The cost of the charter school cap

Evidence shows low-income, urban students pay the price

IN NOVEMBER, VOTERS in Massachusetts will decide whether to raise the cap on charter school enrollment. The irony is that for most voters—those living in suburban and rural communities with charter enrollment far below the current cap—the vote is inconsequential. The charter cap applies to the percentage each school district’s spending which can be sent to charter schools and most communities remain far below the cap. However, for many parents living in communities which are bumping up against the current cap—cities such as Boston, Holyoke, Chelsea and Lawrencethe stakes are very high. In November, their fellow citizens will determine their children’s future educational options.

Students at Roxbury Preparatory Charter School in Boston.

Students at Roxbury Preparatory Charter School in Boston.

So, the question is, should voters statewide limit the educational choices of parents in low-income, urban communities? And, if so, on what basis might they do so? For instance, is there any evidence that parents are being misled, that charter schools are actually diminishing rather than improving their children’s achievement? Is there any evidence that charter schools are discriminating against English language learners or special education students? Are charter schools really undercutting district schools financially?

Over the past seven years, the Massachusetts Department of Elementary and Secondary Education has worked with university-based researchers to examine the track record of charter schools in Massachusetts. Below, I summarize that evidence in light of the four primary concerns voiced by charter school critics:   (Go here for a list of studies.)

Concern 1: “High-achieving charter schools are just selecting better students and dismissing those who can’t meet their standards. They are not actually raising student achievement.”

At least in Boston and Lynn, we now know that the higher achievement of the oversubscribed charter schools is not simply due to selective recruitment and retention. When schools hold admission lotteries—as required by state law when charter schools have more applicants than slots available— they conduct the equivalent of a randomized controlled trial. At the time of the admission lottery, those applicants who are offered a slot at a charter school and those who are denied are indistinguishable; they have the same prior achievement, parental engagement, and motivation. Yet, when I and a group of researchers from Harvard, MIT, Duke and the University of Michigan subsequently tracked down the admission lottery winners, and compared their outcomes to the lottery losers, we found large differences in achievement.  (We counted all those offered admission at a charter school against the charter school ledger, whether they actually attended or subsequently dropped out. So our results were not driven by selection in recruitment or retention.)

Our evidence suggests that the oversubscribed charter schools in Boston and Lynn are having large, positive, causal impacts on student achievement. Moreover, the magnitude of those impacts is striking: Oversubscribed charter schools in the Boston area are closing roughly one-third of the black-white achievement gap in math and about one-fifth of the achievement gap in English—in a single school year!

Fifty years ago, as mandated by the Civil Rights Act of 1964, a group of sociologists led by James Coleman documented large differences in academic achievement by race, ethnicity, and income in the United States. In Coleman’s results, family background explained a larger share of the variation in student outcomes than schools. Many took Coleman’s findings as implying that schools alone would never close the achievement gap, without dramatic societal changes, requiring transforming neighborhoods and family structure.  A series of failed education reform efforts over the past 50 years seemed to confirm that conclusion. Therefore, the finding that a group of charter schools in the Boston area is closing a substantial portion of the achievement gap each school year is not just good news for many Boston families; it is historically important. The fact that the finding was based on the equivalent of a randomized clinical trial—the “gold standard” of research designs—makes it all the more noteworthy.

Readers should be careful not to conflate the national debate over charter schools with the local one. The Boston charter schools truly are a cut above charter schools nationally. For instance, in 2013 and 2015, researchers at Stanford’s Center for Research on Education Outcomes (CREDO) compared the effects of charter schools across roughly half of the US states. They concluded, “the average growth rate of Boston charter students in math and reading is the largest CREDO has seen in any city or state thus far.”  [italics added]


Thomas J. Kane

Concern 2: “Charter schools do not accept their share of English language learners and special education students.”

Charter schools in Massachusetts have always been prohibited—as a matter of state and federal law—from discriminating against students with special needs or English language learners. Nevertheless, students with special needs and English language learners were underrepresented in charter schools for many years.  As a result, the last time the state raised the cap on charter enrollment, in 2010, the legislation required charter schools to submit plans to actively recruit students with special needs and English language learners. Moreover, the legislation required the state board to consider schools’ success in recruiting and retaining students with disabilities and English language learners when renewing a charter schools right to stay open.

The new law seems to have worked. A recent dissertation by Elizabeth Setren, a graduate student at MIT, documents that the proportion of charter applicants in Boston who are English learners and special education students is now similar to the average for Boston Public Schools. Moreover, just as they do for other students, the oversubscribed charter schools are having large positive impacts on the academic achievement of English language learners and special education students.

Concern 3: “The charter schools are undermining the traditional public schools financially.”      

Critics of charter schools argue that they are undermining the financial health of public school districts, because tuition payments for charter schools are deducted from the state’s aid to local districts.   However, that’s only part of the story. When a student leaves a district school to attend a charter school, the district is also relieved of the cost of educating that student. The school district will see a reduction in state aid, but they also have fewer students to teach.

Nevertheless, when students leave to attend charter schools, the state does not cut its aid immediately.   The state law recognizes that district schools have commitments regarding staffing and facilities which are difficult to adjust quickly when demand declines. Therefore, the state law seeks to soften the transition, by paying both the district and the charter school for the first year after a student has left and by continuing to reimburse the district for one quarter of the students’ cost for the subsequent five years. In other words, taxpayers pay twice for the student in the year of the transfer and one and a quarter times for each of the subsequent five years.

Until two years ago, the state appropriated all the necessary funds to reimburse districts for such transition costs. However, in the most recent two fiscal years, the state covered only about two-thirds of the cost of those reimbursements.

When students choose to attend charter schools, a school district suffers a loss in enrollment, analogous to having its geographic boundaries redrawn. However, it maintains the same property values and ability to raise revenue as before. Eventually, a school district must right-size its staffing and facilities to reflect its smaller size. But successful districts come in all sizes. Once the adjustment is made, there’s no reason to believe that the school district will be any less viable.

For instance, Newton and Wellesley have similar average residential property values (the main source of local tax revenues), but Newton public schools have roughly twice the total enrollment.  Are the students in Newton better off than those in Wellesley because the Newton school district has double the number of students? On one hand, having fewer students makes it harder to spread the cost of a math curriculum expert across the district. On the other hand, the difficulty of managing a district increases exponentially with its size.

If a school district fails to make adjustments in the face of rising charter school enrollment, and it keeps the same number of staff and facilities despite having fewer students, it will pay a double penalty:  Because charter school tuition payments are pegged to a district’s average spending per student, a school district’s charter payments rise when costs per student rise.

In the short run, districts which have seen declining enrollment such as Boston face some difficult choices. Unfortunately, the City of Boston’s budget policy has allowed the Boston school district to postpone the hard decisions, making the adjustment all the more difficult. State aid for education is a line item on the city’s budget, not the district’s budget. When more of Boston’s students switch to charter schools (and state aid declines), it does not automatically lead to a reduction in the school district’s budget—unless the city explicitly cuts the school district budget. In fact, the city has done just the opposite.  According to the Boston Municipal Research Bureau, the school district’s budget grew by 25 percent between 2011 and 2016, despite a 4 percent decline in enrollment.  Ironically, the primary effect of the city’s revenue loss from rising charter payments may have been to slow the growth in expenditures in public safety and other city departments, where expenditures rose more slowly than the school budget. (Expenditures on public safety and other city departments grew by 18 percent and 13 percent, respectively, rather than 25 percent in the schools.)

Concern 4: “Charter schools undermine district schools by drawing away the most engaged students and parents.”

Because the parents who are drawn to charters are presumed to be more engaged and more focused on their children’s education, many worry about the loss of positive peer influences and parental social capital for the students remaining in district schools.  Any such loss would be additive to the financial impacts.

But how large are such effects? Could they be large enough to offset the positive gains in achievement described above?

Between 2009-10 and 2012-13, the proportion of Boston 6th graders applying to a charter school more than doubled, from 15 percent to 33 percent. The proportion of 9th graders applying to charter schools also grew, from 11 percent to 15 percent. As charter applications in Boston have soared, the difference in baseline achievement between charter applicants and non-applicants has narrowed.  Before they moved to charter schools, recent charter applicants in Boston did have higher achievement than the average BPS student, but the difference was less than one-sixteenth of the black-white achievement gap in math and one-ninth of the black-white achievement gap in English. Researchers have typically found that a one-point decline in the average baseline achievement of a student’s classmates leads to a smaller (e.g. one-tenth of a point) decline in the achievement of that student. Therefore, even if there are spillover effects of charter departures on the achievement of students left behind, such effects are likely to be small.


More than two decades ago, the Massachusetts Legislature authorized a series of small-scale experiments called “charter schools.” We now know that many of those experiments worked beyond expectation. In fact, charter schools in Boston and Lynn are generating gains in achievement that are large enough to close achievement gaps by race and income in a few school years. It is an historic achievement, and it’s no wonder that thousands of families in Boston and other cities are now wanting to move to charter schools.

Question 2 in November is not a referendum on local public school districts, of which voters in many communities in Massachusetts are understandably proud. Because the referendum involves the cap on charter school enrollment, it only impacts parents and students in communities that are at or near the current cap. Those communities are primarily low-income urban communities, such as Boston, Chelsea, Holyoke, and Lowell.

Naturally, many voters want an easy choice, free of trade-offs. If so, the evidence is not cooperating. If charter schools were having no impact on student achievement, if such schools were merely selecting better students, then it would be an easy choice to vote against the charter cap increase. The charter school law might be reshuffling public resources for no good reason. But the evidence is clear that that’s not true.

If charter schools were discriminating against English language learners and special education students, many voters would likewise find it easy to oppose the charter cap increase. Charter schools are public schools and many voters believe public schools should serve all students. But, again, the evidence does not support that commonly heard allegation.

That leaves only the concern over the financial and educational impacts on those remaining in district schools. Those costs are widely misunderstood. In Boston, the students remaining in the district schools have been enjoying higher spending per student following the charter departures. Eventually, it’s true that school districts will need to right-size their staffing and facilities to reflect their lower enrollment. Those adjustments become more wrenching the longer they are put off. However, once the adjustments are made, there is no reason to believe that the new, smaller districts will be any less viable than before.

Some voters will question whether the urban districts deserve anything more than the funding that is due to them based on the number of students they teach. Others, more concerned about the potential disruption in district schools, need to recognize that leaving the cap in place is not their only option. For instance, if voters in Newton and Wellesley want to maintain the current staffing and facilities of district schools in Boston and other low income communities, they could advocate increasing the state’s payments to urban districts after charter students depart. Such a policy would at least share the cost of the charter departures broadly across the state. However, leaving the cap in place imposes the entire burden on those parents in the capped communities who are seeking a better education for their children. Witnessing the results of the admission lotteries, we have measured those costs in terms of the diminished achievement of children, and they are sizeable.

Meet the Author

Thomas J. Kane

Professor, Harvard Graduate School of Education
Thomas J. Kane is the Walter H. Gale Professor of Education at the Harvard Graduate School of Education. John Hansen, a Harvard graduate student, provided research assistance for this essay.