The abstract of a paper by Cynthia (CC) DuBois andDiane Whitmore Schanzenbach says,

This paper examines the effect of a court-ordered hiring guidelines intended to increase the share of black teachers employed in a school district in Louisiana. We find that the court-ordered hiring policy significantly increased the share of teachers who are black in the district relative to the rest of the state, and to a matched synthetic control sample. The policy also increased the share of new teachers hired who are black, and decreased the student-teacher representation gap, defined as the difference in enrollment share black among students and teachers in a district. There were increases in the share of black teachers observed in both predominately white and predominately black schools in the district. The policy had no measurable impacts—either positive or negative—on district-level measures of student achievement.

In 1994, Benjamin S. Bloom wrote,

Using the standard deviation (sigma) of the control (conventional) class, it was typically found that the average student under tutoring was about two standard deviations above the average of the control class (the average tutored student was above 98% of the students in the control class). The average student under mastery learning was about one standard deviation above the average of the control class (the average mastery learning student was above 84% of the students in the control class)

The reader who sent me the link to this paper asked whether it invalidates the Null Hypothesis. The researchers experiments rather than observational studies, so I will give them that. But

1. Usually, educational interventions have small effects. The effect of “mastery learning” of one standard deviation seems rather implausibly high, considering its definition.

Formative tests (the same tests used with the conventional group) are given for feedback followed by corrective procedures and parallel formative tests to determine the extent to which the students have mastered the subject matter.

That does not sound like something that would cause a one-standard deviation difference. My guess is that these findings would not replicate if they were undertaken by different researchers.

2. Even when interventions show large effects for a single subject in a single year, the effects tend to fade out. That is, if you examine the experimental group and the control group three years later, any difference has vanished. Even if these results replicate in the short term, they do not invalidate the Null Hypothesis if they suffer from fade-out.

3. We do not know how what would be necessary to enable tutoring to scale. Bloom seems to believe that tutoring works by adapting to the needs of the student. If so, then my guess is that the process of matching tutoring style to student characteristics would be quite a challenge.

Salim Furth writes,

we do not, in the suburbs, have a system of public schools. We have private, government-run schools. A public good is something available to all—non-excludable and non-rival in consumption, like clean air or a radio broadcast. But access to local school is eminently excludable: those who do not buy or rent a home in the right area cannot access it. And it is at least somewhat rivalrous in consumption, since crowding and peer effects play such a large role, at least in the perception of educational quality.

He says that universal school vouchers are a political non-starter.

The two most stable organizing principles of the political economy of the American family in the twenty-first century are that educational access is purchased with one’s home, and that established suburbs do not change their character.

I liked the distinction drawn by a commenter.

distinguish two situations. One, the student wants to learn about something, because she finds it interesting or wants to use it for something. She wants to learn and wants the knowledge to stay with her.

In the other situation, the student is not inherently interested and/or does not expect to use the information. The student may well want to learn enough for long enough to pass but doesn’t mind if all that knowledge just decays away once the course (or the test!) is over.

Right. The first situation calls for providing content and feedback. The second situation calls for providing reward and punishment (assuming that we know what is best for the student). Note that the reward can be psychic reward, such as the student feeling closer affinity with a teacher that the student respects.

He wrote,

I just returned from a week in China meeting with parents whose focus on kids’ education is extraordinary. One of the areas I found fascinating is how some of the most advanced parents are teaching their kids new languages: through games. On the tablet, the kids are allowed to play games, but only in French. A child’s desire to win fully engages them and drives their learning rapidly.

He also puts in a plug for the “illustrated primer” of Neal Stephenson’s The Diamond Age.

Read the whole thing. Plenty of interesting ideas, but keep in mind the null hypothesis.

In the course of the podcast with Russ Roberts, Tyler Cowen says

I think that a Harvard/California could work. I believe normatively Harvard should do it. I see zero signs they are about to. It would mean a dilution of control, a lot of headaches, a lot of new legal issues. You know, some reputational risk. But you could increase the number of people getting into some version of Harvard by really quite a bit. And that would be a wonderful thing for the country. And the world.

This is during a long digression on whether elite colleges could expand by orders of magnitude.

Suppose Harvard set up branches around the country, thinking that it could use its brand name to expand to, say, 250,000 students. Think about how this would play out. Many more students could get into Harvard. Assuming that other elite schools did not expand, Harvard would become by far easier to get into than Princeton or maybe even Maryland. So I think you ruin the Harvard brand.

It seems to me that this is an example in which value depends on scarcity.