Matt Yglesias has a piece today on the importance of teachers that he openly described as trollish. The basic idea of the piece is that if you accept the viewpoint that socioeconomic conditions are more important for educational achievement than teacher performance, then you must also think that it makes sense to cut teacher salaries and funnel the savings to the families of poor kids. From twitter and his prior posts, it is clear that this is not Yglesias’ view because he rejects the idea that teacher performance is not that important.
This post received the sort of angry backlash it was designed to receive, especially on the left. Most of the complaints did not attack the argument head-on, but complained about how it was framed, among other things. Earlier today I wrote a very similar post endorsing the basic outline of Yglesias’ claim (although my post came before his). Focusing on teacher salaries is unnecessarily narrow and clearly trollish, but it seems obviously true that at some point, giving a marginal dollar to a poor kid’s family will improve educational outcomes more than spending the same dollar on schools.
From what I gather, Yglesias rejects this idea because conceding it harms his support for the school-focused education reform program, but that’s absurd. Surely Yglesias would concede that if we were spending $100k/year per pupil and then received an extra dollar of revenue, that it would improve educational outcomes more to give the dollar to a poor kid’s family rather than spending it on school-side improvements or reform. This seems undeniable. So what exactly are we arguing about, and how is this particularly damning to anyone’s position?
Once you concede the existence of a funding point where it is better to give marginal dollars to the poor than to schools, the only remaining debate is a technical one about where that point is. But once we have identified that point, it also does not necessarily entail that we must cut school funding in order to fund transfer payments to the poor. In a forced choice scenario, we would want to spend the dollar on the poor and not schools, but the reality of the present scenario is that there is almost certainly other spending that is less helpful than both transfers to the poor and school funding. Maximizing behavior would entail cutting that spending (e.g. military) and financing both transfers and school funding.
I think it is useful to point out that funding the poor is so important that, at a point, it is even better than funding schools. It really helps clarify the nature of the problem being measured by the achievement gap. I endorse that kind of analysis for that reason, but it should not be understood as rendering any obvious policy conclusions. Pointing out the marginal trade-offs in a hypothetical forced choice between paying teachers and transferring to the poor will get people roiled, which is of course why Yglesias did it. But there isn’t much in the form of policy implications that spring from that point because the hypothetical forced choice is not the actual choice we face.