Opening School Choice to the Suburbs

One of the big problems with the No Child Left Behind Act is its failure to deliver on the promise to allow kids in low-performing schools to transfer to better performing institutions.  Only about 1% of students transfer under the Act’s provisions, in part because in many urban districts, there are very few good schools to transfer into.  To fix this problem, conservatives propose private school vouchers while some progressive education reformers (see here and here) are looking at the possibility of inter-district public school choice: allowing reasonable numbers of students in failing urban schools to transfer to high performing public schools in the suburbs.  We need better information on how many students would benefit from cross-district public school choice.  Unfortunately, a deeply flawed report just published by the Education Sector think tank, may mislead policymakers into underestimating the potential of inter-district public school choice.

On most matters, I’m a fan of Ed Sector.  In addition to my full time position at The Century Foundation, I’m an Education Sector nonresident senior fellow and I have great respect for its two founders, Andy Rotherham and Tom Toch.  I was asked to review an early draft of the paper and hoped to learn a lot about an important question.  Unfortunately, the draft, and the final product, fail to live up to Education Sector’s normal standards of excellence.  The study, “Plotting School Choice: The Challenges of Crossing District Lines,” is based on a fatally flawed methodology.  Employing highly dubious assumptions, the report’s conclusion -- that 80-90 percent of students would not be able to transfer to better performing schools using interdistrict choice -- is essentially meaningless.

In estimating the potential of cross-district choice, two variables are critical: driving distances and space constraints in high achieving schools.   The Ed Sector report makes large miscalculations on both fronts.

1. Driving Distances. The heart of the Ed Sector report is an innovative and useful exercise: using Geographic Information Systems (GIS) software to map the number of high performing schools within driving distance of failing schools.  To my knowledge, this analysis, which Ed Sector conducts for schools in Texas, Florida and California, has never been done before and has the potential to yield helpful information.  Unfortunately, the report’s author, Erin Dillon, chose to assume that children trapped in failing public schools would be willing to travel no more than 20 minutes driving distance in order to attend a superior school.  It’s a bizarre assumption, considering the report’s discussion of the highly successful Metco program in Boston, in which minority students routinely travel "an hour or more" each way to attend suburban schools.  Metco, the report notes, has a waiting list of 15,000 students - one quarter of Boston's student population.  These families are clearly willing to send their children on bus rides of far more than 20 minutes because what is at the end of the ride is considered valuable to them.

2. Capacity. But it turns out that the report’s incorrect assumption about driving distances is even less significant to the final result than an arbitrary assumption it makes about the available space in receiving schools.  One of the oddest findings of the report is that roughly the same number of students would be eligible for seats in better schools whether the driving distance were 20 minutes or 60 minutes. How can that be?  Expanding driving distances increases not only the number of receiving schools, but also the number of sending schools and the report finds that there will be fierce competition for seats.  Why?  Because the report assumes – without any empirical justification whatsoever – that a receiving school will only be able to take in 10% more students than currently attend the school.

In other words, the report's fundamental conclusion (that only a limited number of students would benefit from interdistrict choice) is ultimately based less on the interesting and innovative use of GIS mapping technology and more on the arbitrary assumption that capacity will be capped at 110% of current enrollment. 

So where does the critical and all-determinative assumption about capacity come from?  Thin-air.  The report notes that the Aspen Commission on No Child Left Behind recommends that higher performing schools be required to make at least 10% of seats available to transferring students.  But the Aspen Institute itself does explain its choice of the 10% figure, not does it claim that its recommendation is based on an empirical finding that there is only 10% more space in fact available.

What’s the right number?  It’s very hard to say, in part because school capacity is dynamic, not static.  Capacity is dependent not only on demographic shifts (declining or rising enrollment) but on the decisions of policymakers.  Consider a small school with one class per grade.  At the classroom level, a 10% increase means going from 20 students to 22. It's the equivalent of one family with twins moving into the neighborhood. If a second family moves in, adding a 23rd student, schools don't say, "we're filled, no more capacity." School boards and school officials make adjustments.  Likewise, at the interdistrict choice level, providing a financial incentive to suburban schools to take in low income students (something the Ed Sector report itself recommends) could encourage receiving schools to expand capacity. Likewise, increased magnet school funding in urban areas to attract suburban middle class students could create new space in suburban schools

Social Science 101 suggests that when a variable – in this case, school capacity – is unknown, researchers don’t simply assume the validity of an arbitrary figure.  Instead, a careful social scientist would consider the known variable – empirical findings about the number of schools within a reasonable driving distance – forthrightly admit that the capacity variable is unknown, and then calculate the impact under various assumptions about increases in available space (e.g. 10% capacity, 20% capacity, 30% capacity etc.)  To her credit, the author does note in the fine print of a sidebar (p. 4) that assumptions about capacity have an extraordinary impact on the findings.  “If 12 percent of California students enrolled in lower-performing grade three schools could transfer under interdistrict choice with a 10-percent capacity assumption, 24 percent could transfer if we increased our capacity assumption to 20 percent.”  But by presenting the findings in the main body of the report using only the 10% capacity assumption, the author gives a false sense of precision to the findings. 

There are other problems with the report – including its unduly pessimistic discussion of the benefits of socioeconomic and racial integration – but its failure to provide a meaningful analysis of how many students would benefit represents a major missed opportunity.  Everyone agrees that there are large political obstacles to providing school children with economically and racially integrated school environments.  But there are large benefits as well.  According to a forthcoming paper by Amy Stuart Wells and Jennifer Jellison Holme (summarized here) , interdistrict programs aimed at promoting economic and racial integration can pay extraordinary dividends.

The Education Sector report, with its interesting use of GIS software, could have provided valuable information to policymakers about the potential of interdistrict public school choice by identifying the number of good receiving schools available to students within a 60 minute driving radius, and then applying different scenarios about capacity.   Instead, by adopting an artificially short driving distance (20 minutes), and picking a capacity figure (10% expansion) out of a hat, the report understates the potential of inter-district public school choice.  In so doing, it provides timid politicians with yet another excuse not to give deserving poor and minority students the opportunity to attend good solidly middle-class schools that more affluent students take for granted.   

Comments

[Pam Korcuska]

I have a question about No Child Left Behind. Our school district recently built a school specifically for the children with higher test scores. It is called Mesa Academy for Advanced Studies. A family must apply for their child to go and get teacher letters, turn in test scores and then only 15 children total from any elementary school in the district will be admitted. The children that attend are all given their own lap-tops, they have access to the latest technology and they have smaller class sizes than our traditional school. Is this fair under NCLB?

[newport driving school]

Hi!
Great stuff it is.I just went through your information, hope i enjoyed it.He knows very well about risks which are faced when ever you open a driving school.
Thanks for sharing your information with us.

Thank you!