Recently, I wrote a column suggesting that U.S. K-12 public schools spend more than is strictly necessary, and that public school administrators, and policymakers more broadly, need to be more mindful of cost-effectiveness. The response has been overwhelmingly negative, and more than one critic has accused me of being, well, stupid. Two issues have come up more than once.
First, I referenced the extraordinary success of the Recovery School District (RSD) in New Orleans, where student outcomes have improved dramatically without a substantial increase in per-pupil expenditures. I’ve literally heard that it is silly to use the RSD as an example, as the reason the RSD has fared so well is that its composition is whiter and more affluent than it was before Hurricane Katrina. Since 2005, the share of New Orleans public school students eligible for free or reduced price lunch (FRL) increase by 7 percentage points, to 84 percent as of February 2014, yet performance has increased by 41 percent. The FRL share in the New Orleans public schools is quite high even for a large urban school district. It is higher than it is in Newark, a school district that I briefly discuss in my column, where performance hasn’t increased by much. It is true that African-American share of the student population has declined, from 93 percent pre-Katrina to 86 percent in 2011. It is also true, however, that it is black students who have made the most dramatic gains. As David Osborne reports in “Born on the Bayou,” if one only counts black students, New Orleans had the lowest test scores in the state before Katrina, while in 2011 it was above the state average. Is this because it was only the most affluent black families that remained in New Orleans? This is hard to square with the fact that the share of FRL-eligible students increased substantially in the post-Katrina period.
So what exactly happened in New Orleans, and is my enthusiasm for the RSD model an indication of hatred for teachers? Those of you who read my column carefully will notice that I don’t posit teachers as villains. Rather, I explicitly call for giving school leaders more autonomy so that they can “figure out what works best for their teachers and their students.” I chose to reference both teachers and students for a reason: if you have unhappy teachers, you’re likely to have ineffective teachers. The reason I am so enthusiastic about New Orleans is that the RSD has found an approach that appears to be benefit teachers as well as students.
Neerav Kingsland, the outgoing CEO of New Schools for New Orleans, attributes the success of the RSD to its success in attracting high-quality charter networks to the city and facilitating their expansion. Kingsland differentiates between “reformers,” district leaders who seek to improve system-wide performance by leveraging the power of centralized, top-down management, and “relinquishers,” who cede control over the day-to-day management of schools and networks of schools, yet who invest time and energy in monitoring outcomes, ensuring transparency, and helping good school grows and bad schools either improve or go out of business, all without causing too much disruption. The former role is a familiar one, albeit it an extremely difficult one, particularly in the context of large urban school districts, which serve diverse constituencies with varied needs. The latter role is in many respects a more manageable one, as it draws on the decentralized knowledge of a wider array of educators, yet it is also less familiar and thus more politically fraught, and threatening to those who’ve grown accustomed to the current way of doing business. Many teachers, for example, rightly see collective bargaining as their chief defense against the irrational, inconsistent, and often politicized demands of district leaders. Yet as Kingsland argues, teachers in a well-functioning charter market might find their labor conditions improve not because of their political clout, but because schools and networks of schools are actively competing for their services. Unionization is a natural response to monopsony in the labor market. But if teachers find that they don’t have to move to the suburbs or to another city to find a public school willing to offer them more favorable terms, they are in a much stronger negotiating position. Instead of just demonizing teachers unions, those of us who favor charter districts and more flexible labor markets for educators need to address the underlying demand for them. And the demand is driven in no small part by the compliance costs and the bureaucratic rigidities that plague traditional public school systems.
Second, many of my detractors focused on the rise of special needs students. Let’s start with a conceptual question. What exactly is a special needs student, and is the concept an immutable one? I’d suggest that it is not. Rather, I’d suggest that incentives shape this landscape. Consider, for example, Alan Schwarz’s reporting on attention deficit hyperactivity disorder for the New York Times. Recent data from the CDC reveals that 15 percent of high school-age children have been diagnosed with ADHD, and the number of children medicated for the disorder has increased from 600,000 in 1990 to 3.5 million. Is it necessarily true that that the prevalence of the underling disorder has increased by a factor of 6, or could it be that the interaction between a pressure campaign by pharmaceutical campaigns to “raise awareness” of ADHD by preying on the anxieties of parents and educators, a desire on the part of parents to improve the performance of their students, and (well-intentioned) publicly-sponsored coverage expansion efforts that helped medical providers expand their customer base? Moreover, federal and state governments have increased funding for students with special needs, which creates an incentive for school districts to designate more and more students as special needs students. What we’ve seen over the years isn’t so much an increase in the number of students with “hard” disabilities, but rather an increase in the number of students with “soft” disabilities, where diagnosis is often a non-obvious call. If you’re on the fence as to whether or not a student ought to be designated as having special needs, is it possible that you will choose to do so if there is a financial incentive attached?
This is not to say that there aren’t students with profound disabilities who need special attention. There are indeed special needs students who are expensive to serve. But does this reflect some intrinsic cost, or does it reflect laws, regulations, and staffing requirements that prevent the emergence of innovative instructional models that might allow educators to meet the needs of students with special needs at lower cost?
Regardless, the point my critics are making is that my fixation on high per-pupil spending is silly because special education students are driving the increases. There is some truth to this: one of the leading experts on the subject, Thomas B. Parrish, has observed that the ratio of total spending per special education and general education student appears to have remained fairly constant over time at about two to one and that special education students represent a rising share of enrollment. (The data on special education expenditures is terrible, by the way. Something tells me this reflects the incompetence and mismanagement of K-12 administrators — not teachers — that is the focus of my column.) Yet Parrish he’s also noted that “while the costs of educating special education students have increased substantially over time, the cost per general education student has increased at a comparable rate.” That is, no, it is not the case that all cost growth, or even most cost growth, can be attributed to the rising prevalence of special education students: the cost per general education student has been increasing substantially over time.
How should we think about special education in the context of Milwaukee’s school choice program, which I also referenced in my column? In 2012, Patrick J. Wolf, John F. Witte, and David J. Fleming released a report on precisely this subject. They found that while the Milwaukee choice schools (MPCP) served a smaller share of disabled students than public schools (MPS) — between two-fifths and three-quarters the rate of MPS — the MPCP schools were far less likely to classify students as having disabilities than MPS:
[W]e conducted site visits of 13 MPCP schools in part to learn about how they serve students with disabilities. What we observed during those visits confirmed claims in the research and policy literature that most private schools lack the incentives, personnel, protocols, and organizational culture that lead public school systems to label students with disabilities as requiring special education services. In many cases, private school personnel hesitate to count a student as having a disability, even if public school personnel would recognize the student as such. However, that does not mean that private schools do not enroll students who would be formally designated as students with special needs if they were in the public schools.
Though I doubt that any of this analysis will convince those who believe that U.S. K-12 public schools don’t spend nearly enough, and that they currently deploy their resources efficiently and responsibly, of anything at all. But I enjoy beating my head against a wall. I think of it as a kind of aerobic exercise.
And finally, I was very pleased to see Pat Brennan and Justus Myers weigh in on the question of how public schools compensate teachers. If you’re interested in the subject, I recommend reading Stretching the School Dollar, and particularly Marguerite Roza on per-unit costs in education and Steven F. Wilson on “the efficient use of teachers.” Containing costs doesn’t have to mean layoffs. It can also mean deploying teaching resources more efficiently by, say, increasing class sizes for the most effective teaches, or making capital investments that seem expensive when viewed in absolute terms but which seem quite reasonable when you factor in labor cost savings over time.