A new working paper by Helen Ladd and John Singleton, The Fiscal Externalities of Charter Schools: Evidence from North Carolina, appears to confirm the long-held suspicions of district school advocates that public charter schools drain money from districts.  The authors find “large and negative fiscal impact from $500-$700 per pupil in our one urban school district and somewhat smaller, but still significant, fiscal externalities on the non-urban districts in our sample.”  The study raising some valid questions about the use and distribution of scarce taxpayer resources but the input-focused paper answers few of them.

The report is limited in scope, examining only six of the 59 counties that had one or more charter schools during the 2015-16 school year.  In most cases, the six counties in the sample – Durham, Buncombe, Cabarrus, Iredell, Orange, and Union – share few characteristics of the other 53 counties excluded from the study.  That is why it would be unwise to generalize the results to other counties, particularly those outside of the Triangle and Metrolina (Charlotte) regions.

Yet, even for the counties included in the study, I have doubts about one of the central features of the working paper – the determination and treatment of variable and fixed costs.  Variable costs, such as personnel and supplies, can be changed depending on enrollment and programmatic changes in the school district.  For example, if the district loses student population, school district officials can decrease expenditures on teachers or supplies because they are no longer needed.

It is much more difficult to reduce fixed costs, such as school buildings, because they remain necessary regardless of changes in student population.  But fixed costs can be reduced, just not in the same way as variable costs. As Kennesaw State University economics professor Ben Scafidi points out, “fixed costs are only fixed for a given period of time. In the long run, all costs are variable.”

That is why I question the author’s claim that “Intuitively, the money that follows the students who leave public schools to charters requires that a district reduce its variable spending per pupil (as the district cannot reduce its spending on fixed costs).”  An acquaintance who highlighted this statement to me in an email pointed out, “To the extent the N&O is touting this, McClatchy sure as hell has been reducing “fixed cost” over the years or they would already be in bankruptcy.”  No doubt he is correct.

Much of the debate about variable and fixed costs in education also depends on how the researcher classifies them.  Facilities and building maintenance are fixed for longer periods than other fixed costs.  Ladd and Singleton contend that central office administration, school building administration, behavioral support, child and family support, Limited English Proficiency programs, and alternative programs and schools are all fixed costs.  They argue that “charter expansion is unlikely to reduce the demand” for support programs because high-need populations usually remain in district schools.  I am not convinced.

First, it does not address the district school practice of routing a dispropotionate share of racial and ethnic minority students to behavioral support and alternative programs, thus unnecessarily multiplying expenditures on supplementary services.  Moreover, it does not acknowledge that charter schools are becoming less, not more, homogenous. Enrollment of English Learners in charter schools, for example, increased by 72 percent since 2015.  The number of special needs and low-income students in charter schools increases nearly every year due to the proliferation of charter schools that focus on serving these populations.

I also question the exclusion of debt service on public school capital expenditures.  Facilities are the largest fixed expense for school districts, and yet county expenditures on debt incurred for the building and renovation of schools were not included.  Regrettably (but predictably), debt service is rarely included in empirical studies that use public school expenditures as a variable  But the counties, including those included in the study, annually use millions in county tax revenue for school capital outlay.  Most charter schools do not have the luxury of having a portion of their capital expenditures paid on their behalf and must use a portion of their operating funds to pay for these fixed costs.

But even if we grant Ladd and Singleton’s entire analysis without question, the fact is that there was no discernable drop in student performance among the six school districts in their sample.  As alluded to above, Ladd and Singleton acknowledge that their study does not consider academic outcomes, such as student performance on state standardized tests.  For the districts analyzed, all have increased graduation and/or proficiency rates over the last four years.  Buncombe increased their graduation rate by 5.1 percentage points from 2014 to 2017 and had a 1.7 percentage point increase in their grade-level proficiency rate.  While Durham’s graduation rate was largely unchanged compared to 2014, the district had a 2.3 percentage point increase in proficiency during that period.  Orange County Schools gained 2.8 percentage points in their graduation rate and 3.8 percentage points in their districtwide proficiency rate.

If there was evidence of harm, then perhaps Ladd and Singleton’s recommendation that districts receive “transitional aid” should be considered.  But the districts analyzed in the study appear to be improving academically, perhaps a product of competition from charter schools. A 2014 working paper by Stanford University researchers, for example, found that “competition from charter schools with higher than average quality is associated with increased growth in both math and reading” at Washington, D.C. Public Schools.  The research literature is hardly conclusive on the matter, as limitations on charter enrollment and other fiscal barriers to expansion inhibit the development of a truly free market and the competitive pressures that follow.

As I have said to the News & Observer, these types of studies drive a wedge between districts and charters, further discouraging them from collaborating for the betterment of all public school students.  District school advocates complain that the charter school movement is a failure because districts have not adopted their policies or practices.  But the real failure is the unwillingness of district advocates to begin to see charters as collaborators, not rivals.