though, that in a standardized, rule-from-the-top system, outlier

Seeing that it's a Cato Institute article, I anticipate a lot of noise here. I'd like to note, though, that in a standardized, rule-from-the-top system, outliers create immense problems, whether the outliers are really smart kids or kids who need additional attention. In a distributed, self-optimizing system, this is not the case. Outliers can be handled in various ways. This observation isn't political. You can observe the same thing in stuff all over the place, like network traffic. If you had universal rules for everything, the internet would tank. Instead we have a (somewhat) distributed and adaptive system using common protocols. Best of both worlds. Perhaps the argument begins at how to create such adaptive distributed systems. If so, that's cool, but that should be the starting place, not a discussion of free markets or social concern, at least in my opinion. (I was very discouraged to hear Bill Gates blow right through this concept when talking about helping education systems. He's trying to quantify and create the universally-optimized teacher. Good luck with that pipe-dream, Bill.)