Over at Slate, the normally astute Matthew Yglesias has advanced an odd argument about the role of teachers in education. In a post on Friday afternoon, “Teachers’ Curious Embrace of Their Own Irrelevance,” Yglesias takes an article I wrote supporting socioeconomic integration of schools in the current issue of American Educator, published by the American Federation to Teachers, to suggest that if we follow my logic, we should cut salaries for teachers.

He writes: “But if it’s true that socioeconomic integration is much more important for student achievement than teacher quality, then it seems like a no-brainer to reduce expenditures on teachers (accepting that some good ones may leave and be replaced by somewhat worse candidates) and reinvest the funds directly in the key driver of achievement.”

Yglesias seems excited to have caught teacher unions and researchers who believe in teacher voice (like me) in a logical contradiction. This morning, he tweeted: “Again if ‘The real problem with education is poverty’ shouldn’t we spend less on teachers and more on transfers?”

But his argument fails to take account of the fact that a central reason I and other individuals advocate for socioeconomic integration policies is that we believe that teacher quality matters a great deal and that economic segregation makes it much less likely that low-income students will be educated by strong teachers.

In the American Educator article Yglesias cites, I outline three reasons why low-income students benefit from attending middle-class schools: they “are surrounded by (1) peers who, on average, are more academically engaged and less likely to act out than those in high-poverty schools . . . (2) a community of parents who are able to be more actively involved in school affairs and know how to hold school officials accountable; and (3) stronger teachers who have high expectations for students” (p. 3).

If you believe that teachers are important, as do I and almost every education policy analyst, then the fact that economic school segregation keeps great teachers away from the low-income students who need them most is deeply troubling.

Polls find that teachers care even more about working conditions—a safe and orderly environment where there is strong parental support and an excellent principal—than salary. Because middle-class school environments are much more likely to provide all these working conditions than high-poverty schools, it is very difficult to recruit and retain high quality teachers in high-poverty schools.

Stanford economist Eric Hanushek and his colleagues, for example, estimate that in order to get non-minority female teachers to stay in urban schools, school officials would have to provide a salary premium of between 25 percent and 43 percent for teachers with zero to five years experience (p. 5).

Washington, D.C., provides a powerful illustration of the way in which segregation affects the distribution of excellent teachers. If we assume for sake of argument that Michelle Rhee’s teacher evaluation system (IMPACT) accurately determines which teachers are adding the most value, it seems apparent that economic school segregation is keeping great teachers from educating large number of poor kids. According to an August 2011 analysis in the Washington Post, “only 71 of the 663 teachers who received top ratings on this year’s IMPACT evaluation worked in the 41 schools in Wards 7 and 8,” which have relatively high levels of poverty. “By contrast, the 10 schools in [wealthy] Ward 3 have 135 ‘highly effective’ educators.”

Teachers know that poverty concentrations are bad for education, which is part of why teachers unions have supported socioeconomic integration in places like La Crosse, Wisconsin, Raleigh, North Carolina, and Louisville, Kentucky. This support hardly suggests that teachers view themselves as “irrelevant,” or that every dollar for integration should be offset by a reduction in teacher salaries. Surely there is room for our public policies to support teachers and support integration as mutually reinforcing strategies.