The Education Gadfly Weekly

The Fordham Institute’s new report, High Stakes for High Achievers: State Accountability in the Age of ESSA, examines whether states' current or planned accountability systems for elementary and middle schools attend to the needs of high-achieving students, as well as how these systems might be redesigned under the Every Student Succeeds Act to better serve all students. It finds that the overwhelming majority of states provide schools with few incentives to focus on their high-achieving students. This is a problem.

Accountability has been a central theme of education reform for almost two decades, driven by the unchallenged central finding of James Coleman’s seminal 1966 study: Although some interventions are demonstrably more effective than others, there’s no direct link between what goes into a school by way of resources and what comes out by way of student learning. Sage policy makers have recognized that trying to micromanage school and district “inputs” is a waste of time. Instead, the prudent course is to (a) clearly state the results that educational institutions ought to produce, (b) assess how satisfactorily those results are being achieved, and then (c) hold schools and school systems to account, with rewards of various sorts for success and interventions of various sorts in the event of institutional failure.

This strategy has worked fairly well. After years of stagnation in the late 1980s and early 1990s, achievement began to rise again in the late ‘90s—particularly in the earlier grades and most notably in math—as states set new academic standards, started testing their students regularly, and installed their own versions of “consequential accountability” systems. Once No Child Left Behind (NCLB) made this reform regime inescapable, “late adopter” states—those jurisdictions that hadn’t already moved in this direction on their own—also started to see gains. Rigorous studies have shown that accountability deserves at least some of the credit for these improvements. That isn’t too surprising, considering that just about every person and institution does a little better at any number of undertakings when consequences follow from success and failure.

So far, so good. Yet we must not gloss over critical details. Early proponents of accountability in public education tended to speak in generalities; it was said, for example, that we needed to hold schools accountable for “raising student achievement.” But whose achievement? All students? In which subjects? Measured how?

NCLB provided its own answers to these questions. Schools would be held to account for getting greater proportions of their students—and greater proportions of key subgroups—to “proficiency” in reading and math. States would define “proficiency” as they saw fit, but they would eventually need to sanction any school that didn’t raise all of its students to that level.

Faced with these requirements, most states did the rational thing and set the proficiency bar low. That move, combined with NCLB’s mandatory cascade of sanctions, created a powerful incentive for schools to pay close attention to students below proficiency. Conversely, there was absolutely no incentive to worry about the achievement of those who had already reached (or were likely to reach) that bar. To put it bluntly, NCLB did some good for America’s struggling pupils, but for high achievers, it mostly just hit the education pause button.

Research has demonstrated that students just below the bar were most likely to make large gains in the NCLB era, while high achievers made lesser improvements. Those most victimized by this regime were high-achieving poor and minority students—kids who were dependent on the school system to cultivate their potential and accelerate their achievement. (Equally able youngsters from middle-class circumstances have other people, supports, and educational resources to keep them moving forward.) The good news is that accountability works: Districts, schools, and educators do respond to its incentives and disincentives. The bad news is that kids can get left high and dry when policy makers incentivize schools to pick winners and losers.

Why Focus on High Achievers?

Many education reformers look at results for the National Assessment of Educational Progress (NAEP) and other macro measures and see some positive trend lines in recent decades. Gaps are indeed closing, especially between low- and high-achieving students. Isn’t that what we want?

Yes, of course—up to a point. Our K–12 system has historically done the greatest harm to our lowest-performing students, who tend to come from poor and minority families. Using accountability (as well as school choice and other strategies) to improve matters for disadvantaged children has therefore been, and should remain, a policy focus.

But it should not be the only focus. The policy challenge going forward is to devise accountability systems that raise the ceiling as well as the floor. This is partly about fairness: It’s wrong for any child to miss out on academic challenges at school, and we should do everything we can to develop the full potential of all our students, including high achievers. We must also remember, though, that the country’s future economic competitiveness, scientific leadership, and national security depend disproportionately on how successfully we maximize the learning of our ablest children. If we want tomorrow’s scientists, entrepreneurs, and inventors to “look like America,” our schools need to take special pains with the education of high-ability kids from disadvantaged circumstances. They, too, should have the chance to realize the American Dream.

There’s a political argument as well. How can we expect parents to support public education when many of their children aren’t a priority for the schools they attend?

And there’s a powerful case to be made for accelerating social mobility by educating high-ability, low-income children. These are the poor kids—many of them from minority groups—who have the best chance to succeed in selective universities, become leaders in their communities, and climb the ladder to the middle class. Yet they are also the kids most dependent on the education system to recognize and draw out their potential. Research from Fordham, the Jack Kent Cooke Foundation, and elsewhere shows that these low-income “high flyers” are likeliest to “lose altitude” as they make their way through school. The result is an “excellence gap” rivaling the “achievement gaps” that have been our policy preoccupation.

NCLB-style accountability is partly to blame for that. After all, low-income high achievers are likely to attend high-poverty schools, which face the greatest pressure to raise the test scores of their lowest-performing students and neglect their top pupils. They’re also schools that typically face a host of other challenges.

Going forward, policy makers who care about their low-income high achievers should take full advantage of their newfound authority under the Every Student Succeeds Act (ESSA) to ensure that their schools have ample incentives to educate those children, and all children, to the max.

Mindful of both the challenges the country faces and the new opportunity that state leaders have to set matters right, the analysis in High Stakes for High Achievers: State Accountability in the Age of ESSA does two things. First, it advances specific ideas for how state accountability systems can be redesigned to demand strong performance and growth from high-achieving students while meeting the requirements set forth in ESSA. Second, it rates current (or proposed) accountability systems in the fifty states and the District of Columbia based on how well they draw attention to high achievers. The evidence, regrettably, is that few of them are doing it well today, which is to say that few now incentivize their public schools to pay attention to learning gains among those who have already cleared the proficiency hurdle. 

Arkansas, Ohio, Oregon, and South Carolina are the only states that can be considered leaders on this issue. By contrast, just four states base a majority of schools’ summative ratings on "growth for all students," though this is the best way to evaluate a school’s effect on student achievement. Seven states and the District of Columbia assign no weight to this measure. Only five states treat high-achieving students as a subgroup and separately report their results at the school level.

On a slightly less negative note, fourteen states and D.C. rate or plan to rate schools’ achievement using a model, such as a performance index, that gives additional credit for students achieving at an “advanced” level. That’s good. But it’s not much.

The problem is sizable, in other words, but the opportunity to solve it is at hand. The report offers a number of recommendations to state policy makers, who have the opportunity to dramatically upgrade their current accountability systems. The authors also offer one recommendation to the Department of Education, which is finalizing its ESSA regulations: Going forward, Washington should allow states to rate academic achievement using a performance index that gives schools additional credit for getting students to an advanced level.

The impetus for the analysis was an explicit desire to influence these policy makers in the short term. Yes, much of what was unearthed about state accountability systems could be out of date within a year’s time. But that same year offers state leaders a rare opportunity to do things differently and better. Many issues will be debated as states design their new accountability systems. The hope is that the educational needs of high-achieving students get the attention they deserve—and that they didn’t get in the NCLB era.

Let us say to educators and policy makers who are already retooling their state accountability systems: Those children are counting on you. Their futures, and the nation’s, depend in no small part on the decisions you are making.

A new analysis from the Manhattan Institute’s Scott Winship shows that, measured correctly, the U.S. child poverty rate declined from 13.1 percent in 1996 to 7.8 percent in 2014—a drop of almost two-fifths.

This has huge implications for many policy areas, including education reform, and it’s a development that all parties must wrestle with.

For the teachers’ unions and other traditional education groups, it raises hard questions about their familiar contention that America’s lackluster student achievement is due to poverty—that we must “fix poverty first” before our schools will improve. We haven’t fixed poverty, but we have most certainly decreased it.

It also raises hard questions for reformers about why we haven’t seen greater progress in student outcomes over the past twenty years, considering that these socioeconomic trends should put the wind at our backs. We like to point to achievement gains in the late 1990s and early 2000s, especially for the poorest and lowest-performing students, as evidence that testing and accountability boosted learning.

But what if that was only part of the story?

Let’s excavate a bit deeper. Ramesh Ponnuru has a great overview of Winship’s study at Bloomberg View. Ponnuru writes, “The Census Bureau’s official measurements, it is true, show only modest improvement in child poverty since 1996. Winship argues that a more accurate measure of trends should take account of non-cash benefits that poor households receive, such as food stamps. We should also place some economic value on these households’ health care benefits. He makes a strong case that the inflation measures of the Congressional Budget Office and the Federal Reserve are superior to the one used by the Census Bureau. Because the Census Bureau overstates inflation, it understates improvements in poor people’s purchasing power.”

I’m far from an expert on welfare policy—or inflation--but Winship’s adjustments strike me as reasonable. They also imply, as Ponnuru acknowledges, that low-income Americans would be markedly worse off without social programs (like food stamps and Obamacare) that many Republicans and conservatives have opposed.

Let’s also acknowledge that plenty of people are suffering even when their incomes and benefits put them slightly over the poverty line. If you’re an urban single mom with two kids making $20,000 per year, you are struggling to make ends meet—even if you’re getting food stamps, and even if you’re not officially poor. I quite literally cannot imagine how they do it.

Nor is there any doubt, regardless of these promising trends, that a strong link remains between students’ socioeconomic status and achievement. If anything, the association has grown more robust over the past twenty years, owing to increased income inequality. With the upper middle class growing larger and richer, their kids (OK, our kids) appear to be pulling further away from everybody else’s. In part, that’s due to (mostly private) spending on “enrichment,” but it also has to do with the intensive parenting being practiced today by college graduates.

Still. Those of us engaged in policy debates have a responsibility to speak and write with precision. If the real child poverty rate is anywhere near 7.8 percent, everyone needs to stop pointing to “poverty” as the cause of our educational underachievement. Less than ten percent of the student population can’t possibly be responsible for our disappointing PISA and TIMSS scores. (In fact, American performance is mediocre at every point along the socioeconomic spectrum.) Nor can it be true, as some newspapers have irresponsibly reported, that “more than half” of America’s students are now living in poverty. (A majority are eligible for subsidized lunches, but nobody these days views that as a precise gauge of family income.)

Yes, the United States remains a stratified society with a widening gap between the haves and have-nots, mostly because of big gains for the top 30 percent. But let’s not overlook the good news: We are doing better by our poorest citizens, including our poorest children, than we were 20 years ago. And we should expect them to be doing better in school as a result.

The discussion we should be having is whether the gains we’ve seen in student achievement are commensurate with the gains apparent in children’s other life experiences.

Editor’s note: This article was originally published in a slightly different form in the 74.

A recent study published by Johns Hopkins’s Institute for Education Policy sets out to uncover how many elementary and middle school students are performing one or more years above grade level. The authors undertake this study to challenge the current education policy focus on achieving grade-level proficiency without accounting for students who perform above grade level.

To answer this broad question, the study examines data from state, multi-state, and national level assessment datasets (five in all). At the state level, the authors delve into data from three assessments: Smarter Balanced in Wisconsin and California and the Florida Standards Assessment in grades 3–8.

In evaluating all three state-level datasets against the states’ respective measures of grade-level proficiency, the authors found significant percentages of students scoring at or above grade level in the spring of their current grade level. In Wisconsin, 25–45 percent of students in grades 3–8 scored at or above grade level. For the same set of grades, 11–37 percent of California students scored at or above grade level. Florida features the highest percentage of students performing at or above grade level, at 30–44 percent for ELA (grades 3–9) and mathematics (grades 3–7).

Turning to the Northwest Evaluation Association’s Measures of Academic Progress (MAP) multi-state assessment, the authors looked at what percentage of fifth graders were already achieving at fifth-grade end-of-year proficiency levels in reading and mathematics at the beginning of their school year. They also measured how many were way ahead, performing at eighth-grade end-of-year proficiency levels. In math, 14 percent of the fifth-grade students started the year in the first group (achieving fifth-grade end-of-year scores), and 2.4 percent were in the second (at an eighth-grade elevation). In reading, those figures were 35 percent and 10 percent respectively.

Finally, the authors utilize nationally representative data from the National Assessment of Educational Progress collected in grades four, eight, and twelve to examine how many fourth-grade students outscored their eighth-grade counterparts. The authors found that over 5 percent of fourth-grade students outscored eighth-grade students in mathematics, while almost 10 percent of fourth-grade students outperformed the bottom half of eighth-grade students in reading.

On the basis of this diverse pool of data, the authors conclude that a relatively high number of American elementary and middle school students perform above grade level—approximately 20–40 percent in reading and 11–30 in math. They also note that K–12 students performing above grade level fall outside of the primary focus of many federal and state education policies, which are frequently aimed at achieving grade-level proficiency. To prevent these students from continuing to fly below the radar, the authors recommend that states require districts and schools to report the numbers and percentages of students performing above grade level. This would constitute a first step toward heightened transparency and accountability in meeting the learning needs of these students. Our recently released report High Stakes for High Achievers: State Accountability in the Age of ESSA provides additional actionable recommendations for upgrading current accountability systems to better attend to the needs of students performing above grade level.

This study provides a needed reminder that a “one-size-fits-all” approach to education policy cannot successfully meet the needs of all students—and that students performing one (or more) years above grade level are equally deserving of academic support as those working to achieve grade-level proficiency.

SOURCE: Matthew C. Makel, Michael S. Matthews, Scott J. Peters, Karen Rambo-Hernandez, and Jonathan A. Plucker, “How Can So Many Students Be Invisible? Large Percentages of American Students Perform Above Grade Level,” Johns Hopkins Institute for Education Policy (August 2016). 

Although recent analyses show that the child poverty rate isn't as high as many people believe, the fact remains that millions of American students attend under-resourced schools. For many of these children, well-resourced schools are geographically close but practically out of reach; high home prices and the scarcity of open enrollment policies make it all but impossible for low-income families to cross district borders for a better education.

Some research shows that low-income children benefit from attending school with better-off peers. Middle- and upper-income children may also benefit from an economically diverse setting. In short, income integration is a win-win for everyone involved. So why do the vast majority of school districts in the United States remain segregated by income? The answer isn’t much of a mystery: Schools are mainly funded by locally raised property taxes, which functionally “give wealthier communities permission to keep their resources away from the neediest schools.”

In order to examine just how isolating school district borders can be for low-income students, a relatively new nonprofit called EdBuild recently examined 33,500 school borders for school districts in 2014 and identified the difference in childhood poverty rates between districts on either side of the boundary line. (For poverty rates, the report uses the Census Bureau’s Small Area Income and Poverty Estimates, also know as SAIPE.) While a typical school district border has a student poverty rate difference of seven percentage points, EdBuild identified fifty school district borders where the difference ranged from thirty-four to forty-two percentage points. These fifty districts are located in just fourteen states, and Ohio claims nine spots in the top fifty—more than any other state.

The top four most segregated borders are between: 1) Detroit and Grosse Pointe, Michigan; 2) Birmingham and Vestavia Hills, Alabama 3) Birmingham and Mountain Brook, Alabama; and 4) Clairton and West Jefferson Hills, Pennsylvania. Clocking in at number five is Fordham’s hometown of Dayton, which has a 40.7 percent difference in student poverty rate from neighboring Beavercreek. Dayton shows up again in the seventh spot with a 40.3 percent difference from neighboring Oakwood. An interesting caveat to Dayton’s story is that Ohio actually has an inter-district open enrollment program. State law permits kids in Dayton—many of whom are enrolled in schools that aren’t just under-resourced but also academically failing—to enroll in one of the surrounding suburban school districts, but only if the receiving district chooses to allow it. Unfortunately, though not surprisingly, both Beavercreek and Oakwood have declined to accept open enrollment students.

The overall picture presented by the fifty most disparate district borders is bleak. For instance, the poorer of these districts have an average poverty rate of 46 percent compared to their wealthier neighbors’ average of just 9 percent. The average home in the affluent districts is worth approximately $131,000 more than the average home in the neighboring high-poverty district; as a result, wealthy districts are able to generate more local funds via property taxes—about $4,500 more per student. This disparity exists despite the fact that several high-poverty districts tax themselves at a higher rate than their affluent neighbors.

EdBuild’s report doesn’t offer any specific policy recommendations. However, in a related piece in the Atlantic, the organization’s CEO, Rebecca Sibilia, calls for decreasing the importance of district boundaries by “creating a larger tax pool that can fairly resource schools.” These ideas are undoubtedly horrifying to defenders of the public district monopoly and champions of so-called local control, but EdBuild’s report already offers its response to such opposition: “School district boundaries have become the new status quo for separate but unequal. It’s time to rethink the system.”

SOURCE: “Fault Lines: America’s Most Segregating School District Borders,” EdBuild, (August 2016).

A recent American Enterprise Institute study dispels myths about charter schools by comparing them to nearby district schools in a few novel ways.

Author Nat Malkus gathered data on school type, locale, enrollment, proficiency, discipline rates, demographics, and the number of English language learners and special education students they serve. Sources included the National Center for Education Statistics, the Department of Education’s Office of Civil Rights, and EDfacts.

Instead of looking at large groups of charter and districts schools across the country or a state, as charter opponents are wont to do, Malkus compares each charter school to five neighboring district schools that a given charter student might otherwise attend. Obviously, this makes for much more of an apples-to-apples comparison.

A recurring theme throughout virtually all of Malkus’s analyses is the great amount of variance between charter schools. He compares randomly selected district schools, which he terms “reference schools,” to five neighboring district schools, just as he did with the charters. Through the study’s various lenses—school discipline, student enrollment, achievement, or something else—charter schools are repeatedly shown to differ more from one another than district schools do. (There is also more variance between charters than between charter schools and their matched district schools.)

This is borne out by the report’s important findings on charter school discipline. Contrary to popular belief, the charters he studied weren’t plagued by significantly higher suspension rates than nearby district schools. Indeed, regardless of whether one looks at schools with rates that are high, low, or middling nationally, the difference in suspension rates between sectors was no greater than three percentage points. Moreover, charter schools were twice as likely as reference district schools to exhibit comparatively low out-of-school suspension rates. They were also, however, twice as likely to claim comparatively high rates—a prime example of the aforementioned variance.

Similar patterns appeared in student achievement and demographics. Compared to neighborhood district schools, 23 percent of charters had substantially lower rates of proficiency, while 28 percent had substantially higher rates. Likewise, 12 percent of charter schools had substantially fewer white students than nearby schools, and 16 percent had substantially more.

These findings also buck another accusation commonly thrown at charter schools—that they “cream skim.” As Malkus writes, “If charters were generally cream skimming, we would expect to see more uniform differences, with charters having fewer poor black and low-performing students and more white, non-poor, and high-performing students.”

The one area that might give charter proponents pause is special education enrollment. Malkus finds that charter schools do tend to enroll fewer such students than neighboring district schools. For example, 50 percent of the charter schools he examined had a special education enrollment of less than 10 percent, whereas only 24 percent of the neighboring district schools did. And at the other end of the spectrum, only 13 percent of charters had special education enrollment rates of 20 percent or more—seven percentage points less than nearby district schools.

Malkus nevertheless denies that this is evidence of skimming, citing both insufficient research to support that assertion and studies that have found other factors to be responsible for the gap (such as parents of students with disabilities being less likely to apply to charters in the first place).

Overall, the study’s greatest strength is its demonstration of the pitfalls of unsophisticated charter-district comparisons. Charters are too diverse to be compared as a unit, and their settings tend to be too urban to be compared to all district schools. As the author suggests, future studies of charters should move away from more broad generalizations and account for the diversity in charter school options that they were intentionally created to provide.

SOURCE: Nat Malkus, “Differences on Balance: National Comparisons of Charter and Traditional Public Schools,” American Enterprise Institute (August 2016).