The Education Gadfly Weekly

One of the best kept secrets in education policy is that Ohio policymakers have set achievement goals for Buckeye State students. These math and reading targets—and whether students meet them—almost never make headlines, largely because they are buried in a bureaucratic document appended to the state’s federally-required ESSA Plan. Under federal law, every state must establish “ambitious” annual goals for students statewide, as well as designated subgroups (race/ethnicity, economically disadvantaged, students with disabilities, and English learners).

These goals shouldn’t fly under the radar or be deemed a mere paperwork exercise, as they remain crucial to the state’s improvement efforts. They should be a motivating tool that encourages Ohio schools to help students reach higher levels and be touted as a regular checkup on whether the state is making progress against its own expectations. They are also used in the state’s report card system, as districts and schools receive credit (or not) within the Gap Closing component based on whether students meet these targets. If students are falling short, state policymakers and school officials shouldn’t just shrug their shoulders and hope no one notices. Instead, they ought to seek ways to drive the achievement gains needed to meet them.

With that in mind, around this time last year, I took a look at whether Ohio had achieved its achievement goals for the 2021–22 school year. Given the serious academic setbacks during the pandemic, the results were astonishing. Ohio students—including student groups across the board—topped state goals for the year.

The twist: Prior to the release of those test scores, state policymakers substantially lowered Ohio’s achievement goals. As allowed by federal officials, this “recalibration” of the state’s original math and reading goals—established back in 2017—aimed to set more realistic targets in light of pandemic-era learning losses. But with an academic crisis raging, policymakers seem to have eased expectations a bit too much, at least for that year.

Fortunately, Ohio’s achievement goals do escalate over time, somewhat like a balloon mortgage. For the purposes of goal-setting, the state relies on performance index scores, a composite measure of achievement on all state exams. Statewide, the performance index goal increased by 3.2 and 2.4 points in math and English language arts (ELA), respectively, from 2021–22 to 2022–23. Depending on student group, goals rose by about 1–3 points.

How did Ohio students do against the state’s slightly higher 2022–23 goals? As figures 1 and 2 display, students once again fared well against these targets. In math, the overall statewide goal for the year was a score of 74.3, and students narrowly exceeded that target (74.8). Every student group save for one (students with disabilities) outperformed their respective math goal for the year.

Figure 1: Math performance index scores versus state goals, 2022–23

A similar story emerges in ELA. Statewide and across every group—again with the exception of students with disabilities—students outperformed state goals in 2022–23. The results versus goals were slightly more impressive in ELA, likely reflecting the more robust post-pandemic recovery in ELA compared to math. Statewide, for instance, students surpassed the state’s overall goal by 2.3 points in ELA versus just 0.5 points in math.

Figure 2: ELA performance index scores versus state goals, 2022-23

While students continue to exceed state goals, no one should be resting on their laurels. For one, the goals for the past year were still very low—in math, for example, the state set a goal that Black students would score 31.4 points below their White peers in math. That’s an enormous gap and is much larger than Ohio’s original expectation for 2022–23, which was that the Black-White performance index gap would be 18.9 points. Moreover, none of the 2022–23 goals expected any student group to achieve at their respective pre-pandemic levels from 2018–19. And, reflecting the persistent effects of the pandemic, no student group actually outperformed their pre-pandemic counterparts in math or ELA (those comparisons are available in this table).[1]

Another indication of just how far students—especially disadvantaged subgroups—still need to go is to compare current performance to the state’s long-term achievement goals. As noted above, Ohio’s goals escalate each year, and the long-term targets are thankfully more ambitious. For example, Ohio will need to boost the math performance index scores of students with disabilities by 24.1 points by 2030–31 to meet its long-term goal;[2] for Black students, the necessary increase is 20.3 points. There is also a long way to go to achieve the statewide goal of 100 on the performance index (as context, the highest-performing district scored 111.6 last year). Those are sizeable increases that will take serious effort to achieve.  

Figure 3: Math performance index scores versus long-term state goals

Figure 4: ELA performance index scores versus long-term state goals

* * *

With Covid-era disruptions in the rearview, achievement has ticked upward from the pandemic nadir, and students are—for now—meeting state performance goals. But most students still haven’t fully recovered lost ground, and state performance targets are about to get more challenging moving forward. As we start making resolutions for the new year, let’s once again resolve to make higher student achievement our top priority. 

Imagine for a moment that William Shakespeare came back to life for one day and promised to devote all of that time to young students—answering their questions about writing and theater and sharing details about the unique aspects of life in England centuries ago. Or that LeBron James and Michael Jordan offered to come to a school to run basketball drills with aspiring hoopsters. Or that Marie Curie showed up ready to discuss her scientific inspirations, methods, and discoveries with physics or chemistry students. Imagine any of these once-in-a-lifetime learning opportunities being made available to students, only to have the adults who run their schools say, “Sounds like a lot of trouble. We’re going to pass.”

Ridiculous, right? Well, the rough equivalent is currently happening right here in Ohio.

In case you missed it, a total solar eclipse will occur on April 8, 2024. It is a stunning astronomical phenomenon, and the path of totality—the darkest part of the eclipse—will pass through portions of northern Ohio for just over four minutes that afternoon. The event will give observers the chance to examine the solar corona, to view solar winds in action, and to study the Earth’s own ionosphere without the glaring brightness of the sun as an impediment. This is a golden opportunity to expose students to real life science learning. It’s also a once in a lifetime experience: The last total solar eclipse visible in Ohio was in 1806, and the next won’t occur until 2099. In short, the combined knowledge of Galileo, Newton, Einstein, and Hawking will soon come alive for thousands of Ohio students for one day only.

Or at least it would have if their school leaders hadn’t unanimously decided to close their entire districts for the whole day and provide zero educational opportunities whatsoever. Missing out on the rare astrophysical learning opportunity is bad enough, but no English or math or history or art will be taught to students that day either. Many districts are opting to extend their spring breaks by a day, while others jumped through administrative hoops to change their academic calendars and create a three-day weekend. Despite this rare celestial opportunity, nothing of educational value will happen.

Why? Fear, it seems—and of a distinctly Earthly variety.

Large crowds and traffic are expected in prime viewing areas. Spotty Wi-Fi and cell service are deemed to be possible. Police and safety services are expected to be in high demand all day for reasons not fully articulated. It seems local safety officials are driving the concerns, strongly pressuring even those districts that want to hold classes that day to give in and shut down. School leaders are listening, and amplifying the fearful talk.

To be fair, the time of totality is right around 3:00 pm, the traditional time of school dismissal. And, indeed, the idea of school buses and carpools trying to navigate to and from schools at the same time that lots of other folks are stopped in their tracks, staring (through safety glasses, of course) at the sky, is properly concerning. But totality is less than five minutes in duration. And even if police and fire departments are right to expect gridlock for a long time beyond that, that still doesn’t explain why the default response to even that level of concern is to shut down the entire district for the entire day. On the contrary, it’s an overreaction that feels akin to Y2K (remember how that turned out) or to medieval doomsayers prophesying the end of the world because of an approaching comet. Today’s message is the same as in those past examples: Stay home, take shelter, and hope for deliverance from the evils we foresee.

Obviously, student safety is important. But given the extraordinary event in question, is this really the wisest response? Must the educational opportunity of a lifetime be squandered because of overblown worries regarding traffic and dropped phone calls?

There are better responses available. For example, why not have an extended or flexible schedule that day? Perhaps a district-wide science day that starts at noon and ends at 6:00 pm, with livestreams from NASA or in-person demos and explanations guided by local astronomers? With months to plan and local safety services as partners, any of a number of interesting and engaging possibilities could be developed to alleviate safety concerns and maximize the scientific enrichment of students. For the traffic-jam pessimists that refuse to be swayed by the educational potential of an in-person event held safely on school grounds, why not offer some kind of online option led by teachers and supplemented by resources from far and wide?

As we have learned in recent years, fears are easy to stoke, and shutting down is an equally easy response. Finding clever solutions that balance competing priorities—learning and safety simultaneously, new ways of conducting business while still maintaining some traditions—is hard. But taking the easy path in our schools also has consequences—and those consequences are still being felt far and wide. Do we really want to go down this road again?

NOTE: The Thomas B. Fordham Institute occasionally publishes guest commentaries on its blogs. The views expressed by guest authors do not necessarily reflect those of Fordham.

In late November, two large urban Ohio school districts publicly engaged in academic goal-setting exercises. They ended very differently.

In Dayton, the interim superintendent announced big hairy audacious goals—goals so dramatic, they seem unlikely to actually be achieved during his tenure. In Columbus, the school board moved in the opposite direction, rolling back what were already decidedly unambitious goals first adopted just two years ago. In their own ways, however, these recent efforts in Dayton and Columbus help illustrate why it is so hard to improve large, underperforming school systems—and how adult politics gets in the way.

Start with Dayton. Two newly-hired district administrators announced a series of new “Power Metrics”—including lifting the third-grade reading proficiency rate from the current 33.6 percent to 65 percent, and raising eighth grade math proficiency from 10.5 percent to 40 percent.

While setting lofty goals for improvement can be admirable, the district did not provide a time frame for when it expects the targets to be reached, raising some doubts about the seriousness of the district’s commitment to accomplishing them.[1] Instead, one of the Dayton officials said vaguely, “We know school transformation takes three to five years, so we’re trying.”

Given recent trends in Dayton, even if the current interim superintendent does get the job permanently, it seems unlikely he would still be in that position when enough time has passed to evaluate whether these ambitious goals have actually been achieved. Without the real prospect of accountability for meeting these goals, it’s hard to believe that this is a serious goal-setting process.

If Dayton has—at least on paper—swung for the fences, Columbus leaders indicated they would be satisfied just to get a walk to first base. In a unanimous vote, the board adopted a set of new five-year academic goals that represent a dramatic lowering of ambitions.

Back in 2018, the district’s draft strategic plan promised to raise the third-grade reading proficiency rate to 56 percent by 2022–23.[2] (The actual proficiency rate was just 36 percent last spring.) In 2021, the board formally moved the deadline out to June 2026 and lowered the proficiency goal to 55 percent. In last month’s vote, board members pushed the finish line back again, to 2028, and further lowered the target, aiming to only reach a 45 percent rate of proficiency.

The new plan also, for the first time, includes a math goal. That’s at least a start. But the board decided to aim to raise seventh grade math proficiency from below 15 percent last year to a still shockingly low 23 percent by 2028.

All of these numbers can make one’s head spin, so here is the bottom line: At this pace of improvement, it will take Columbus at least twenty years to bring the math achievement of its seventh grade students up to the current statewide average and at least fifteen years to do the same in third-grade reading. (Hopefully, the rest of the state won’t be just standing still in the meantime.) In effect, the new goals write off an entire generation of Columbus school children.

It is worth noting that each variant of the Columbus goals—the 2018 draft strategic plan, the 2021 schedule, and the new plan adopted last month—were drafted by different superintendents. And therein perhaps lies the first indication that such goal-setting efforts reveal some of the underlying dysfunctions that makes these goals so hard to achieve.

Indeed, one explanation Columbus district officials gave for lowering the bar was that they wanted to make the targets realistic and achievable. Aim too high, they warned, and you’ve set the new superintendent up for failure—and perhaps firing—depriving the district of the long-term leadership stability it needs.[3] The diverging Dayton and Columbus strategies raise two important and interrelated questions: What sort of medium-term goals are actually achievable for a large urban district serving an overwhelming number of disadvantaged students of color? And, when trying to figure out what is possible, should we benchmark against other dysfunctional, low performing districts—effectively baking low expectations into the goals—or look to the best performers, while understanding that their success might be difficult to achieve even under the best of circumstances?

To start with the first question, it is useful to look at other models from around the country. In particular, Denver, New Orleans, and Washington, D.C., stand out as exemplars of urban education reform so far this century. In each case, a chronically dysfunctional, under-performing school district experienced a burst of achievement growth—dramatically narrowing the gap with the relevant state or national average during a period of substantial improvement lasting between five and ten years.

But producing these gains required forces of nature—for example, Michelle Rhee in D.C. and a literal hurricane in New Orleans. In each case, the improvements required contentious reforms to teacher evaluations (or outright mass teacher firings), the closure of under-enrolled low-performing school buildings, and major political upheaval. In neither Dayton nor Columbus have education leaders indicated any willingness to pursue policy changes on anywhere near the scale necessary to accomplish the improvements in performance their students so desperately need—or to pay the associated political price from angry, entrenched adult interests. Dayton appears to incredulously promise big gain with little pain, while Columbus has simply thrown in the towel and acknowledged that such gains are all but impossible.

An alternative way to establish a reasonable benchmark is to examine other high-performing non-traditional options already available to students in these cities. Both Dayton and Columbus are home to robust school choice markets, with a number of high-quality charter operators that serve similarly disadvantaged student populations. Why not use look to these schools as evidence of what is possible and set the goals appropriately?

Leaders in Columbus reject this comparison, arguing that running a big city district—with over 44,000 students—presents unique challenges that smaller charter operators simply don’t face. This may be true, but it again risks baking political and operational dysfunction into expectations. After all, if reforming the state’s largest school district is akin to trying to turn the Titanic, there is no reason why Columbus students should have to be remain on a sinking ship. If district officials truly believe it is impossible to improve quickly, as their comments in November clearly indicate, why not encourage families to pursue better options elsewhere in the meantime?

By adopting the new goals, the district seems to have conceded its own inability to achieve significant academic improvements over the next half decade. At the same time, however, Columbus public officials (not only district administrators) have done all they can to prevent existing high-quality charter operators from expanding. Earlier this year, the city council blocked an annexation plan from KIPP Columbus, part of the highly regarded national KIPP chain, in retaliation for the school’s effort to fend of what was at the time an ongoing teacher unionization campaign. And back in 2020, the Columbus school board rejected an offer by the United Schools Network, a homegrown highly successful charter management organization, to purchase one of the district’s vacant buildings, even though it has continued to sit as an empty eyesore for years and the district has no real plan to use it.

Indeed, one reason Columbus leaders have dragged their feet on closing more schools, despite sharp enrollment losses, is fear that the building might be snapped up by charter operators, which have the right of first refusal to purchase unused buildings under state law. At the same meeting where the Columbus school board lowered the district’s academic goals, officials also previewed a forthcoming plan to dramatically reduce intra-district choice and have more students attend their zoned neighborhood schools. One major motivation for this shift is that it would allow the district to reduce the transportation it must provide to high school students attending charter and private schools. (Dayton, too, has faced its own share of controversy over transportation services provided to charter and non-public students.)

Having given up on meaningfully improving student outcomes in district-run schools, Columbus officials seem to have now turned their focus on making it as hard as possible for families living within their district to find better options.

Despite their quite different numerical targets, the Dayton and Columbus academic goals share a lot in common. Both appear to be designed primarily to influence public perception—to, in the words of one Columbus board member, “shape narrative” about each district. Both also seem calibrated to serve the employment interests of their districts’ respective superintendents—to help the interim win the permanent position in Dayton, and in Columbus, to have a chance to remain in the job longer than her recent predecessors. Sadly, neither is likely to meaningfully move the needle on student learning in the coming years.

Vladimir Kogan is a Professor in The Ohio State University’s Department of Political Science and (by courtesy) the John Glenn College of Public Affairs. The opinions and recommendations presented in this editorial are those of the author and do not necessarily represent policy positions or views of the John Glenn College of Public Affairs, the Department of Political Science, or The Ohio State University.

Between 2007 and 2022, California saw its K–12 public school enrollment decrease by more than 390,000 students, or more than 6 percent statewide, according to data from the state’s Department of Education. The baby bust is a large contributor to that decline, as is the rising cost of living driving outmigration from many parts of the state. Inevitably, hundreds of schools have closed in response—nearly 700 between 2012 and 2021. Without a drastic change in circumstances, officials predict a further 524,000 students will leave Golden State public schools in the next ten years, leading to even more building closures. A similar story is playing out in other states as well. A trio of Stanford University researchers, concerned over the possibility that school closures disproportionately impact Black and Hispanic students, looked at both California and national data to quantify the racial and ethnic dimension of closures. Their recent working paper suggests an outsized influence of race as a factor, but fails to take into account the positive impacts that can accrue to all students whose schools close for good reason.

The research team, led by Francis A. Pearman, II, uses data from the National Center for Educational Statistics’ Common Core of Data from 2000 to 2018 to look at school closures both in California and nationwide, focusing on the racial composition of district and charter schools which were or weren’t closed during that period. They also utilize school-level achievement statistics from Stanford’s Educational Opportunity Project (EOP) to build a subset of schools—mainly elementary and middle schools, both closed and not—containing at least one grade level tested annually on state assessments (grades 3 through 8) between 2008 and 2016. The big questions: Do school closures disproportionately impact Black and Hispanic students versus their White peers? And do other differences between schools—like achievement levels, enrollment patterns, and socioeconomic characteristics—drive or mitigate any disproportional impacts found?

The topline finding: California schools enrolling higher proportions of Black students were at significantly increased risk of closure relative to those enrolling fewer Black students. In schools where less than 20 percent of students are Black, fewer than 1 percent closed during the study period. Once half or more of the student body is composed of Black students, closure rates rose to 2.5 percent and exhibited a sharp incline thereafter. The same pattern manifested in the national data, but California’s rates more than doubled those nationally. Above 80 percent Black enrollment, 10 percent of California schools closed, as compared with 3.5 percent nationally. For Hispanic students, closure rates were generally low—both in California and nationally—until their share of the student population reached 80 percent, above which point roughly 2.5 percent of schools experienced closure.

In controlling for possible alternative explanations for school closures—such as differing enrollment declines, school types, poverty rates, and achievement levels—the analysts find some mitigation of the race- and ethnicity-focused disparities, but not enough to fully explain them in either Golden state or national data. In California, for instance, the odds of closure still increased by nearly 25 percent for every 10-percentage-point increase in the share of Black students—even in schools that were otherwise equivalent in terms of charter status, achievement levels, urbanicity, poverty rates, and enrollment trends, and that were situated in districts under the same level of “strain” (in terms of per-pupil expenditures, total enrollment, and level of school choice). In short, Black student enrollment appears highly predictive of which schools are more likely to close, independent of any other explanation, especially in California.

The research team recommends more analysis of these and other factors related to school closures. At minimum, they say, states should adopt policies that require school leaders to make their closure decisions data-focused, transparent, and subject to non-discrimination laws. As team member Francis Pearman told EdWeek: “Deliberate or not deliberate, [race] is showing up, and we have to be really careful about the processes that govern closures to ensure that those processes themselves are equitable.”

All of this makes sense as far as it goes, but the researchers should probably take their own advice first. One vital factor they did not consider in their analysis: the positive impacts of closures. It is noted in the paper that data from elementary school closures in Chicago and Michigan, as well as Fordham’s own research in Ohio, all indicate that many students moved to higher performing schools than the closed buildings they previously attended, and tended to have significantly higher academic gains in subsequent years. There are downsides to closures—disruption for students, possible transportation issues, and socialization with new peers among them—but any credible analysis of impacts must take into account the well-researched academic benefits that can accrue to all students whose schools close and not simply assume that closures are a universal bad. If Black and Hispanic students are indeed the primary constituents whose schools close, then all impacts should accrue to them. School closures are likely to continue—or even ramp up—in California and elsewhere. They are not automatically negative, and maximizing their positive impacts requires acknowledging and maximizing them.  

SOURCE: Francis A. Pearman, II; Camille Luong; and Danielle Marie Greene, “Examining Racial (In)Equity in School-Closure Patterns in California,” Policy Analysis for California Education Working Paper (September 2023).

Foreword

By Amber M. Northern and Michael J. Petrilli

In many communities, public school systems compete with an increasingly complex and frequently overlapping set of alternatives for students and families, including charter schools, public schools in other districts, voucher programs, tax-credit scholarships, and education savings accounts, as well as traditional private schools, microschools, and homeschooling.

We at Fordham view this as a healthy development, both because we believe in the fundamental right of parents to choose schools that work best for their children, and because of the large and ever-growing research literature demonstrating that competition improves achievement in traditional public schools. That “competitive effects” are largely positive should be seen as good news for everyone, as all of us should root for every sector of American education to improve. And it means that the whole “school choice versus improving traditional public schools” debate presents a false dichotomy; we can do both at the same time. Indeed, embracing school choice is a valuable strategy for improving traditional public schools.

Yet, despite the amount of attention that school choice receives in the media and among policy wonks, politicians, and adult interest groups, the extent of actual competition in major school districts is not well understood. We were curious: Which education markets in America are the most competitive? And which markets have education reformers and choice-encouragers neglected or failed to penetrate?

Those questions prompted this analysis, conducted by David Griffith and Jeanette Luna, Fordham’s associate director of research and research associate, respectively. The study seeks to quantify the extent to which competition is occurring by estimating the number of students enrolled in charter, private, and homeschools in each of the nation’s 125 largest school districts in spring 2020 and then dividing that sum by an estimate of a given district’s total student population (which includes students in traditional public schools). The resulting quotient—the report’s measure of the competition facing a district—is the combined market share of all non-district alternatives. While this is not a perfect measure (we can’t account for inter-district open enrollment, for instance), it is as good an estimate as current data allow.

In addition to calculating the competition districts face for all students, David and Jeanette also crunch the numbers for individual subgroups, including Black, White, Hispanic, and Asian students, to determine the percentages of school-age children in those subgroups who are attending non-district alternatives.

As the report explains, even calculating these seemingly simple figures required a bewildering number of decisions. For example, because of the variation in state and local kindergarten policies and the data conundrum relative to high school dropouts, the report relies on enrollment in grades 1–8.

In the end, the study’s findings reveal several interesting patterns. First, most of America’s largest districts face only modest competition for students (though there is considerable variation). In the median district, approximately 80 percent of students enroll in a district-run school, and approximately 95 percent of students do so in the country’s “Least Competitive Large District,” Clayton County, Georgia. Meanwhile, the country’s “Most Competitive Large District”—Orleans Parish, Louisiana, where there are no district-run schools due to reforms implemented in the wake of Hurricane Katrina[FW1]—is a clear outlier. In the country’s second and third most competitive districts, the San Antonio Independent School District (TX) and the District of Columbia Public Schools (DC), only half of students enroll in non-district schools.

In addition to facing low levels of competition in general, most large districts face more competition to educate White students than they do to educate their non-white peers. After all, White students are much more likely to attend private schools. Yet between 2010 and 2020, the ground shifted: More districts are now competing to educate non-white students, thanks largely to the growth of charter schools. Specifically, 116 of 125 large districts saw an increase in non-white students’ access to non-district alternatives in the last decade (with the median large district seeing a 7 percentage point increase). The co-authors commend this trend even as they conclude that we need more, as well as more affordable, non-district alternatives, especially for traditionally disadvantaged groups.

The full report includes several interactive figures that allow readers to see how specific forms of competition have evolved in specific districts and racial subgroups. So, by all means, jump in and click around to your heart’s content.

But before you do—especially if you’re an education reformer keen to expand school choice for kids who need it—take a look at the following table, which shows major urban districts where relatively few students of color attended non-district alternatives as of 2020 (ranked one through seventy). In our view, the districts that appear towards the top of this list are ideal targets for charter school expansion.

Table FW-1: Major urban districts that face the least competition to educate non-white students

//--> //--> Notes: Urban districts are those where private, charter, and traditional public schools with urban locale codes account for more than half of total public and private enrollment.

Let's focus on the twenty major urban districts with the least competition. Table FW-1 shows that two of these districts (Cherry Creek School District 5 and Mobile County) are located in Colorado and Alabama, which have the second and third best “state public charter school laws” (respectively) in the nation, according to the National Alliance for Public Charter Schools. Fayette and Jefferson County school districts in Kentucky and several urban districts in Texas—including in Plano, El Paso and Round Rock—are also in dire need of more options for students of color. The same can be said for Black and Brown kids in multiple urban districts in charter-friendly North Carolina, including in the counties of Forsyth, Cumberland, Wake, Guilford, and Mecklenburg. Making more progress in Nevada could also have a big payoff, given the underperformance of Washoe County (i.e., Reno), where just 12 percent of students of color attend an option other than their district-run school.

So, a big heads-up to school choice advocates in search of fertile terrain. Take a good look at these districts because they need you. They need your energy, your resources, and your commitment to expand charter schools and other options.

Because, at the end of the day, just 7 percent of all public school students attend a public charter school. It’s a proverbial yet precious drop in the bucket. Look, we already know that the growth of high-quality charter schools benefits low-income, Black, and Hispanic students academically. So, it follows that we need more of them.

Who’s listening?

Introduction

Exposing traditional school districts to greater competition is a central but typically unstated goal of education reform in the United States. The case for greater competition rests on two pillars: First, basic economic theory suggests that competitive markets are more efficient than monopolies, at least when appropriately regulated. Second, a growing empirical literature suggests that, despite the inevitable market failures, competition really does work in K–12 education, as it does in most other fields. Monopolies, in the view of many informed observers, are by their nature indifferent to the needs of those they are meant to serve, even when they reflect the will of a fractious majority and especially when they are in thrall to one or more interest groups (as is too often the case).

Yet, despite widespread agreement that competition is desirable, quantifying it is difficult. Outside of track meets and soccer games, we seldom observe competition between schools and districts. Rather, we infer its presence or absence from the movement of students and the presence or absence of enabling conditions. For example, when two charter schools with similar grade spans, missions, and pedagogies locate on the same New York City block, it seems reasonable to assume that they are in competition with one another (after all, families who live within striking range are free to enroll their children in either school, space permitted). In contrast, when there are no charter or private schools within driving distance of a traditional public school, it seems reasonable to assume that competition for the students who live in that school’s attendance zone is minimal (though because Americans have the right to homeschool their children, no school district in the United States is a pure monopoly).

Of course, traditional school districts can also compete, both with one another and with the various alternatives. And in rare cases, enterprising jurisdictions have created programs that have succeeded in fostering internal competition (that is, between district-run schools or programs). Yet, while there is evidence that it has fostered healthy competition in some places, in practice interdistrict open enrollment rarely benefits traditionally disadvantaged students.[1] And even in the growing number of systems with some variant of intradistrict school choice, the incentive to improve is often blunted by the dearth of meaningful consequences for adults when low performance persists.

In short, districts that don’t have to compete for traditionally disadvantaged students can afford to take them for granted, which increases the odds that they will be poorly served. Hence the decades-long quest to create truly independent alternatives to America’s largest school districts—to encourage them to improve and to provide students with viable alternatives should they fail to do so—and the need to periodically assess that endeavor’s success.

Introducing The Education Competition Index: a comparable measure of the “competition” facing the 125 largest school districts in the United States.

WHAT DO WE KNOW ABOUT THE EFFECTS OF COMPETITION?

Research suggests that schools perform better when students aren’t obliged to enroll in them.[2] For example, at least twelve studies that rely on student-level data have found that competition from charter schools has positive or neutral-to-positive effects on traditional public school achievement,[3] while just three studies[4] have found negative effects.[5] Moreover, research also suggests that charters (which nearly always are obliged to compete for students) tend to boost the achievement of students who enroll in them, at least in urban areas.[6] For example, the most recent national analysis by the Center for Research on Education Outcomes (CREDO) finds that students in charters gain the equivalent of sixteen days of learning in reading and six days of learning in math per year.[7] And the literature on charters’ systemic effects—that is, their effects on the average achievement of all publicly enrolled students including those in traditional public schools—is similarly positive.[8]

If anything, the research on competition from private schools is more positive.[9] For example, at least nineteen studies of publicly funded private school choice programs have found positive or neutral-to-positive effects on the academic progress of students who remain in traditional public schools (though in many cases, these effects were small).[10] Yet, to our knowledge, no study has found negative effects. Although some studies overlap geographically, positive competitive effects from public and/or private school choice have now been detected in Florida, Louisiana, Massachusetts, Milwaukee, New York City, North Carolina, and Texas, as well as several national analyses.

Admittedly, the research on the effects of participating in private school choice programs is more complex, with studies of urban programs finding more positive effects and those of statewide programs finding more negative effects, at least in the United States.[11] Access to private school choice, of course, is not confined to voucher programs. But assessing the overall performance of private schools is a fraught endeavor, absent a natural experiment and common assessments (from which private schools are typically exempt). And for similar reasons, we know almost nothing about the effects of homeschooling. Finally, we know less than we would like about how competition of any sort affects outcomes other than academic achievement, though the limited evidence that does exist on such outcomes is also positive.[12]

Data and Methods

To generate a comparable measure of the “competition” facing school districts, we first estimate the number of students enrolled in charter, private, and homeschools in a given geographic school district in the 2019-2020 schoolyear (i.e., the most recent year for which data on all three alternatives are available) and then divide the resulting sums by our estimates of the total student population (which include students in traditional public schools). In other words, our chosen measure of competition (which is reminiscent of market concentration measures) is the combined market share of all non-district alternatives.

Although this measure could theoretically be supplemented with other commonly used proxies for competition, such as measures of entry rates and/or barriers to entry, in practice these alternatives are challenging to operationalize (for example, one of the principal barriers to charter school entry is a hostile political environment, which is difficult to quantify). So, for simplicity’s sake, the report focuses on the most indisputable evidence of competition (i.e., non-district enrollment).

Similarly, although competition and school choice are closely related concepts, our estimates do not include those forms of school choice, such as magnets schools and other intradistrict choice programs, that do not generate competition from the district’s perspective. Nor do we include interdistrict open enrollment, although we do include district-authorized charter schools on the grounds that they typically lack attendance zones (or any sort of neighborhood preference) and thus compete with traditional public schools for students.

Because a nontrivial percentage of students drop out at some point in high school, we also exclude grades 9–12 from our analysis. And because some states/districts don’t require or fund Kindergarten, we also exclude this year (in addition to preschool). In other words, the report’s focus is the competition districts face for students in grades 1–8.

Because education markets are often segmented in practice, in addition to overall rankings, we also estimate competition within the four biggest racial subgroups: White, Hispanic, Black, and Asian. Furthermore, although the first half of the report is based exclusively on data from 2020, the latter half also includes estimates of total and subgroup competition for the preceding decade (i.e., from 2010–20).

To generate those estimates, we first assigned traditional public, charter, and private schools to geographic school districts based on their longitude and latitude and then aggregated their enrollment data for grades 1–8 at the district level (as discussed below, the process for homeschooling is different).

Data on charter and traditional public school location and enrollment come from the National Center for Education Statistics’ Common Core of Data and are essentially comprehensive for the years and districts in question; however, because many virtual schools’ catchment areas extend far beyond the borders of their districts, we excluded these schools when estimating charter and traditional public school enrollment in individual districts.[13]

Data on private schools come from the biannual Private School Universe Survey (PSUS), which is also conducted by NCES. To estimate private school enrollment in the years when the survey wasn’t administered, we first aggregated the data at the geographic school district level and then interpolated (i.e., took a simple average of the values from the two adjacent years). To adjust for nonresponse, we used the school-by-year weights provided by NCES[14] and took a simple moving average of up to five years of data.[15] Because the PSUS doesn’t include school-by-grade level data on race, we estimated White, Hispanic, Black, and Asian enrollment in grades 1–8 by multiplying total enrollment in these grades by the relevant school-by-year-level percentages and then following the other steps outlined above.

Finally, data on homeschooling enrollment come in two forms: First, after conducting a search of state department of education websites and contemporaneous news articles, we were able to locate at least one year of official, pre-pandemic, district-level data on total K–12 homeschooling enrollment for twelve of the thirty-four states with one or more “large” school districts, of which ten had counts that we deemed reliable enough to utilize.[16] Second, to estimate homeschooling rates for districts in the other twenty-six states, we relied on the Parent and Family Involvement in Education Survey of the National Household Education Surveys Program, which was administered in 2012, 2016, and 2019. Of the various data sources, the survey is the least informative if generating district-level estimates is the goal. So, to generate district-specific estimates of homeschool enrollment in the relevant states, we took three steps: First, we took a weighted average of the national estimates for the relevant grade levels for each wave of data. Second, we averaged the resulting estimates, on the grounds that whatever trends they exhibited were as likely to be driven by sampling error as by actual changes.[17] Finally, we adjusted the resulting national averages based on the percentages of students in a district’s traditional public, charter, and private schools who were White, Hispanic, Black, Asian, and “other,” as well as the percentages of all students in those institutions whose schools were classified as urban, suburban, town, and rural.[18] In practice, taking this approach means our estimates for homeschooling in these states vary considerably between districts but relatively little within them.

Although these methods preclude the generation of precise confidence intervals, the limitations of the data mean there is some uncertainty associated with our estimates. Accordingly, we have rounded the resulting estimates to the nearest percentage point for the purposes of the report’s narrative as well as the tables.

In general, our sense is that these estimates are relatively insensitive to the specific choices we have made with regard to the data (see Technical Appendix). However, it is important to recognize that they are all “pre-Covid” and likely understate the current level of competition in some locales, given the documented increases in charter[19] and private school enrollment, as well as homeschooling, in the last few years.[20]

The National Context

Nationally, the proportion of students in grades 1–8 who were not enrolled in a district-run school increased by about three percentage points between 2010 and 2020, from roughly 15 percent to approximately 18 percent. Per Figure 1, this increase was driven by the growth of charter schools, which more than doubled their market share from 3 percent to 6.5 percent; however, of the students in grades 1–8 who didn’t attend a district school in spring 2020, nearly half still attended a private school and perhaps one in five were homeschooled.

Importantly, these national averages mask considerable differences between the major racial/ethnic subgroups. For example, of the White students in grades 1–8 who didn’t enroll in a district-run school in 2020, nearly three-fifths attended a private school, while just one-fifth were enrolled in a charter school, and roughly the same percentage were homeschooled. In contrast, more than half of non-White students in grades 1–8 who weren’t enrolled in a district-run school in 2020 attended a charter school, about one-third attended a private school, and approximately one in seven were homeschooled.

Figure 1: Nationally, non-district enrollment in grades 1-8 increased slightly between 2010 and 2020 thanks to the growth of charter schools.

800) { vizElement.style.width='800px'; // Adjusted width vizElement.style.height='527px'; } else if (divElement.offsetWidth > 500) { vizElement.style.width='800px'; // Adjusted width vizElement.style.height='527px'; } else { vizElement.style.width='100%'; vizElement.style.height='727px'; } var scriptElement = document.createElement('script'); scriptElement.src = 'https://public.tableau.com/javascripts/api/viz_v1.js'; vizElement.parentNode.insertBefore(scriptElement, vizElement); //--> //--> //-->

Notes: This figure shows the percentage of U.S. students in grades 1–8 who enrolled in a charter, private, or home school between 2010 and 2020. Data on charter, private, and homeschool enrollment come from the Common Core of Data, the Private School Universe Survey, and the Family Involvement in Education Survey, respectively.

Overall, about 19 percent of White students, more than 18 percent of Black students, roughly 17 percent of Asian students, and approximately 14 percent of Hispanic students in grades 1–8 were enrolled in non-district schools heading into the pandemic (Figure 2).

Figure 2: The national increase in non-district enrollment in grades 1–8 was concentrated among Black and Hispanic students.

Notes: This figure shows the percentage of U.S. students in grades 1–8 in each of the four biggest racial subgroups who did not enroll in a district-run school between 2010 and 2020.

Per the figure, the share of White students not enrolled in a district-run school increased by about two percentage points between 2010 and 2020 (or about 10 percent). The share of Asian students in non-district schools increased by about three and a half percentage points (or about 25 percent). And the equivalent shares of Hispanic and Black students increased by five and six percentage points respectively (or approximately 50 percent).

Below, we present results for America’s largest school districts.

FINDING 1: MOST OF AMERICA’S LARGEST SCHOOL DISTRICTS FACE ONLY MODEST COMPETITION FOR STUDENTS

Within the 125 largest districts in the country, the percentage of students in grades 1–8 who were not enrolled in a district-run school in 2020 ranged from 5 percent to 100 percent; however, as Figure 3 illustrates, most large districts face only modest competition.

Figure 3: In most large districts, most students still enroll in district-run schools.

Notes: This figure shows the distribution of non-district competition in the largest 125 geographic school districts in the United States in 2020, as measured by the percentage of students in grades 1–8 who did not enroll in a district-run school.

Unsurprisingly, the country’s “most competitive large district” is the Orleans Parish School District, where there were no district-run schools by 2020 due to reforms implemented in the wake of Hurricane Katrina; however, per the figure, New Orleans remains a clear outlier, at least among the larger districts that are the focus of this study.

Per Table 1, in most places, the local education market looks more like those of Clayton County School District in Georgia and Corona-Norco Unified School District in California, which our estimates suggest are the country’s “least competitive large districts” and where approximately 95 percent of students enrolled in a district-run school in 2020. In the median large district, roughly four out of five students in grades 1-8 still enroll in a district-run school.

Table 1: Which large school districts face the most competition for students?

//--> //--> Notably, there is a moderate correlation between urbanicity and the percentage of students not enrolled in a district-run school (r = 0.49), with highly urban districts facing roughly twice as much competition on average as nonurban districts (Figure 4).

Figure 4: On average, urban districts face more competition for students than non-urban districts.

Notes: This figure shows the relationship between the percentage of students in grades 1–8 who attended a school in urban locale and the percentage of students in grades 1–8 who did not enroll in a district-run school in the largest 125 geographic school districts in the United States in 2020.

Given the practical barriers to greater competition in more rural areas, the urban focus of many reform efforts, and the comparatively strong performance of urban charter schools and private school choice programs, this relationship makes sense.

FINDING 2: BECAUSE NON-WHITE STUDENTS HAVE LESS ACCESS TO PRIVATE SCHOOLS, MOST LARGE DISTRICTS FACE MORE COMPETITION FOR WHITE STUDENTS

As of 2020, non-White students had less access to private schools than White students in every major school district in the United States, with the median district exhibiting a White/non-White gap of approximately fourteen percentage points (Figure 5).

Figure 5: In most large school districts, White students have more access to private schools than non-White students.

Notes: This figure shows the distribution of the White/non-White private school enrollment gap in grades 1–8 in the largest 125 school districts in the United States in 2020. Data on private school enrollment come from the Private School Universe Survey.

Similarly, White students were less likely to be enrolled in a district-run school than non-White students in 112 of the nation’s largest 125 districts, with the median district exhibiting a gap of about twelve percentage points (Figure 6).

Figure 6: Most large school districts face more competition for White students than non-White students.

Notes: This figure shows the distribution of the White/non-White “competitiveness gap” in the largest 125 geographic school districts in the United States as of 2020, as measured by the percentages of White and non-White students in grades 1–8 who did not enroll in a district-run school.

Per Table 2, the large school district with the biggest White/non-White “competitiveness gap” is Santa Ana Unified, where roughly 82 percent of White students were not enrolled in a district-run school in 2020 compared to just 17 percent of non-White students (a difference of 65 percentage points).

Table 2: Where is the gap between the competition districts face for White and non-White students largest?

//--> //-->

Still, focusing exclusively on gaps threatens to obscure the bigger story in some districts. For example, despite having an unusually large competitiveness gap, San Antonio ranks second when it comes to the competition districts face for non-White students’ (see Table 3). And some districts have notably competitive markets for particular non-White subgroups. For example, about half of Milwaukee’s Hispanic students and roughly two-thirds of Atlanta’s Asian students are not enrolled in district-run schools.

Click on the column headings to see which districts face the most/least competition for students in the different racial/ethnic groups, as measured by the percentage of students NOT enrolled in a district-run school.

Table 3: Which large school districts face the most competition for students in different racial/ethnic groups?

//--> //-->

FINDING 3: OVER THE LAST DECADE, MOST LARGE DISTRICTS FACED INCREASING COMPETITION, ESPECIALLY FOR NON-WHITE STUDENTS

Of the country’s 125 largest school districts, 105 faced increasing competition for students between 2010 and 2020; however, in most cases, this increase was modest. In the typical large district, the share of students not enrolled in district-run schools increased by fewer than ten percentage points, and only one district saw an increase of more than twenty percentage points (Figure 7).

Figure 7: In most large districts, the share of students not enrolled in district-run schools increased modestly between 2010 and 2020.  

Notes: This figure shows the distribution of the change in non-district enrollment in grades 1–8 in the largest 125 geographic school districts in the United States between 2010 and 2020.

Unsurprisingly, the district that experienced the biggest increase in competition was New Orleans, where the share of students not enrolled in a traditional public school increased from 75 percent to 100 percent between 2010 and 2020. In contrast, the share of students not enrolled in a traditional public school declined in Cincinnati (from 40 percent to 35 percent) and at least a dozen other major school districts (Table 4).

Table 4: Which large school districts saw the biggest increases in non-district enrollment?

//--> //--> Importantly, non-district enrollment increased more quickly among non-White students than it did among White students, with the typical large district seeing about a four-percentage-point increase among White students and a seven-percentage-point increase among non-White students. Of the 125 largest districts, 114 saw an increase in non-district enrollment among non-White students between 2010 and 2020, eighty-nine saw an increase among White students, and eighty-four saw an increase among both White and non-White students.

Finally, some districts saw notably rapid change for particular non-White subgroups. For example, Hispanic students’ non-district enrollment increased by twenty-three percentage points in Indianapolis. Places like Chandler, Arizona, saw even larger increases in Black students’ non-district enrollment. And in Douglas County, Colorado, Asian students’ non-district enrollment increased by forty-two percentage points.

Figure 8: How has competition for students in different racial/ethnic groups evolved in your school district?

800 ) { vizElement.style.width='600px';vizElement.style.height='527px';} else if ( divElement.offsetWidth > 500 ) { vizElement.style.width='600px';vizElement.style.height='527px';} else { vizElement.style.width='100%';vizElement.style.height='727px';} var scriptElement = document.createElement('script'); scriptElement.src = 'https://public.tableau.com/javascripts/api/viz_v1.js'; vizElement.parentNode.insertBefore(scriptElement, vizElement); //--> //--> //-->

Notes: This figure shows how the percentage of students in grades 1–8 who did not enroll in a district-run school changed within each of the major racial/ethnic subgroups between 2010 and 2020.

FINDING 4: IN MOST LARGE DISTRICTS, NEARLY ALL OF THE INCREASE IN COMPETITION FOR NON-WHITE STUDENTS WAS ATTRIBUTABLE TO THE GROWTH OF CHARTER SCHOOLS

Across and within the largest 125 districts in the country, the increase in charter school enrollment between 2010 and 2020 was responsible for most of the increase in competition during that same time period. For example, the dramatic increase in New Orleans’ overall competitiveness was driven by an inexorable increase in charter school enrollment. And the story in most other large districts is similar, especially for non-White students. In New York City, the increase in competition for Black students was entirely attributable to the growth of charter schools, as was the increase in competition for Hispanic students in Los Angeles and many other districts.

Figure 9: Which sector drove the increase or decrease in competition in your district and/or community?

800 ) { vizElement.style.width='1200px';vizElement.style.height='527px';} else if ( divElement.offsetWidth > 500 ) { vizElement.style.width='1200px';vizElement.style.height='527px';} else { vizElement.style.width='100%';vizElement.style.height='727px';} var scriptElement = document.createElement('script'); scriptElement.src = 'https://public.tableau.com/javascripts/api/viz_v1.js'; vizElement.parentNode.insertBefore(scriptElement, vizElement); //--> //--> //-->

Notes: This figure shows the percentage of students in grades 1–8 who enrolled in charter, private, or home schools between 2010 and 2020. Data on charter, private, and homeschool enrollment come from the Common Core of Data, the Private School Universe Survey, and the Family Involvement in Education Survey, respectively. 

Takeaways

1. The death of traditional public schools has been greatly exaggerated.

Per Finding 1, the typical American student still attends a school that is administered by his or her local school district. And because the typical school district is only exposed to modest levels of competition, that school still has little incentive to change its behavior if that student is ill-served.

2. Most communities would benefit from the creation of more and/or more affordable non-district alternatives, especially for traditionally disadvantaged groups.

Per Finding 2, while the gap between White and non-White students’ access to non-district schooling has narrowed, it has not disappeared in most large districts. And despite the evidence that traditionally disadvantaged groups are particularly likely to benefit from non-district alternatives, the approach many places take to K–12 education policy is still reminiscent of that old and morally dubious adage: “Choice for me, but not for thee.”

3. Progress is possible.

In the wake of the Covid-19 pandemic, there is little to cheer when it comes to public education. Yet for advocates of equal opportunity and better schooling, the decade just before the pandemic was a time of real progress, at least when it came to the expansion of choice and competition in the communities served by our country’s largest school systems.

Let’s get back to that.

LIMITATIONS

The data that are the basis for this analysis are imperfect. For example, when it comes to charter schools, the fact that we don’t have students’ home addresses means we can’t definitively link them to the districts in which their schools are physically located (though, “competition” is arguably also about perception, which is necessarily impacted by the presence of non-district schools).

In the case of private schools, we are again limited by the absence of home addresses. Moreover, we are also limited by the fact that our data source is a survey, which means that approximately 20–30 percent of schools are missing (though we are at least partly addressing this problem by weighting for nonresponse).

Finally, when it comes to homeschooling and/or virtual schooling, we lack reliable district-specific data for most locations, and our estimates are perhaps better characterized as educated guesses. Fortunately, of the three types of competition, homeschooling is the least prevalent, so whatever error is baked into our estimates is unlikely to affect the overall story (see Technical Appendix).

In addition to these challenges, our analysis doesn’t account for interdistrict open enrollment, which research suggests is yet another salutary form of competition.[21] Nor does it capture the inevitable differences in districts’ responses to competition.

In short, the Index is a necessarily imperfect measure of competition, not the thing itself.

Technical Appendix

Table A1 shows the estimated percentages of students in grades 1-8 who were not enrolled in a traditional public school in the twenty-five ""large"" U.S. school districts where reliable district-level data on homeschooling enrollment could be located for the 2019-2020 school year with and without the inclusion of these data as opposed to Fordham's homeschooling enrollment estimates. Per the table, the official homeschooling data differ from Fordham's estimates for many of these districts (though in some cases, these differences are obscured by rounding). However, because homeschooling accounts for a modest share of total non-district enrollment, the implications for districts' overall competitiveness and ranks are modest (with the possible exception of Marion County's rank, which does change somewhat due to the number of districts with similar levels of competitiveness). Among other things, this suggests that the overall estimates and ranks of districts where official homeschooling data could not be located are close to the mark.

Table A1: Sensitivity Analysis

Name	With district-reported homeschooling enrollment		With Fordham's estimates of homeschooling enrollment
	% Non-District	Rank	% Non-District	Rank
Boston School District	34	22	36	18
Brevard County School District	27	38	26	43
Charleston County School District	23	56	24	50
Cincinnati City School District	34	25	35	19
Cleveland Municipal School District	44	8	45	8
Columbus City School District	36	17	37	16
Dade County School District	37	15	37	15
Duval County School District	33	29	32	31
Fairfax County Public Schools	12	101	13	99
Henrico County Public Schools	11	105	12	105
Howard County Public Schools	10	111	11	107
Lake County School District	33	30	30	33
Lee County School District	20	68	21	67
Loudoun County Public Schools	6	121	7	120
Marion County School District	18	79	16	92
Milwaukee School District	47	5	47	5
Montgomery County Public Schools	16	93	16	84
New York City Department Of Education	30	33	31	32
Palm Beach County School District	24	51	23	56
Pasco County School District	17	81	17	83
Philadelphia City School District	45	7	46	7
Prince William County Public Schools	6	123	7	121
Seattle Public Schools	26	42	27	40
Virginia Beach City Public Schools	8	116	9	115

Acknowledgments

This report was made possible through the generous support of the Charter School Growth Fund, as well as our sister organization, the Thomas B. Fordham Foundation. We are deeply grateful to Emily Gutierrez at the Urban Institute, who reviewed an early draft of the report, and to Pamela Tatz for copyediting.

On the Fordham side, we are grateful to Griffith and Luna for authoring the report, to Adam Tyner for developing the report’s interactive data elements, to Chester E. Finn, Jr. for providing feedback on drafts, to Stephanie Distler for managing report production and design, and to Victoria McDougald for overseeing media dissemination.