My colleague Andrew Rotherham and the Century Foundation's Rick Kahlenberg engaged in a bit of a back and forth over the merits of increasing socio-economic integration as a strategy to improve educational opportunities for low-income students. Rick is a strong proponent of public school choice and inclusionary housing policies that enable low-income students to attend predominantly middle-class schools. Andy acknowledges the appeal of this strategy, but points out that geographic and logistical constraints significantly limit its potential to help more than a fraction of low-income children.
Actually, it's worse than that, and neither Rick nor Andy is mentioning the elephant in the room here: Rick's socio-economic integration strategy depends on enabling low-income children to attend low-poverty, predominantly middle class schools. But even in the absence of logistical or geographic barriers, there are just too many poor and low-income kids in the United States for that to be possible.
Consider: In the recent Century Foundation study* that touched off this whole discussion, low-income children reaped the benefits of improve educational outcomes when they attended schools in which no more than 20% of students qualified for free and reduced-price lunch (FRPL).** Children in the study who attended schools with greater than 35% of children eligible for free- and reduced-price lunch reaped no benefits, compared to children who attended schools with much higher rates of low-income students.
Then, consider this: Nationally, 41% of American students are eligible for free and reduced-price lunches. If there were no residential segregation, if poor and low-income students were perfectly geographically dispersed such that their percentage in every district, school, and classroom perfectly matched their percentage in the population at large—The percentage of low-income students in every school, district, and classroom would be too large to produce the benefits of socio-economic integration identified in the Century Foundation's research.
Now, it just so happens that we live in a world in which low-income students are NOT perfectly geographically dispersed. As a result, some schools have very high percentages of low-income students, and some have very low percentages. And because of that, it's possible to move some low-income kids into predominantly middle class schools—and some of the housing and school choice policies Rick supports could in fact enable far more low-income kids to do so than is the case to day. But the success of these socioeconomic policies themselves depends on the continuation of a level of income segregation necessary to create predominantly middle class schools in the first place.
The fact is that, in a nation where 21% of children live in poverty and 42% of children live in low-income families, greater economic integration can be, at best, a solution for only a fraction of children. I'm generally supportive of well-thought out public school choice and housing policies that enable more children to attend more economically integrated schools. But given the demographics and geography of our nation it is absolutely critical that we ALSO have strategies in place to enable low-income children to get a good education in high-poverty schools. We simply can't ensure all our kids access to the educational opportunities they deserve without high-performing, high-poverty schools.
UPDATE: Richard Kahlenberg writes to inform me that the 20% and 35% thresholds for impact in the Montgomery County paper cannot be generalized nationally, and suggests that the actual "tipping point" for high-poverty schools is closer to 50%. But some of the issues that call into question the generalizability of the specific thresholds in the Montgomery County study also raise questions about the generalizability of the results overall. Identifying exactly where the threshold for integration impacts would need to be, and the relative impacts of moving from, say, a school with 85% low-income students to 50% low-income students, or from 50% to 20% seems like an important question for proponents of the socio-economic integration strategy to answer—particularly given the inherent limits our national demographics impose. Note that 40% of students low-income is the threshold for a school to implement "schoolwide programs" under the federal Title I program. If the maximum integration we could possibly achieve would be 42% low-income students in every single school, would that percentage dramatically improve the outcomes for low-income students? Americans like to think of ourselves as a predominantly "middle-class" nation, and so it seems like it should be possible to enable all or most of our students to attend predominantly "middle class" schools. But when you look at the demographics of our children, we're really not as much of a middle class country as our national mythology suggests.
*I do want to give the Century Foundation kudos for focusing in this paper on housing strategies to increase economic integration. While the way district and school attendance boundary lines are drawn does often exacerbate economic segregation, the fact remains that our schools are segregated in large part because the places people live are, too. Inclusionary housing like the Century Foundation talks about here is one solution. Another simple—and cheaper—strategy would be the elimination of zoning and other regulatory policies that are deliberately designed to keep low-income people out of affluent communities. (ie, minimum lot and square footage requirements or bans on renting out "mother-in-law" apartments in single family homes) I'd like to see more creative ideas on how federal educational and housing funds and policies could be used as leverage to push states and localities to eliminate these barriers.
**FRPL status is commonly used as an indicator of poverty in education policy discussions, but it's not actually the same thing as poverty, because the cut-off for FRPL eligibility is much higher than the poverty line—185% of poverty, to be exact.