Arguably, the most important development in K–12 education over the past decade has been the emergence of a growing number of urban schools that have been convincingly shown to have dramatic positive effects on the achievement of disadvantaged students. Those with the strongest evidence of success are oversubscribed charter schools. These schools hold admissions lotteries, which enable researchers to compare the subsequent test-score performance of students who enroll to that of similar students not given the same opportunity. Through careful study of the most effective of these charter schools, researchers have identified common practices—a longer school day and year, regular coaching to improve teacher performance, routine use of data to inform instruction, a culture of high expectations—that have yielded promising results when replicated in district schools.

We have only a limited understanding of how these practices translate into higher academic achievement, however. It may be that attending a school that employs them enhances those basic cognitive skills—such as processing speed, working memory, and reasoning—that research in psychological science has shown contribute to success in the classroom and later in life. Do schools that succeed in raising test scores do so by improving their students’ underlying cognitive capacities? Or do effective schools help their students achieve at higher levels than would be predicted based on measures of cognitive ability alone?

To address this question, we draw on unique data from a sample of more than 1,300 8th graders attending 32 public schools in Boston, including traditional public schools, exam schools that admit only the city’s most academically talented students, and charter schools. In addition to the state test scores typically used by education researchers, we also gathered several measures of the cognitive abilities psychologists refer to as fluid cognitive skills. Our data confirm that the latter are powerful predictors of students’ academic performance as measured by standardized tests.

Yet while the schools in our sample vary widely in their success in raising test scores, with oversubscribed charter schools in particular demonstrating clear positive results, we find that attending a school that produces strong test-score gains does not improve students’ fluid cognitive skills. Put differently, our evidence indicates that effective schools help their students achieve at higher levels than expected based on their fluid cognitive skills. It also suggests that developing school-based strategies to raise those skills could be an important next step in helping schools to provide even greater benefits for their students.

Crystallized Knowledge and Fluid Cognitive Skills

Despite decades of relying on standardized test scores to assess and guide education policy and practice, surprisingly little work has been done to connect these measures of learning with the measures developed over a century of research by cognitive psychologists studying individual differences in cognition. Psychologists now consider cognitive ability (few dare say “intelligence” anymore) to have two primary components: crystallized knowledge and fluid cognitive skills. Crystallized knowledge comprises acquired knowledge such as vocabulary and arithmetic, while fluid skills are the abstract-reasoning capabilities needed to solve novel problems (such as the ability to identify patterns and make extrapolations) independent of how much factual knowledge has been acquired. The terms were coined by the late psychologist Raymond Cattell, who first distinguished two types of intelligence. Cattell noted that one “has the ‘fluid’ quality of being directable at almost any problem,” while the other “is invested in particular areas of crystallized skills which can be upset individually without affecting others.”

Hundreds of studies show that, at any point in time, the two are highly correlated: people with strong fluid cognitive skills are at an advantage when it comes to accumulating the kinds of crystallized knowledge assessed by most standardized tests.

That these capabilities are nonetheless distinct is best illustrated by the fact that fluid cognitive skills decline with age starting even in one’s twenties, while crystallized knowledge tends to rise over the decades, in some cases peaking as late as one’s seventies. In an influential 2002 study involving people ages 20 to 92, University of Texas at Dallas psychologist Denise Park and colleagues found that the fluid cognitive skills of participants in their twenties exceeded those of participants in their seventies by as much as 1.5 standard deviations. In other words, more than 90 percent of participants in their twenties had higher fluid cognitive skills than did typical participants in their seventies. Those in their seventies nonetheless scored higher than participants in any other age range on tests of vocabulary, a key component of crystallized knowledge.

At a more fine-grained level, cognitive psychologists have identified multiple aspects of fluid cognition, including processing speed (how efficiently information can be processed), working memory (how much information can be simultaneously processed and maintained in mind), and fluid reasoning (how well novel problems can be solved). Longitudinal studies tracking individuals from late childhood through young adulthood indicate that gains in processing speed support gains in working memory capacity that, in turn, support fluid reasoning. Each of these abilities has been shown to be associated with academic performance, suggesting that they promote or constrain learning in school.

The strength of the relationship between fluid cognitive skills and academic performance also suggests that schools that are particularly effective in improving standardized test scores may do so by improving fluid cognition along one or more of these dimensions. This is what our research sought to explore.

Data and Sample

We gathered the data for our study during the spring of 2011 from 32 of the 49 public schools in Boston that serve 8th-grade students. The schools that agreed to participate in the study included 22 open-enrollment district schools, five oversubscribed charter schools, two exam schools to which students are admitted based on their grades and standardized test scores, and three charter schools that were not oversubscribed at the time the 8th-grade students in our study were admitted. Boston’s oversubscribed charter schools are of particular interest, as multiple studies have exploited the lottery admissions process to document the schools’ effectiveness in raising student test scores (see “Boston and the Charter School Cap,” features, Winter 2014).

Within those schools, we collected data on all students for whom we obtained parental consent for participation and who were in attendance on the day we collected data. These 1,367 students represent 43 percent of all 8th-grade students attending public schools in Boston and 64 percent of the students in participating schools. Seventy-seven percent of the students in our sample are from low-income families, 38 percent are African American, and 39 percent are Hispanic, in each case closely matching the demographic composition of all 8th-grade students attending public schools in the city and 8th graders attending the same schools.

The fluid cognitive skills we measured for each student included processing speed, working memory, and fluid reasoning. For processing speed, students were asked to translate numbers into corresponding symbols using a number-symbol key, and to indicate as quickly as possible under a time constraint whether either of two symbols on the left side of a page matched any of five symbols on the right side. For working memory, students viewed an array of blue circles, blue triangles, and red circles, and were instructed to count the number of blue circles within 4.5 seconds. After viewing between one and six arrays, they were prompted to record the number of blue circles contained in each. Finally, the fluid-reasoning task required students to choose which of six pictures completed the missing piece of a series of puzzles that became progressively more difficult. Because these three measures are closely related in theory and were positively correlated among the students in our sample, we also averaged them to create a summary measure of students’ fluid cognitive ability.

Fluid Cognitive Skills Predict Test Scores

Our first step is to examine the relationship between our measures of fluid cognitive skills and scores from the state’s standardized tests. We look at the students’ scores on the Massachusetts Comprehensive Assessment System (MCAS) tests in math and reading (ELA) and improvements in those test scores over time. We use simple correlation coefficients to measure the strength of the relationship between fluid cognitive skills and test scores. Correlation coefficients can range from -1 to 1, with a correlation of 0 indicating that there is no linear relationship between the two variables in question.

The correlations between our measures of fluid cognitive skills and 8th-grade math test scores are positive and statistically significant, ranging from 0.27 for working memory to 0.53 for fluid reasoning. The correlation between math test scores and our summary measure of fluid cognitive ability is 0.58, which implies that differences in fluid cognitive skills can account for more than one-third of the total variation in math achievement. The relationships are somewhat weaker for test scores in reading. Even so, variation in our summary measure of fluid cognitive ability can explain as much as 16 percent of the total variation in reading achievement.

Fluid cognitive skills are also related to the rate at which students improve their test-score performance over time. To measure gains in student achievement, we calculate the difference between 8th-grade performance in each subject and the performance level that would have been expected based on performance in both subjects in 4th grade. The correlations between our summary measure of fluid cognitive ability and test-score gains in math and reading were 0.32 and 0.18, respectively.

A high degree of correlation between measures of fluid cognitive skills and test scores is not news. As noted above, fluid cognitive ability has a long track record of predicting how much students know and are able to do. Our findings do suggest, however, that the specific measures of fluid cognitive skills we administered in classrooms as part our research were able to capture academically relevant differences in student cognition.

Schools Improve Test Scores but Not Fluid Skills

We address our central question of whether schools that raise student test scores also improve fluid cognitive skills in two complementary ways. First, we use our entire sample to analyze the extent to which the schools that students attend can explain the overall variation in student test scores and fluid cognitive skills, controlling for differences in prior achievement and student demographic characteristics (including gender, age, race/ethnicity, and whether the student is from a low-income family, is an English language learner, or is enrolled in special education). Second, we focus on the subset of students who entered the admissions lottery at one of the five oversubscribed charter schools in order to study how attending one of those schools affected test scores and fluid cognitive skills.

Consistent with other research on school effects, we find that the school a student attends can explain a substantial share of the overall variation in test scores: that single factor explains 34 percent of the variation in math scores and 24 percent of the variation for reading. In contrast, after accounting for prior achievement and demographics, the school attended explains just 2.3 percent of our summary measure of fluid cognitive ability.

This pattern suggests that schools may influence students’ test scores but not by affecting their fluid cognitive skills. However, this analysis does not account for the possible sorting of students into particular schools based on characteristics not captured by their prior achievement and demographic characteristics. Such “selection effects” could in theory account for the apparent school impacts on test scores, or even the apparent absence of impacts on fluid cognitive skills.

Our second analysis aims to address this concern. Because the oversubscribed charter schools in our sample admit students via random lotteries, comparing the outcomes of lottery winners (most of whom enrolled in a charter school) and lottery losers (most of whom did not) is akin to a randomized-control trial of the kind often used in medical research. Evaluations led by Harvard’s Tom Kane and MIT’s Josh Angrist have used this lottery-based method to convince most skeptics that the impressive test-score performance of the Boston charter sector reflects real differences in school quality rather than the types of students charter schools serve.

Due to the limited coverage of our sample, we cannot claim for our analysis the same level of rigor as these previous lottery-based evaluations. Of the roughly 700 applicants for the lotteries used to admit students in the 8th-grade cohort in our study, only 200 of them are in our evaluation sample. Focusing on lottery applicants is nonetheless useful because it enables us to hold constant whatever unmeasured differences lead some students to apply for a seat in a charter school and others to remain within the district. When comparing lottery winners and losers, we also control for prior achievement and the same set of demographic characteristics used in our broader analysis. We use standard methods to account for the fact that not all lottery winners enrolled in a charter school and remained there throughout middle school (and some lottery losers eventually obtained a seat). This approach enables us to generate estimates of the effect of each additional year of actual attendance at a charter school between 5th and 8th grade.

Our results show that each year of attendance at an oversubscribed Boston charter school increases the math test scores of students in our sample by 13 percent of a standard deviation. This is a noteworthy effect, equivalent to roughly a 50 percent increase in the academic progress students typically make in a school year (see Figure 1). Charter school attendance also appears to have a modest positive effect on reading scores, though this estimate falls short of statistical significance due to the relatively small number of students in our lottery sample. Even as students benefit academically, however, their fluid cognitive skills hardly budge. The estimated effect of charter school attendance for each of our measures is very small in magnitude; none is statistically significant.

Are Test-Score Gains “Real”?

There is ample reason to believe that the test-score gains generated by these schools are meaningful, despite the lack of corresponding improvement in fluid cognition. State tests are aligned to standards that specify the knowledge and capabilities students are expected to acquire—the very things cognitive psychologists call crystallized knowledge. And there is strong evidence that crystallized knowledge, which also bears a strong resemblance to E. D. Hirsch’s notion of Core Knowledge, matters a great deal for success in school and beyond. Recent studies by Harvard economist Raj Chetty and colleagues confirm that teachers who improve student test scores also improve their students’ earnings as adults (see “Great Teaching,” research, Summer 2012). Moreover, lottery-based evaluations of the Boston charter sector show that attending high schools affiliated with three of the charter schools in our sample increases Advanced Placement test-taking and performance and the likelihood of attending a four-year college.

Indeed, in our view, the unique data we gathered for this study make these schools’ accomplishments all the more impressive. They show that the schools that are most effective in raising student test scores do so in spite of the strength of the underlying relationship between math achievement and fluid cognitive skills. In other words, these schools have figured out ways to raise students’ academic achievement well above what is expected given the students’ baseline fluid cognitive skills.

A compelling way to see this is to look at the relationship across schools between the average test-score gain students make between the 4th and 8th grade and our summary measure of their students’ fluid cognitive ability at the end of that period (see Figure 2). Each dot represents a school, and the diagonal line shows the overall relationship between test-score gains and fluid cognitive ability across the full sample of schools. The extent to which a school is above or below that line indicates whether the average test-score improvement among its students has been greater or less than would be predicted based on their fluid cognitive skills.

Most schools fall relatively close to the regression line, indicating that their students’ academic progress is roughly as expected given the students’ fluid cognitive skills, but there are clear exceptions. Most notably, each of the five oversubscribed charter schools is well above the regression line. A few open-enrollment district schools also show the ability to drive similarly outsized gains, an important reminder that while governance matters, what counts in the end is effective practice. Finally, while exam-school students have considerably higher fluid cognitive skills (as would be expected of students who gain admission via test scores and grades), attending one of these locally renowned schools in the company of other bright students confers no systematic advantage. This last finding is consistent with recent evidence showing no academic benefits of attending a Boston or New York City exam school for students who just met the admissions criteria (see “Exam Schools from the Inside,” features, Fall 2012).

What do these differences in school performance mean in layman’s terms? Among students who fell below the midway point on our summary measure of fluid cognitive ability, only 20 percent of those attending a district school were deemed proficient in math as defined by Massachusetts on its 8th-grade math test. In oversubscribed charter schools, 71 percent of such students were deemed proficient. This is a remarkable difference for students who rank lower than their peers on a key enabling capacity. For district students, success is the rare exception (2 in 10), while for oversubscribed charter school students, it is closer to the rule (7 out of 10).

At the same time, fluid cognitive skills remain potent predictors of academic progress even among students attending oversubscribed charter schools. While these schools succeed in generating test-score gains for students of all cognitive abilities, it is still the case that students with strong fluid cognitive skills learn more. Indeed, the strength of the correlation between fluid cognitive skills and test-score growth in oversubscribed charter schools is statistically indistinguishable from the correlations we observe among students in open-enrollment district schools and exam schools.

Could Schools Boost Fluid Cognitive Skills, Too?

Our research sought to examine whether schools that have demonstrated success in raising test scores also boost students’ fluid cognitive skills—either as a byproduct or perhaps as a principal pathway for improvements in test scores. That turns out not to be the case. This result does not, in our view, call into question the value of the improvements in crystallized knowledge captured by improvements in test scores.

What we do not yet know, however, is which long-term outcomes are more strongly influenced by fluid cognitive skills and which by crystallized knowledge. One reason is that, as we see in our study sample, fluid cognitive skills and crystallized knowledge tend to be highly correlated. In fact, it may be accurate to say that schools like the most effective schools in our study may be the first to produce students for whom these two types of cognitive ability are consistently decoupled, providing an opportunity to study just which kinds of outcomes are enabled by gains in crystallized knowledge alone. For example, it is possible that the oft-discussed challenges some students from high-performing urban schools experience in college (see “‘No Excuses’ Kids Go to College,” features, Spring 2013) stem in part from deficits in fluid cognitive skills.

Indeed, perhaps the most important implication that we draw is that educators seeking to innovate should get about the business of developing and rigorously testing the effects of interventions to raise these fluid cognitive skills. Improved abstract-reasoning capacity likely has important benefits in its own right and is highly related to important skills such as reading comprehension. Deficits in students’ fluid cognitive skills may also prevent even the most effective schools from raising all of their students’ academic performance to the desired level.

The question of whether processing speed, working memory, and fluid reasoning skills can be developed through intentional efforts is an area of active debate among cognitive psychologists. Several researchers have published studies claiming that they have improved these skills through deliberate practice aimed at one or more of these skills and, in a few cases, have shown that such improvements have translated into gains in other, broader measures of cognitive ability. None of these interventions has yet been shown to improve long-term outcomes such as college completion or earnings, however, and other researchers have failed to replicate even the narrower impacts that have been reported. Meanwhile, private companies such as Lumosity are aggressively marketing software-based training programs derived from this line of research to the general public as “brain training.”

This is a perfect time for cognitive psychologists, educators, and perhaps even game and software developers to join forces in rapid-cycle experimentation to explore whether and how schools can broadly and permanently raise students’ fluid cognitive skills. Successful schools have demonstrated their ability to dramatically increase crystallized knowledge and thereby raise test scores, improving other important student outcomes in the process. Boosting fluid cognitive skills might have an equally profound impact on students’ academic and life outcomes.

Martin West is an associate professor at the Harvard Graduate School of Education (HGSE) and deputy director of the Program on Education Policy and Governance at the Harvard Kennedy School. Christopher Gabrieli is adjunct lecturer at HGSE and executive chairman of the National Center on Time & Learning. Matthew Kraft is assistant professor of education at Brown University. Amy Finn is a postdoctoral fellow at the Massachusetts Institute of Technology, where John Gabrieli is professor of health sciences and technology and cognitive neuroscience.

This article is based on a study published in the March 2014 issue of Psychological Science.

The Knowledge Is Power Program (KIPP) is a network of charter schools designed to improve the educational opportunities available to low-income families. KIPP schools seek to boost their students’ academic achievement and ultimately prepare them to enroll and succeed in college. To achieve these objectives, KIPP schools leverage strong student-behavior policies with rewards and sanctions; contracts between students, parents, and teachers; longer school days and school on Saturdays; substantial autonomy for principals; and close monitoring of school performance in terms of student achievement and college readiness. KIPP has grown from two middle schools established in the mid-1990s to a nationwide network of more than 140 elementary, middle, and high schools in 20 states and the District of Columbia in 2014.

The majority of KIPP schools are middle schools that serve grades 5 through 8. Several rigorous studies have confirmed that many KIPP middle schools have large, positive impacts on student performance on math and reading tests. Our own study (the largest rigorous study of KIPP) estimated that over three years KIPP middle schools have an average cumulative impact of 0.21 standard deviations in reading and 0.36 standard deviations in math, roughly equivalent to an additional eight to 11 months of learning. KIPP’s ability to operate a large number of high-performing schools has attracted the attention of policymakers, including Secretary of Education Arne Duncan, who notes that “the KIPP network has debunked the myth that great schools are one-offs that cannot be replicated.” In just the past five years, KIPP has received tens of millions of dollars in federal grants to open new schools, in addition to a $50 million grant through the Investing in Innovation Fund to develop its school-leader training program.

Not everyone is convinced, however. Skeptics have asked whether attrition of students from KIPP schools may explain their apparent success and thereby raised doubts about the prospects for replicating that success at scale. For example, commentator Richard Kahlenberg has argued that “the big difference between KIPP and regular public schools…is that whereas struggling students come and go at regular schools, at KIPP, students leave but very few new students enter. Having few new entering students is an enormous advantage not only because low-scoring transfer students are kept out but also because in later grades, KIPP students are surrounded only by successful peers….”

Critiques like this do not deny that KIPP schools improve the achievement of the students who attend them, but rather argue that these improvements reflect advantageous enrollment patterns at KIPP that are not possible at traditional public schools. The key question is whether KIPP’s positive effects on learning are attributable to a peer environment that is more conducive to academic achievement than the peer environment found in traditional public schools. If so, traditional public schools may find it difficult to achieve comparable impacts by replicating the KIPP model.

Little systematic evidence exists on this question for the charter sector in general, much less for KIPP schools in particular. In this study, we use detailed student-level data to compare patterns of entry, attrition, and replacement in 19 KIPP middle schools and in traditional public middle schools in the districts in which the KIPP schools are located.

We find that, on average, KIPP middle schools admit students who are similar to those in other local schools, and patterns of student attrition are typically no different at KIPP than at nearby public middle schools. In both groups of schools, students who leave before completing middle school are substantially lower-achieving than those who remain. KIPP schools replace fewer of these students in the last two years of middle school, however, and, compared to district schools, KIPP schools tend to replace those who leave with higher-achieving students. Nonetheless, while this difference in replacement patterns is noteworthy, it cannot account for KIPP’s overall impact on student achievement. In particular, the literature on peer effects suggests that KIPP’s student replacement pattern could produce only a small fraction of KIPP’s actual impact on student achievement. A large part of KIPP’s cumulative effect occurs in students’ first year of enrollment, before attrition and replacement could have any effect.

Data

We use data from 19 KIPP middle schools located in nine states and the District of Columbia for which we were able to obtain at least three years of complete administrative data that track students over time. These 19 middle schools constitute a little more than half of the 35 KIPP schools that opened their doors in 2005 or earlier. As is typical of KIPP middle schools, each of the schools in the sample serves students in grades 5–8. The key pieces of information for the study are test scores in reading and mathematics, demographic characteristics, and schools attended. These data include 7,143 students who enrolled in a KIPP school between school years 2001–02 and 2008–09 and 1,202,060 non-KIPP students enrolled in districts where a KIPP school is located.

The students who attend the KIPP schools in the sample are quite disadvantaged, both in absolute terms and relative to the school districts in which they are located. In the year prior to entering a KIPP school, 80 percent of the KIPP students are from low-income families, as measured by eligibility for free or reduced-price school breakfast and lunch (FRPL); 96 percent are either black or Hispanic; 7 percent are English language learners; and 7 percent receive special education services (see Figure 1a). Compared with students district-wide, KIPP students are more likely to be black or Hispanic and from low-income families but slightly less likely to be English language learners or in special education. The same basic pattern holds if we compare KIPP students to only those non-KIPP students who attended the local elementary schools that KIPP students attended, i.e., those that effectively serve as KIPP “feeder” schools.

In terms of academic performance, KIPP students’ achievement in grade 4 (before entering KIPP) is lower than the district average by 0.09 standard deviations in reading and by 0.08 standard deviations in math, or roughly one-quarter of a grade level in each subject. KIPP students’ grade 4 achievement is very similar, however, to that of non-KIPP students who attended the feeder elementary schools (see Figure 1b). In other words, KIPP students start out academically behind the average student in their local school district, but this appears to reflect the communities in which KIPP schools have decided to open. Neither comparison would seem to indicate that KIPP schools succeed by serving a select group of students, but skeptics contend that these kinds of overall comparisons obscure more subtle patterns of selection that emerge as students in these districts transfer out of and into different middle schools.

Students Leaving KIPP Schools

To compare attrition rates at KIPP middle schools with rates at other public middle schools, we identified students who left their original schools either during or immediately after each middle school grade. We considered school-specific grade ranges and disregarded transfers caused by a normal grade progression, such as a move from an elementary school at the end of 5th grade to a middle school in 6th grade. All transfers out of the district or to private schools were classified as attrition.

For this portion of the study, we compare KIPP students with two groups of district school students: all those attending non-KIPP middle schools in the same district, and a smaller comparison group of students attending middle schools that accepted an above-average number of students from the feeder elementary schools. In other words, if students from feeder elementary schools eventually attend five different non-KIPP middle schools, the two most commonly attended of these middle schools would be included in the “comparison” middle-school group.

We do not find a consistent pattern of differences in attrition rates between KIPP and district schools (see Figure 2). At KIPP schools, the rate of attrition tends to decline moderately over the course of middle school. KIPP’s grade 5 attrition rate of 16 percent declines to 13 percent in grade 6 and to 9 percent in grade 7. The grade-specific attrition rates in the comparison middle-school group do not decline, with average rates of 11, 13, and 14 percent, respectively. Attrition rates in the full district fall between those of the KIPP schools and the comparison middle schools in each grade.

Over the entire course of middle school, cumulative attrition rates—the percentage of students entering 5th grade who change schools before completing 8th grade—is similar at KIPP schools and district schools. On average, the cumulative attrition rate at KIPP is 34 percent, compared with 34 percent in the comparison middle-school group and 36 percent in the district as a whole.

Despite the similarity in overall attrition rates in KIPP and comparison middle schools, cumulative attrition rates for black and Hispanic students are substantially lower in KIPP schools than in comparison middle schools. The attrition rate for black students is 37 percent at KIPP vs. 44 percent in the comparison group, and the attrition rate for Hispanic students is 24 percent at KIPP vs. 29 percent in the comparison group. There do not appear to be any notable differences between KIPP and comparison middle schools in attrition rates for students eligible for a free or reduced-price lunch.

We also examine the characteristics of students who transfer early to determine the extent to which attrition may change the composition of students at KIPP and at the comparison middle schools. For both groups, students who leave early tend to have markedly lower grade 4 test scores than students who stay (see Figure 3). At KIPP, on average, students who leave score 0.25 and 0.22 standard deviations below the district-wide mean in math and reading at baseline (or at the 40th and the 41st percentiles, respectively); students who stay score 0.02 below the mean (the 49th percentile) in both subjects. In comparison middle schools, students who leave score 0.23 and 0.21 standard deviations below the mean in math and reading (or the 41st and the 42nd percentiles, respectively) at baseline; students who stay score 0.01 above the mean (the 50th percentile) in math and at the mean in reading.

The demographic composition of students transferring out of KIPP schools and out of comparison middle schools is also broadly similar. At both KIPP schools and comparison schools, students who leave early are significantly more likely to be black males and significantly less likely to have limited English proficiency than students who remain. Those who leave early from KIPP are also less likely to be Hispanic, whereas at comparison middle schools the proportion of Hispanic students who leave is nearly identical to the Hispanic proportion of students who remain enrolled. In addition, students who leave KIPP are no more likely than those who stay to be FRPL-eligible (both groups have high eligibility rates) or to be in special education, whereas students who leave comparison middle schools are more likely than those who stay to be FRPL-eligible and to participate in special education programs. Thus, if anything, it appears that the attrition from comparison middle schools is more likely than attrition from KIPP to draw away students who are disadvantaged relative to the students left behind.

Replacing Students Who Leave KIPP Schools

Attrition is not the only way that a school’s student body can change. Students who take the places of departing students represent another potentially important source of changes in a school’s overall enrollment. We define a late entrant as a student enrolling in a given school for the first time at any point other than that school’s normal entry grade. For example, in the case of a KIPP middle school serving grades 5 through 8, late entrants include students who enroll for the first time in grades 6, 7, or 8.

We examine two key issues related to late entrants. First, we calculate the number of late entrants relative to the number of students who leave the school early, in other words, the extent to which late entrants are admitted at all. Second, we examine whether the average achievement level of the new students is higher or lower than that of the students who enter the school in the normal entry grade.

Both KIPP and comparison middle schools admit late entrants, but at KIPP they are concentrated in earlier grades. All KIPP middle schools admit students after the normal entry grade, and most schools continue to admit at least some new students in every grade. The number of new enrollees declines substantially after grade 6, however. The KIPP schools in our study enroll an average of 13 new students per year in grade 6 (accounting for 18 percent of average total enrollment in that grade), 7 new students per year in grade 7 (12 percent of total enrollment), and 3 new students per year in grade 8 (6 percent of total enrollment).

KIPP schools admit more new students in grade 6 than the number of students who left through attrition in grade 5. Part of the explanation for this trend is the comparatively high rate of students who repeat grade 5 at KIPP schools. This trend reverses in later grades. In grades 7 and 8, KIPP schools admit slightly fewer new students than the number of students who left in the prior grade.

Comparison middle schools also consistently admit students after the normal entry grade. Unlike KIPP, however, these schools do not show any decline in new enrollment levels in grade 8 compared with grade 7. Rather, comparison middle schools consistently admit more new students, on average, than the number who exited through attrition in the prior grade. As a result, in grades 7 and 8, comparison middle schools are more likely than KIPP schools to replace the students who leave through attrition. Overall, however, the proportion of late entrants at KIPP (15 percent) is nearly identical to the overall proportion in the comparison middle schools (14 percent).

Although overall late-entry rates are similar, KIPP schools tend to differ from comparison middle schools in how those new students change the composition of the student body. Students who enroll late at KIPP tend to be higher-achieving than those who enroll on time, as measured by their grade 4 test scores, whereas the reverse is true at comparison middle schools (see Figure 3). On average, the 4th-grade math and reading test scores of KIPP late entrants were 0.15 to 0.16 standard deviations above the district average, putting them 0.19 standard deviations above the scores of students who enrolled in the normal intake grade. Conversely, late entrants at district schools had dramatically lower average 4th-grade test scores than on-time enrollees: 0.30 and 0.32 standard deviations lower in reading and math, respectively (in both cases, 0.29 standard deviations below the district average).

In addition to the substantial difference in test-score patterns, modest differences also exist between KIPP and comparison middle schools in the demographic characteristics of late entrants. Compared with on-time enrollees, KIPP’s late entrants are equally likely to qualify for FRPL and less likely (by 4 percentage points) to be in special education; in the comparison middle schools, in contrast, late entrants are significantly more likely (by 13 percentage points) to be FRPL-eligible and more likely (by 3 percentage points) to be in special education.

The Changing Composition of KIPP Students

Are the differences among late entrants large enough to notably change the types of students attending KIPP relative to those in district schools? To answer this question, we compare the characteristics of KIPP students in grades 5, 6, 7, and 8 with the characteristics of their peers across the district as well as of those non-KIPP students who attended the feeder elementary schools.

We find several noteworthy differences. Within KIPP schools, students in early grades have lower entering achievement levels, on average, than those in later grades, a pattern that is not evident at district schools. This is because those who left early and late entrants at district schools resemble each other—both groups of students tend to be relatively low-achieving. In contrast, a KIPP student can expect a large increase over time in the average grade 4 achievement of his peers. The increase in peer prior achievement from 5th to 8th grade at KIPP schools was 0.15 standard deviations greater in reading and 0.19 standard deviations greater in math than for students who attended feeder elementary schools (see Figure 4).

Unlike entering test scores, most demographic characteristics at KIPP schools remain stable throughout middle school. Throughout every grade of middle school, KIPP schools had a substantially higher proportion of blacks and Hispanics than district schools.

KIPP’s attrition and late-enrollment patterns do not appreciably change the proportion of students who were in special education, had limited English proficiency, or were FRPL-eligible when they entered. In district schools, the proportion of students with these attributes also remains similar in each of the grades we examined. In each grade of middle school, KIPP schools have significantly more FRPL-eligible students, significantly fewer special education students, and significantly fewer English language learners than district schools.

How Much Could Peer Effects Explain KIPP Impacts?

In light of research showing the positive impacts of KIPP schools on achievement, student transitions into and out of KIPP remain of interest. Policymakers supportive of KIPP wonder how much of the student population the KIPP network might grow to serve, and skeptics ask whether KIPP’s results depend on excluding students who are the most disadvantaged or the most difficult to teach, either in admissions or subsequent attrition.

Our findings provide a picture of KIPP’s student population from several key angles. In terms of prior achievement, KIPP schools generally admit students who are disadvantaged in ways similar to their peers in nearby district schools. These disadvantaged populations have high rates of educational mobility, but rates of exit from KIPP schools are typically no higher than rates at nearby district schools. Students leaving KIPP schools have similar prior achievement to those leaving nearby middle schools, but the late entrants at KIPP schools tend to have had higher achievement than late entrants at these other schools.

How much of KIPP’s impact on student achievement might be explained by the fact that KIPP tends to attract higher-achieving new students in the upper grades? One way to estimate the possible size of peer effects at KIPP is to combine our findings with other research on how peers’ prior scores affect student achievement. Unfortunately, published estimates of the effect of peer ability on student achievement range widely, from close to zero to nearly half a standard deviation impact for each standard deviation of difference in peer achievement. Even if the largest estimates of peer effects are correct, however, the improvement in peers’ prior test scores would appear to benefit KIPP students’ achievement only by about 0.07 to 0.09 standard deviations after four years at KIPP. KIPP’s cumulative impacts in middle school are three times that size, so even the largest estimates of the size of peer effects suggest that they are unlikely to explain more than one-third of the cumulative KIPP impact.

Moreover, the best available evidence shows that KIPP produces large impacts on students in their first year at a KIPP school—before late-entering students could possibly have any effect. Consequently, the true peer effect resulting from late entrants is likely to be substantially below the back-of-the-envelope estimate of 0.07 to 0.09 standard deviations.

A potentially important limitation of this study is that there could still be unmeasured differences between the students attracted to KIPP and those enrolling in other schools. We analyze the peer environment at KIPP as measured by demographic characteristics and prior achievement, but we do not have direct measures of parent characteristics, prior motivation, or student behavior. For example, KIPP students might benefit from attending school with peers who are especially motivated to accept KIPP’s academic and behavioral demands. If this were true, the presence of motivated peers at KIPP could help bolster the effectiveness of the KIPP model, and it would call into question whether KIPP’s approach would be equally effective in conventional public schools.

But the data available for this analysis clearly show that KIPP’s impacts cannot be explained by advantages in the prior achievement of KIPP students. Even when attrition and replacement throughout the middle school years are taken into account, the limited range of potential peer effects at KIPP schools does not explain the large cumulative impacts on student achievement identified by prior studies. One implication of these findings is that the KIPP model may include practices that could be effective outside schools of choice. Whether these practices can be replicated in traditional public schools or raise academic achievement across the full range of traditional public-school students remains to be seen.

Ira Nichols-Barrer is a researcher at Mathematica Policy Research, Inc., where Brian Gill and Philip Gleason are senior fellows and Christina Clark Tuttle is also a senior researcher.

Each year, as predictably as classes end in June, the School District of Philadelphia faces a budget crisis for the coming school year. In 2014, the School Reform Commission, the school district’s state-imposed governing body, for the first time and in violation of the city charter, refused to pass a budget, arguing that there were insufficient funds to run the schools responsibly. Philadelphia’s mayor Michael Nutter said, “It is a sad day in public service that we find children being held on the railroad tracks awaiting some rescue to come from somewhere.” And yet, casting the school children of Philadelphia in the Perils of Pauline has become a yearly ritual.

In the summer of 2013, the district superintendent announced that schools might not open on time, as severe budget cuts had led to insufficient staffing. The budgets of many individual schools allowed for no counselors, no secretaries to assist principals or answer telephones, and no arts or sports programs. With a last-minute financial-aid pledge from the city, some laid-off personnel were recalled, and schools opened on time. But the district was still in such dire straits that Philadelphia’s newspapers launched a drive to obtain pencils, paper, and other basic supplies. This is no way to run a school system, much less the eighth largest in the United States.

We investigate why these school crises keep recurring. The most recent spate of crises actually started in the 2012 school year  (herein school years are referenced by the spring in which the school year ended). Between 2011 and 2012, the district cut almost 17 percent of its workforce, including nearly 2,000 teachers. How did the district get into this mess? Some people blame managers for failing to look ahead and budget carefully, in particular, for failing to plan for the predictable end of federal stimulus funds. Structural factors have also been cited: It may be that the State of Pennsylvania does too little for public education. The City of Philadelphia may either tax itself too lightly or allocate too much money to other city services at the expense of the school district. Some of the financial burden may be the result of student flight to charter schools, whose share of Philadelphia students has grown dramatically in recent years. We find supporting evidence for all of these explanations. We also find that the crisis was certainly not due to excessive spending relative to that of surrounding districts (see sidebar titled Philadelphia Support for Education for details on city support, charter school enrollments, district spending, and teacher salaries). We conclude that the unwieldy process for financing the district mean that such crises are bound to recur unless that process is changed.

Funding the Philadelphia Schools

Three specific events triggered the 2012 crisis: an abrupt reduction in federal and state funding (see Figure 1), the inability of the district to cut many of its costs, and political pressures on the district to spend available revenues in a given year.

In response to an earlier financial crisis, in 2001, the State of Pennsylvania took control of the Philadelphia school district. The Republican governor at the time worked with the Democratic mayor of Philadelphia to arrange a “friendly” takeover, which replaced the local school board with a five-member School Reform Commission (SRC). Three members are appointed by the governor and two by the mayor.

With no independent taxing authority, the district depends for funding on annual allocations from the city, state, and federal governments. Of the district’s $2.7 billion in revenues in 2013, 50 percent came from the state and 14 percent from the federal government; city and local contributions made up the remainder. In a typical year, the governor presents a draft state budget to the legislature in February, including allocations for K–12 education. In April, the district uses the governor’s proposed budget and estimates what it will get from the city to prepare its own draft budget. Much of the funding the School District of Philadelphia gets from the city is fairly predictable, but it must nevertheless appear before the city council in May to present a request for funds.

From the perspective of the district, multiple sources of funding create three significant planning problems. First, the state legislature can alter the governor’s proposed budget. Second, the city council may not grant the district’s request for funds, especially any requests that go beyond its typical allocations. When the allocations from the state and city deviate from what the district expects, it must adjust its budget. Third, there is a “who goes first?” problem. In years when the district appeals to the state and city for additional funding, each is reluctant to pledge new funds without knowing how much the other will commit. If one governmental entity steps up, this reduces the burden on the laggard. Such gaming between the city and state can leave the district uncertain about its exact funding until well into its fiscal year.

When the recession began in late 2007, it severely reduced tax revenues at all levels of government. In order to preserve jobs and maintain educational spending, the federal government in early 2009 initiated an economic stimulus program. About $113 million a year in 2010 and 2011 came directly from the federal government to the school district, mainly as supplemental grants to enhance educational services for disadvantaged students. Other stimulus funds went to the state to allocate to school districts.

In 2010, the state faced reduced revenues and cut its own support for K–12 education but used the stimulus funds to increase overall allocations to school districts. For Philadelphia, combined state and federal funding grew by 10.2 percent. Local funding for the district declined by 0.7 percent, but given the large share of state and federal funding, overall district revenues still grew by 6.7 percent.

The subsequent fiscal year was similar, but the growth in revenues was more constrained. Funding to the district from state-generated revenues was essentially flat, but federal funding increased by 13.3 percent, largely due to the stimulus package. Local funding fell by 2 percent. On net, the district’s total revenues grew by 2 percent. This modest increase created budget pressures for the school district that required some midyear cuts, but the district managed to end the fiscal year with a small surplus.

District administrators clearly understood that the district’s budget was heavily dependent on the state’s basic education subsidy, and that the state had increased this subsidy in fiscal years 2010 and 2011 by using federal stimulus funding. In 2010 and early 2011, district management warned repeatedly that severe cuts would have to be made if the state or the city did not increase support for K–12 education once the federal stimulus funding ended in 2011. Arlene Ackerman, who became superintendent in 2008 but would be forced out in 2011, emphasized two factors that she thought might win over legislators who were demanding greater school choice and more accountability from traditional schools. Ackerman pointed to the district’s support for the growing numbers of students selecting charter schools and to rising test scores by district students on the state’s standardized tests.

Spending Cuts

Political developments did not provide much hope for additional state funding, however. In November 2010, voters elected Tom Corbett, a Republican, as governor, replacing Ed Rendell, a Democrat. In addition, the Republican Party gained control of both houses of the state legislature. Corbett and many Republican state legislators had run on a “no new tax” pledge. With the state still mired in the recession and the end of federal stimulus funding, Corbett clearly had to cut state spending to balance the budget.

The school district could not know exactly what Corbett planned to do with the education budget, but it recognized that a big increase in state funding to offset the elimination of federal stimulus funding was unlikely. Almost immediately after Corbett’s election, the School District of Philadelphia began to make plans for how it might adjust to major budget cuts. But the governor surprised the district (and school districts throughout the state) in March 2011 when he proposed a state budget that cut support for basic education by nearly 10 percent. The termination of almost $1 billion in federal stimulus funding for education in Pennsylvania was to be offset by only a very modest replacement by the state. After heavy lobbying by school districts around the state, Corbett and the state legislature settled on a 7 percent cut in the basic education subsidy, but they also made deep cuts to other state grants for education. Reductions in state funding fell disproportionately on Philadelphia. The district, which educates about 10 percent of the state’s children, shouldered about 30 percent of the state cuts.

For Philadelphia, the end of the stimulus and the state’s decision to offset only a small share of the federal cuts meant a 14 percent drop in state and federal education funding between 2011 and 2012. As the magnitude of the coming funding cuts materialized, the district appealed to the city for additional revenues to help it head off drastic disruptions to its educational program. The city responded by increasing its funding and other local revenue to the district by 11.3 percent. The net result, however, was an almost 7 percent decline in district revenue (see Figure 2).

Limited Options

The school district had to cut spending, but much of the district’s budget is fixed in the short run. Collective bargaining agreements set the scale for salaries and benefits. The federal government mandates certain levels of special education services. State law and enrollments govern how much the district must transfer to charter schools. School buildings must be heated and maintained. Bond issuances determine required debt payments. In fact, the district estimates that between 66 and 80 percent of its expenses are predetermined from one year to the next, and many of these costs, such as negotiated wage contracts, increase automatically. Thus, the 7 percent decline in total operating revenues in fiscal year 2012 fell heavily on a small portion of the overall budget.

The largest component of educational expenditure is personnel, so it is not surprising that the district looked to the workforce for potential cuts. Between 2003 and 2009, district employment of classroom teachers had declined, but the student-to-teacher ratio had held relatively stable as the district lost students to charters and other schools. In 2010 and 2011, the district added teachers while continuing to lose students, and student-to-teacher ratios dropped dramatically. But when the stimulus funding ended and the state failed to replace those funds, the district decided to cut its workforce by almost 17 percent (see Figure 3), which would save more than $300 million annually.

The district also netted one-time savings of well over $100 million, mainly from a debt refinancing, and it pushed some expenses into the subsequent fiscal year. It delayed $42 million in payments into a union health plan and borrowed $35 million from the regional transportation authority. Such measures simply postponed the need for additional cuts.

The tight financial situation continued into 2013, but the district avoided significant cuts that year by borrowing $302 million in the bond market. Much of this went to cover the short-term debts the district had incurred in the previous year. Recognizing that it could not continue to borrow to cover operating costs, the SRC announced that it would close almost 10 percent of the district’s 240 schools in 2013. For several years, the SRC had been closing schools, but well below the rate the schools were losing students. Every announced school closing was vociferously contested by supporters of that school.

In 2013 the relatively new superintendent, William R. Hite, announced that schools might not open on time if they could not be adequately staffed. Committed to a balanced budget, Hite and the SRC put forth a “doomsday” budget that severely cut the number of noontime aides, counselors, and teachers, and created “split” classrooms, that is, classrooms with two grade levels in the same room. That summer, the district mailed layoff notices to nearly 3,800 employees. At the same time, the SRC asked the city for $60 million in additional funding, the state for $120 million, and its unions for $133 million in labor concessions. The city pledged to find $50 million in additional revenue for the schools. Exactly how the city would come up with this money was still to be determined, but the pledge was sufficient for the district to reinstate 1,600 laid-off employees and open on schedule.

Could the Crisis Have Been Avoided?

Critics of the district management argue that the administration should have cut costs in order to conserve resources and prepare for the predictable end of the stimulus money, even when, relatively speaking, funding was flush. It certainly did not do this. When the district had sufficient funds prior to 2011, it agreed to a series of wage increases for union employees, added remedial summer-school classes, reduced class sizes, operated underutilized buildings, and devoted extra resources to improve poorly performing schools. Some of these measures could be quickly reversed, but some created ongoing commitments.

People who defend the district’s spending prior to the crisis point to a number of institutional and political realities. For one, the federal stimulus funding that came directly to the district had to be used for program enhancements, that is, the district had very limited legal ability to “bank” the funds. Moreover, graduation rates in the district are strikingly low, as is student performance on standardized tests. SRC members and state and local politicians naturally pushed the district to improve these metrics. The school district’s response was to spend available funds on initiatives that might help in the short run, even if it left the district vulnerable in the long run. And, as noted earlier, the district must approach the state and city each year with its hat in hand to ask for resources. If it did so while banking previously appropriated funds, this would undermine its case. Finally, prior to the November 2010 election, the district assumed that the state would largely replace the federal funding that the state used to maintain its support for K–12 education during the recession. It did not anticipate that the state would continue its reduced level of funding and cut state grants to the district. But once federal stimulus funding ended and the state and city refused to step up, the district was, like the Coyote in the Road Runner cartoon, standing on thin air.

Conclusion

Despite a vibrant downtown, Philadelphia is a much poorer city than many people realize. With one-quarter of its residents living below the poverty level, Philadelphia is the ninth-poorest U.S. city with a population over 250,000. Relative to Pittsburgh and the Philadelphia suburbs, the school district is significantly underfunded by the state and its city government, especially when one adjusts for the comparatively large percentages of special education, English language learners, and low-income students. In short, the district faces huge challenges with limited resources.

In addition to this fundamental fiscal weakness, the district is undermined by its governance structure. With no independent taxing authority, every year the district administration must plead for funding from the state and city. As with the boy who cried wolf, the politicians who hear the message repeatedly become skeptical. In addition, the city and state have an incentive to delay committing funds as long as possible in the hope that the other will take the lead.

In the summer of 2014, the district once again faces a dire fiscal situation and is appealing to the city and state to rescue the schools. The danger is that the district may have entered a vicious cycle in which persistent financial crises encourage more parents to leave the city or move their children to charter schools, further undermining the district’s financing and reinforcing the exodus of students.

John Caskey and Mark Kuperberg are professors in the department of economics at Swarthmore College. Erica Kouka provided research assistance.

Not far from the heart of Houston, unlikely alliance between a school district and nearby charter schools is bringing the best of both worlds to area students. As with most breakthroughs, the catalyst was simple curiosity. Duncan Klussmann, superintendent of schools for the Spring Branch school district, wondered just what was going on in those high-performing charter schools peppered around Houston, where both YES Prep and KIPP (Knowledge Is Power Program) operated several schools. Did they know something he needed to know?

Klussmann is a folksy, salt-of-the-earth kind of guy and a very traditional superintendent. But he did something most superintendents would never do: he scheduled a visit to a YES Prep school and spent some time with founder Chris Barbic. “I was really impressed with the culture, the environment, the interaction with the kids,” Klussmann recalled. “As I tell people, I often gauge things by whether my own three kids are missing out on something by not being part of a really good program.”

About a year later, Klussmann accepted an invitation to visit a KIPP school. Again, he liked what he saw. What stuck in his mind was the special classroom culture. Most of all, he liked the urgency: these teachers taught like their “hair was on fire,” to borrow a line from famed Los Angeles teacher Rafe Esquith.

How, Klussmann wondered, could I import some of that special sauce into my schools? And then he found a way. Why not make charter operators partners rather than competitors? We have what they want, great buildings and access to thousands of students they otherwise would have little chance of reaching. And they have that special culture we need to adopt.

So he did something else that most school superintendents would never even consider: he invited both YES Prep and KIPP to take up residency in two middle schools, building grade-by-grade expansions that would lead to YES Prep establishing a high school program within an existing Spring Branch high school.

Turns out Klussmann was on to something. More than 20 public school districts across the country, including the large urban districts of Boston, Chicago, Los Angeles, and Philadelphia, have quietly entered into “compacts” with charters and thereby declared their intent to collaborate with their charter neighbors on such efforts as professional development for teachers and measuring student success.

Striking a Bargain

The political appeal of district-charter compacts is evident. For mayors who have long admired the top charter schools in their cities but also remain wary of crossing the politically potent teachers unions that are hostile to charters, the “compact” designation nicely reduces the political heat. No longer is it districts (and their unions) against charters; it’s districts embracing charters. Who could be against that?

Unfortunately, it’s too soon to celebrate the end of the charter school wars. Compacts are still few in number, and not all have moved their districts past long-standing grievances (see, for example, “Boston and the Charter School Cap,” features, Winter 2014).

“Today we are on the battlefield of an ugly, nonproductive war on charters, all while children are stuck in underperforming public schools,” said Joseph DiSalvo, a former San Jose teacher and current member of the Santa Clara County Board of Education, which serves as an appeal board for charters rejected by school districts in the San Jose area.

District-charter compacts are sometimes a mere footnote to what remain caustic charter school battles, such as the one that took place in early 2014 in New York City. Newly elected New York City mayor Bill de Blasio moved to rein in the Success Academy charters run by his former political rival, Eva Moskowitz, by cancelling agreements that enabled charter schools to operate within district school buildings. Like anticharter forces elsewhere in the country, de Blasio chose the city’s highest-performing charter to attack. The higher the performance, the bigger the threat.

On the opposite side of the spectrum are compacts in Denver, the Texas district of Aldine, and Spring Branch, where superintendents are embracing high-performing charters by inviting them into their schools. These efforts couldn’t be more different from the contentious co-locations in New York City that sparked de Blasio’s ire. In cities with well-developed compacts, the charter and district principals and teachers constantly rub elbows to learn from one another. Even within contentious Santa Clara County, where most of the San Jose–area superintendents are doing their best to stiff-arm charter schools, the Franklin-McKinley school district brings in charters as welcome partners.

Why would these school superintendents lower the drawbridge while others are digging deeper moats to keep charters out? The compacts fill needs on both sides. School districts want to import some of the classroom culture and sense of urgency they see in charter schools. Some want charters to take more special education students or to hold low-performing charter-school operators to account. Fair enough. Best of all, they get to “claim” the test scores turned in by the charter students. What mayor or school superintendent doesn’t want to see headlines about rising test scores? Charter school leaders need building space and can pursue the mission of many charters, to help improve the broader system.

Compacts thus help top charter groups deal with two political Achilles heels that now limit their growth: special education and better accountability for low-performing charters. By coordinating special education with the district and setting common performance metrics for low-performing charters, compacts make the charter operators more politically palatable.

Finally, district compacts offer access to thousands of students charters might not otherwise reach. Says Mark DiBella, the Houston superintendent for YES Prep,

Our district partnerships have allowed us to expand beyond our growth model. Since charter schools do not receive any facility funding from the state of Texas, without our district partnerships our growth model showed that we could only fund raise enough to open 11 campuses. However, because of our co-located partnerships, which save us the upfront capital investment when starting a school, today we have 13 campuses—two of which are district partnerships. Furthermore, without district partnerships our growth model for the next five years would only allow us to open five additional schools. With our partnerships, we will be able to open a sixth school as well. This means that by the end of our next phase of growth, partnerships will have added more than 3,000 students to our overall enrollment capacity.

Should they prove effective over time, these compacts could help overcome the limits on growth that have long stymied high-performing charters (roughly defined as the top fifth of all charters): If you can’t find building space, and you can only expand by a grade a year, how can you ever reach a critical mass of students? Suddenly there’s an answer: view the district as a potential partner rather than a competitor, and you’ll be invited to the big dance.

The Gates Foundation now funds district-charter collaboration through formal “compacts” in 20 different cities. Gates support, which usually begins with planning grants of $100,000, pays for joint professional development; for designing a universal enrollment system; for establishing common metrics for evaluating all students, regardless of school type; and for creating more personalized learning for students.

Especially intriguing about the Gates-funded compacts is their top quarterback, Don Shalvey, a former district superintendent who also founded Aspire Public Schools. Shalvey, who is someone with credibility on both sides of the district-charter divide, now runs the Gates program. Think of Shalvey as a Bill Clinton figure. He may no longer be president, but when he comes to town to promote a favored cause, doors open. Suddenly, superintendents dare to think differently.

It’s tempting to dismiss compacts as lightweight fads, particularly if one focuses on districts where district-charter tensions are still paramount. Top charter schools also insist that there’s no way to dilute the brand and still achieve success. Charters that succeed with low-income children go full bore: all-out culture building in the classroom, students and staffers willing to endure longer school hours, bright teachers willing to adapt to precise training regimens and relinquish a fair amount of privacy (giving out their cell-phone number for afterschool homework questions, for example). How could that culture possibly mesh with schools that recruit from traditional education colleges and run according to union contracts?

The four district-charter compacts—Denver, Colorado; Aldine, Texas; the Houston-area Spring Branch district; and Franklin-McKinley in San Jose, California—reveal the variety of forms compacts can take as they seek to answer that question. First up is Spring Branch, a compact that has a promising template for others to imitate.

Spring Branch
Independent School District, Houston, Texas

When I first asked 6th-grade math teacher Jennifer Rivera at Northbrook Middle School how wrapping a YES Prep charter school into half of her school had changed her life, she answered, “Not that much.” Then, when she began talking about the changes at her school, it became clear that all the small changes sparked by the compact were adding up to something big.

Most noticeable, she said, were the classroom culture issues. “It’s the little things, like asking students to tuck in their shirts.” Her side of the school, the traditional Northbrook side, adopted the SLANT program used by YES Prep to shape classroom culture. SLANT, described in Doug Lemov’s book Teach Like a Champion, asks students to: Sit up; Listen; Ask and answer questions; Nod your head; Track the speaker. “It’s working,” said Rivera. “We’re asking for 100 percent of the students’ attention before we move on.”

One significant payoff: students on the traditional side of Northbrook have adopted a new attitude about academic success. “Before, kids were not so keen on doing well academically. They would call each other nerds. Now, it’s the norm to know the answers, to not be labeled. It’s cool to know the answers,” explained Rivera.

So far, there’s no tension or sense of competition between the two sides of the school, she said. The YES Prep students post higher test scores, but that is seen as the logical outcome of spending more time on task: the YES Prep school day is an hour and a half longer, and their school year is two weeks longer. For parents and students, it’s a matter of finding the right fit. Scheduling more academic time is not for everyone. “These are our children,” said Rivera, acknowledging that the YES Prep children would be in regular Northbrook classes if there were no charter offered. “If you want to give parents choices, you can’t argue with it.”

Among the first to receive a Gates compact grant, Spring Branch has now seen more than $2 million from the foundation. Two pieces of the compact stand out. First, the district used the compact as the foundation for a new accountability strategy. Spring Branch chose a single metric: the percentage of graduates who successfully complete some form of higher education. That could be a technical certificate, a two-year degree, or a four-year degree. The goal was to double that rate from 36 to 72 percent over five years.

Second, to reach that goal, Klussmann and his board realized they would have to usher in changes that were deeper than the two middle-school charters evolving into a charter high school. Klussmann hired veteran KIPP leader Elliott Witney. Giving a charter guy a major role in shaping policy for a 35,500-student district was a gutsy move.

Witney moved quickly to ramp up professional development for teachers. Where possible, he ushered in charter-traditional mingling for professional development. Math teacher Rivera, for example, now wants to adopt the YES Prep math curriculum for her students. “Their math program is at an advanced level. Every 8th grader takes high school algebra.”

Recently, Witney won approval to hire 20 academic advisers whose job is to get high school students on track to both enter college and then graduate from college. To train the advisers, Witney turned to the successful KIPP and YES Prep college-counseling programs. All the new counselors visited the outside charters to see how they worked. “What this brought is a unique culture, this idea that you will go to college. It’s not a matter of whether, but where.”

Given the charter-district fighting around the country, why is there no angst in Spring Branch? “I’ve thought about that a lot,” answered Ellen Winstead from YES Prep, who acts as the liaison in Spring Branch and oversees the partnership in Aldine. Each side gets something of real value while retaining the core values. “We’re giving our scores to them,” said Winstead, which means the higher-scoring YES Prep students at Northbrook count toward the district average. That’s not a big sacrifice for YES Prep, which runs high-scoring schools, but it means something to Spring Branch. YES Prep gets “free” state-of-the-art buildings (paying only operational costs), which means everything to the charter group. In Texas, charters don’t get facility funding, which for YES Prep means a start-up cost of $11.5 million per school. The partnership means taking that start-up cost to zero. “Our board loves partnerships. That means $11.5 million less to raise,” explained Winstead.

The core values of each organization are preserved. For YES Prep, that means a longer school day, longer school year, and unique student culture. “These are sacred cows to YES Prep.” For Spring Branch, that means the charter students, whom they consider “their” kids as well, gain access to electives that charter students don’t normally get—the opportunity to play in a band or orchestra, take art or choir. For Klussmann, that was a sacred cow.

When you take care of everyone’s sacred cows, harmony reigns.

Aldine
Independent School District, Aldine, Texas

I met Aldine superintendent Wanda Bamberg in 2009 while working as the project journalist for the Broad Prize. Aldine was a district dealing with considerable poverty and equally considerable language challenges. Some 80 percent of its students qualified for subsidized lunches, and 31 percent were designated as English language learners. Despite that, it was clear that Aldine was punching far above its weight. In every school the Broad evaluation team visited there was a sense of urgency, the exact feeling you get in the best charter schools. Every school had an exacting academic playbook followed by every teacher and administrator.

Given the level of academic success in Aldine, it might seem surprising that Bamberg would choose to bring in a charter organization. What’s to gain? But when YES Prep purchased an old grocery store in a part of the district where Aldine was losing population, the district knew it would lose even more students. Bamberg took the admittedly practical step of inviting YES Prep to join forces.

Soon, however, the partnership went beyond the practical. “We went to their campuses; they visited our campuses,” said Bamberg. “We discovered we had a lot in common.” From Bamberg’s perspective, parents who want to enroll their children in a YES Prep school are no different from parents who want to enroll in a magnet school, an option Aldine already provided. Fitting YES Prep into an existing middle school, with plans to build through high school, created just another choice option for parents.

Aldine received a $100,000 planning grant from the Gates Foundation in January 2014. The top priority for Bamberg now is tapping into YES Prep’s highly regarded college-preparation program, the same program Elliott Witney draws on in Spring Branch. In Aldine, that YES Prep program will be available to all high-school students, not just those sharing space with YES Prep students. It will take more than a year to line up the right resources and teachers for that program. The formal launch comes in the 2015–16 school year.

Here, too, all the sacred cows get respect.

Franklin-McKinley
School District, San Jose, California

Some school superintendents in the San Jose area say they are not anticharter; they just want the authority to control which schools enter their district. But then they never approve charters. John Porter, who runs the high-poverty Franklin-McKinley district here, also wants to decide on charter access, but unlike his colleagues he makes charters a key part of his education plan.

San Jose is home to the Rocketship charters, known nationally for their pioneering work using blended learning, which mixes online and classroom work. Porter could see that Rocketship was doing something different, so he invited them in. Rocketship Mosaic, the first of the schools to open in Porter’s district, took in 426 students its first year (2011–12), in kindergarten through fifth grade, and turned in stunningly high scores in 2012 (for schools serving a high-poverty population, Mosaic’s score of 872 on California’s Academic Performance Index tests was considered a home run). A second Rocketship school opened this school year, and a KIPP middle school opens this coming year.

Having Rocketship within the district gave Porter and his teachers a close-up look at their methods. The teacher exchanges didn’t always turn out as expected. Porter recalls one meeting where the expectation was that Rocketship teachers would be explaining their techniques to the more veteran Franklin-McKinley staff. Instead, the young Rocketship teachers, nearly all from Teach for America, wanted advice from the district teachers. “How do you stay in the profession for years and still keep yourself motivated?” they asked.

“The conversation really surprised the two principals in the meeting,” said Porter.

The creation of Rocketship Mosaic inspired surrounding principals to pick up their game. The staff at Robert F. Kennedy Elementary, which is sandwiched between Mosaic and another charter school, recovered from their initial “demoralization,” said Porter, and established a successful science-themed school that partnered with San Jose’s Tech Museum. “I think all our schools feel the competition,” said Porter. “It has been an amazing transformation. We have schools going door-to-door and showing up at Walmart to recruit. They’re doing the kinds of things charters do.”

That’s what makes the compacts so interesting. From the beginning, the charter movement has been split on the issue of competition versus collaboration. The market-oriented charter advocates expected that students drawn to charters would spur radical reform by district leaders and unions experiencing the losses. In truth, the reaction has been more pushback—trying to deny students access to charters rather than undertaking internal reform. Other charter advocates, inspired by the original intent of charters, remain wedded to the “laboratories of invention” theory of charters. Districts, they assumed, would want to absorb the charter innovations. In reality, that has rarely happened. But here, in this district, both patterns emerged: charters are laboratories of innovation and provide helpful competition.

Porter’s district also received a Gates grant of $100,000 in January 2014 to deepen the charter-district relationships. The two top priorities are drawing together staff from both sectors to deal with Common Core challenges and boosting the number of special education students taken by charters. “Our two-year goal is to bring the charters up to the district (special education) average,” Porter said. “We’ve found a perception among special ed parents that charters don’t want their kids, so we’re going at that head-to-head.”

Denver
Public Schools, Denver, Colorado

The Denver compact, signed in 2010, is rightly considered the nation’s premier compact. Denver was among the first cities to receive Gates compact funding, and the district is slated to receive more than $4 million in total. The compact brought 21 new charter schools into Denver Public Schools, including 11 run by three of Denver’s charter stars: DSST Public Schools (which began in 2004 as the Denver School of Science and Technology), STRIVE Preparatory Schools (which began in 2006 as West Denver Prep), and KIPP. Those three operators’ combined student population grew from 2,343 students in 2010 to 6,100 in 2014.

A 2014 study by Denver’s Donnell-Kay Foundation concluded that the charter schools account for most of the steady increase in test scores within Denver Public Schools (DPS). High-performing charters, concluded the report, “provided a tide that has lifted aggregated data.”

There are many moving parts to the Denver compact, including lessons shared from both sides. But to an outside visitor, the core of the Denver compact appears to be a simple transaction: top charters get access to DPS facilities and agree to shoulder a fair burden with special education students.

One of the most distinctive components of the Denver deal is “Compact Blue,” a program in which DPS teachers and administrators meet with their charter counterparts. I observed the first of those meetings at the STRIVE Prep-Westwood campus. Two teachers, Kristie Burke, a 7th-grade reading teacher at STRIVE, and Kerrie McCormick, a language arts teacher at a struggling DPS middle school, sat side by side while the STRIVE teachers went through a professional development day.

McCormick was stunned to see the kind of rich, up-to-date data Burke had for each student. “I cannot only see if my students have passed or failed,” said Burke, “I can see specifics and break them down into what standards they are mastering, where they need help. We break it down further into reading and writing standards, fiction and nonfiction.” Said McCormick after watching the session, “To be able to go to the computer and see my students by skill level, all neatly lined up with bar and pie graphs—that would be so valuable.”

Can one school’s best practices rub off on another school? It’s worth trying.

Richard Whitmire is author of On the Rocketship: How Top Charter Schools Are Pushing the Envelope , which chronicles the rise of leading charter school network Rocketship and the impact of high-performing charter schools in Denver, Spring Branch, and other cities across the country.

In June, a judge declared California’s seniority protection laws unconstitutional. Citing the 1954 Brown decision, Judge Rolf Treu, in Vergara v. California, declared the laws in violation of the equal protection clause of the California state constitution because they limited minority access to effective teachers.

Teachers unions are aghast that a judge has interfered with their special privileges. Their close ally, Michael Rebell, who has filed multiple adequacy lawsuits against state legislatures, told a reporter, “It is basically unprecedented for a court to get into the weeds of a controversial education policy matter like this.” Invoking judicial restraint doctrine, he expects other judges to “defer to the legislative branch, which…knows more about the workings of these policies.”

Given Rebell’s connections to teachers unions, his conversion to judicial restraint doctrine is only to be expected. More surprising are the criticisms of Vergara voiced by Education Next’s own judicial observers, Martha Derthick and Joshua Dunn. Also calling for judicial restraint, they argue that the principles enunciated in Vergara could allow any child to file suit against his teacher. The reaction of Frederick Hess, of the American Enterprise Institute and Education Next, is no less apocalyptic: “If courts decide that civil-rights claims can be stretched to dictate personnel management in education, it’s hard to see where they would stop.”

But Judge Treu did not dictate to the California legislature; he said only that its current policies are unconstitutional. Nor does he require any specific new remedy, leaving that job to the legislature.

This is hardly the first California education law to be declared unconstitutional. In the famous Serrano case (1977), the court ruled the state’s spending formula unconstitutional on disparate-impact grounds. Subsequently, Rebell and other union-backed attorneys persuaded state courts from Kentucky and New Jersey to New York and Washington to declare state expenditure policies unconstitutional on the grounds that they are “inadequate,” a standard so vague and amorphous as to make Vergara seem as precise as a mathematical equation.

Nor have the courts balked at ordering legislatures to spend millions—even billions—more. At Rebell’s request, a trial court judge in 2006 ordered the New York legislature to send New York City an additional $4.7 billion (or more than $5,000 per student).

Judicial restraint became a celebrated doctrine when Oliver Wendell Holmes said laws should not be declared unconstitutional without considering them “in the light of our whole experience.” That phrase was used by Felix Frankfurter to defend New Deal legislation at a time when President Franklin Roosevelt found himself at loggerheads with the Supreme Court.

But since Brown, courts have felt no compulsion to defer to the legislature when they identify violations of civil liberties and civil rights. The Supreme Court has told schools they must allow free speech, even when duly elected school boards think students are there to learn, not to proclaim. When students are accused of disrupting the school, boards cannot suspend them for more than a few days unless the student has been given the right of counsel and cross-examination. If judges can dictate the mechanisms of discipline enforcement, certainly they can review laws affecting student access to effective teachers.

Nor does Treu violate Holmes’s injunction to look at the case “in light of our whole experience.” The plaintiffs detailed the many ways in which seniority protection placed teacher interests ahead of those of minority students.

Judicial action is most appropriate when legislatures have been captured by special interests. Just as the Brown decision broke the segregation logjam, so Vergara provides an opportunity to break the union stranglehold over teacher-tenure policy. Similar suits are popping up in New York, Connecticut, and New Jersey. The U.S. Department of Education has just asked states to demonstrate the ways in which they are ensuring equal student access to effective teachers. Public opposition to teacher tenure outweighs support by a 2:1 margin. All of a sudden, the powerful California Teachers Association is on the defensive. The court has done its job.

—Paul E. Peterson

In the old world, a great teacher had two choices: stay in the classroom at relatively low pay or leave to become an administrator. There was no way to advance your career or expand your reach while working as a classroom teacher. But traditional school leaders have only an indirect impact on instruction. Their work may include designing processes, planning use of resources, leading professional development, providing observation and feedback, and offering instructional coaching, but they are not teachers. Here at Charlotte-Mecklenburg’s Ranson IB Middle School, there’s a third option: become a multi-classroom leader (MCL). The new-world role of MCL combines planning, coaching, and direct teaching. It is my dream job.

At Ranson, innovative staffing, a flexible teaching schedule, and strategic use of technology combine to create personalized learning for 1,100 students in grades 6–8. This year, as the first MCL at Ranson, I lead two pods of three teachers and one learning coach (or teaching assistant) each, and I am responsible for the learning outcomes of 800 6th and 7th graders. Our blended-learning rotation model for math incorporates up to 40 percent online instruction in 6th and 7th grade (8th grade to follow next year). Sixth graders move in groups of 15 to 17 between face-to-face instruction with their teacher and online instruction in the computer lab, under the supervision of a learning coach and their MCL. In the lab, students work on activities created, assigned, and tracked by the teaching team. The model is similar for 7th grade, but the students stay in their classroom, which incorporates a Chromebook learning lab. A learning coach and I float between the different classes.

A few entries from my journal show what the daily life of an MCL looks like:

Monday

5:45 p.m.: Sixth-grade math teachers Courtney and Clarissa have come for help with a surface area lesson we will co-teach the next day. We find 3-D shapes around the room. “They need to touch all the faces and edges first,” says Courtney. I ask them what part of the lesson the students will need support with. They agree it will be the independent practice. I adjust my schedule. Courtney walks out, uttering her ritual “I feel better now.” I feel better, too.

Tuesday

7:00 a.m.: I check my e-mail and find messages from several young scholars on Edmodo (the social learning network we use). Yesterday, my 6th graders worked on an online review activity I created with Educreations (an iPad-based interactive whiteboard for creating instructional videos). I sent each of them feedback on their work. Their messages this morning are just to say thank-you. It makes the hour I spent doing this yesterday completely worth it.

9:15 a.m.: Off to the 6th-grade computer lab to pull out six to eight scholars to work with me for 30 minutes in what we call the Math Genius Bar. We start the lesson with a problem they struggled with last week. Keaun blurts out, “How did you know we needed this?” I smile and reply that his teacher and I are always talking. He seems both satisfied and worried.

9:55 a.m.: Working in class with Courtney while she finishes her mini-lesson on finding surface area. She did a good job modeling the task (called a Think Aloud), but I feel the need to add a Check for Understanding before they start practicing. I look over at her, she nods, and I start cold-calling students.

11:00 a.m.: Sixth-grade planning period. Clarissa asks, “What could be a highly rigorous problem with surface area?” We create one together and add it to the lesson plan. It involves painting my bedroom but not the floor, the ceiling, or the windows.

11:50 a.m.: I have some planning time before the next Genius Bar pullout at 12:35 p.m. With iPad and stylus in hand, I start responding to students’ Edmodo posts from the morning. I glance at my computer clock; it is already time for the next block and I forgot to eat lunch. When a Frenchman forgets about eating, this is a sign that he loves what he does.

Being an MCL is a great job. I hope more teachers around the country will soon have the opportunity to break down the walls between classrooms and expand their reach as I have.

Romain Bertrand blogs about his work as a multi-classroom leader at Ranson IB Middle School at expandingthereach.wordpress.com.

Teachers Versus the Public: What Americans think about schools and how to fix them
by Paul E. Peterson, Michael Henderson, and Martin R. West
Brookings Institution Press, 2014, $28.00; 144 pages.

As reviewed by David Steiner

Most Americans possess only general, and not quite accurate, notions about their public schools. Moreover, the perceptions of the general public and the teaching profession often differ quite dramatically.

Paul Peterson (well known to readers of Education Next as its editor in chief), director of the Program on Education Policy and Governance and professor at Harvard University, and his fellow authors have published a short book that quantifies this reality. They find that Americans typically overestimate local school performance and underestimate costs and salaries. More than 80 percent of the public support annual student testing, three-quarters favor charter schools, two-thirds favor higher teacher pay, and half are in favor of means-tested vouchers. In contrast to the general public, teachers are less likely to support school choice, testing, and school accountability, and more likely to support higher teacher salaries and raising taxes to pay for them.

These findings are less than surprising, and some of them are familiar. It has long been known that Americans are locally generous to education performance, meaning that they inflate the academic performance of their own children and their own children’s schools. On the latter point, Gallup polling between 2007 and 2012 indicates that about 50 percent of respondents gave a grade of A or B to the schools in their community, while only 20 percent gave those grades to the nation’s schools as a whole. The Peterson results track the Gallup poll results quite closely, with 44 percent of the respondents giving their local schools one of the top two grades and 20 percent giving those grades nationally.

So what’s the news in the research this book reports? Two findings are noteworthy. First, provided with more precise information about the performance of their schools and districts, Americans lower their ratings of local schools and express greater support for charter schools, parent trigger mechanisms, and vouchers (if extended to all children). These views are especially prevalent among Americans who live in low-performing school districts. Give them information about teacher salaries, and Americans become more skeptical about supporting increases in teacher pay.

The second set of findings regards differences between African American and white opinions. There is relative silence, unfortunately, on Hispanic views. The text notes, without any further analysis, that Hispanics rate their schools far more highly than do African Americans, but why? Focusing on African Americans, the authors report that they are (understandably) far more critical of their schools than the rest of the public, and far more supportive of what is repeatedly called “the reform agenda”: merit pay for teachers, charter schools, vouchers, and tax credits to fund private school scholarships. On all of these issues, the gap in opinion between African Americans and teachers is especially marked.

It seems clear that Peterson and his colleagues are not simply reporting these findings for their own sake; a political agenda gradually takes shape in these pages. First, they hint that disseminating real data about school performance and teacher salaries would lead to more rapid reform. (In this context, the authors hold out hope that assessment results based on Common Core standards will increase transparency of educational achievement across the country, a bet that looks less secure than it did a year ago.)

Second, Peterson and his coauthors see an opening in the split between African Americans and teachers unions. The authors note the wealth of teachers unions and the role that their dollars play in keeping the national civil-rights leadership in line with union priorities—against the wishes of most black citizens. But, they argue, African Americans’ growing awareness of the failings of public schools and their enthusiasm for alternatives are putting pressure on the status quo: “The alliance between teachers and minorities within the Democratic coalition can be held together as long as education problems are defined as the by-product of inadequate funding. While successful in the past, today that strategy is becoming problematic…”

The authors tiptoe in these intimations; they do not stride. There is nothing concrete on political strategies. Only occasionally does the tone change. Drawing on comparisons to the airline and communications industries, the authors in their closing pages advocate for “large-scale competition” in the education sector. Their argument: “If large-scale competition is introduced, the government will not need to persuade teachers to go along with reforms, as teachers will be struggling to keep their jobs if students move elsewhere.” But while the call for competition is clear, the strategies to encourage it are not laid out.

What to make of the book? One could certainly quibble with small points. For instance, the authors refer to “the reform” case as though it were monolithic and uncontested, and cite, for example, arguments that certified teachers are no more effective than uncertified, with no indication that this finding is disputed in the literature. Throughout, one might wish that the authors were less coy about the agenda they can see taking shape around their findings.

But for this reviewer, the deeper issue lies elsewhere: the authors’ narrow range of “reform” opportunities. When parents know their children are getting a disastrously poor education, it is unsurprising that they will support alternatives, especially if those alternatives have shown powerful results, as, say, is the case in some New York City charter schools. But there is a certain begging the question in the authors’ reserving of the term “reform” for only the set of policies endorsed by such bodies as Democrats for Education Reform, the U.S. Department of Education and sympathetic state school chiefs, and certain figures, such as Jeb Bush, in the Republican Party.

The survey and the book would have been more provocative had a much broader “reform” agenda been considered, one that proposes rethinking teacher preparation, reshaping the tax structure, or establishing a national curriculum, for instance. Other countries, such as Finland, closed their schools of education and moved them to the universities, requiring subject-matter degrees and raising the bar for entry substantially. Ontario, too, focused its efforts on strong teacher preparation. Alberta, Canada, overhauled its property-tax structure to distribute resources more equitably, and at the same time expanded funding for school choice (including home-school stipends). Still other countries, such as England, are moving toward content-rich curricula, since academic rigor is shown to diminish the achievement gap as few other interventions do. Peterson’s team did not explore any of these reform options.

The authors would naturally respond that such alternatives aren’t seriously on the radar screen. That, of course, is the point. The policymakers in Washington who define “reform” delimit the realm of imaginable alternatives. How would the general public in the United States respond if given an educated choice between the reform agendas in Ontario and Washington? That’s what we need to know.

David Steiner is dean of the School of Education at Hunter College.

Like many high-poverty middle schools, Oakland’s Elmhurst Community Prep is trying to reach students who are academically all over the map. One-third of the students are working at grade level in reading and math, says Principal Kilian Betlach. Another third are one to two years behind. The remaining third are three or four years behind—or more. “You can’t teach them by aiming for the middle and providing these little supports,” says Betlach.

Differentiation—teaching students at very different levels of achievement in the same class—is “the greatest challenge facing America’s schools today,” writes Michael Petrilli (see “All Together Now?” features, Winter 2011).

“Teachers are told to sprinkle your differentiation fairy dust,” says Betlach. With 32 students in a class, and no aides, “it’s not possible.”

What is possible?

A foundation-funded experiment is testing whether “blended learning” can personalize instruction in eight Oakland schools. Blended learning combines brick-and-mortar schooling with online education “with some element of student control over time, place, path, and/or pace” of learning, according to the Clayton Christensen Institute definition of the term.

The Rogers Family Foundation, created in 2003 by T. Gary and Kathleen Rogers, launched a blended-learning pilot in four Oakland Unified schools, including Elmhurst, in fall 2012. Two more district schools and two charters were added in 2013. The foundation focuses on improving Oakland’s troubled schools and is funding the pilot with help from the Quest Foundation, the Bill and Melinda Gates Foundation, and others.

Urban schooling doesn’t get much tougher than in Oakland. More than two-thirds of Oakland Unified School District students are Latino or African American; 80 percent qualify for free or reduced-price lunch. One-third are English language learners (ELLs). Only half of disadvantaged 9th graders earn a high school diploma in four years.

Pressured by community groups, the district created small, autonomous, quasi-charter schools like Elmhurst starting in 2001. Scores began rising but remain low.

Despite this bleak picture, Oakland schools do have one advantage as they attempt to transform through blended learning: nobody thinks the status quo is good enough.

Blend and Rotate

“Student control” over the pace of learning is on display in Will Short’s Math 8 class at Elmhurst. Laptop users are working on Khan Academy quizzes geared to each student’s skill level. At the front of the class, a display charts their “energy points,” a measure that includes “on taskness” and the percentage of correct answers. Students receive instant feedback.

Freed from whole-class instruction by the technology, Short has time to reteach concepts to individual students or small groups. Advanced students can move ahead.

At Bret Harte Middle School, which joined the pilot in the second year, Chantel Parnell divides her 6th-grade math students into three groups. On a day in late October, some are using Khan Academy or Google Drive on Chromebooks, while another group is constructing box plots at their desks. The rest are sitting in a circle with the teacher. “Do you understand why you got this answer?” Parnell asks a student.

At that moment, a girl in the Chromebook group raises her hand. A student walks over to help. Oakland schools can’t afford aides in mainstream classrooms, so Parnell has asked students to volunteer to coach their classmates.

Like Parnell, most pilot teachers use the “station rotation model.” Students move between computers, teacher-led discussions, and, sometimes, group projects or independent desk work.

A few teachers use the PC lab. Some of Patricia Wong’s students are working on adding and subtracting negative numbers. Wong explains 4 + (-5) to a boy. “Plus a minus is subtraction,” she says. “What do you do?…Why?”

Meanwhile, one girl is drawing a number line on the screen. Another has moved on to multiplying and dividing with negative numbers. Advanced students are doing word problems. One boy is taking the “mastery challenge,” which pops up randomly. The challenge asks him to add 2.83 + 3.5.

Blended learning isn’t just for math classes at Bret Harte. Teacher Amy Colt uses the online learning program Achieve3000 to teach English and social studies. The program provides Associated Press news stories rewritten to match each user’s reading level, plus a reading quiz to check comprehension and a writing test. Colt circulates, talking to students individually.

The Launch

The idea for the pilot came from Oakland principals who had received earlier Rogers grants. They “weren’t having the impact they wanted with technology” and wanted to do “something larger and deeper,” says Greg Klein, director of blended learning for the Rogers Foundation. Klein is coauthor of “Blended Learning in Practice: Four District School Journeys,” a case study of the Oakland project written with Carrie McPherson Douglass, who’s now with the Cities for Education Entrepreneurship Trust, an umbrella organization for urban reform groups nationwide.

Teachers said, “I don’t have enough time to meet the needs of my highest-skilled students or my ELLs. It would be great to personalize instruction.” Those proved to be “gateway drug conversations,” says Klein.

Tracey Logan, who manages technical services for Oakland Unified, was enthusiastic. “We saw an opportunity that was aligned to where we want to go,” she says.

Schools competed for the chance to participate. Working with Education Elements, a company that designs “personalized learning solutions,” the foundation initially chose EnCompass Academy (K–5), Korematsu Discovery Academy (K–5), Madison Park Academy (grades 6–9), and Elmhurst Community Prep (6–8). All are district-run schools in high-poverty, high-crime areas.

“We looked for principals who were strong leaders able to implement change,” says Jane Bryson, who directs the education team at Education Elements. Like their principals, all the teachers within the schools that implemented blended learning were “early adopters” who wanted to try something new. They committed to spending one hour a week after school on collaboration and training, which Rogers funded. That time proved to be critical, says Bryson.

But the pilot’s first year was difficult.

Education Elements trained and supported Madison Park and Korematsu teachers. The company’s software platform lets students sign in to a home page that shows all the learning programs available. Teachers can access a data dashboard to see all the data on each student and create assessment tools.

All of Madison Park’s math teachers and two other teachers volunteered to be in the pilot. At Korematsu, two 4th-grade teachers volunteered to use blended learning to teach both reading and math.

Hoping to test a different approach as well, Rogers brought in Junyo, a small company, to work with Elmhurst and EnCompass Academy. Junyo tried to develop its own platform with a portal for students and a data dashboard for teachers but gave up in mid-September, returned the money, and quit the pilot. Klein and the foundation’s two blended-learning specialists took over tech support and training for the two schools. But lingering effects of the Junyo experiment added to what were already significant challenges.

At the former Junyo schools, the lack of a single portal meant that a student might need to log in separately to five different programs. Klein made sure they could use the same username and password at least.

Rogers had planned to try blending in just 4th and 5th grade at EnCompass, but Junyo signed up all the teachers. Many didn’t realize how much work it would take, says Klein. Implementation was “shallow.”

Elmhurst had a “failure to launch” in what Betlach calls “Year 0.” Junyo’s “advice on devices was divisive and faulty.” Teachers never had reliable Internet access.

Elmhurst’s building, shared with another small middle school, was erected in 1906 and partially rebuilt in the 1920s after a fire. To make room for desktop computers, the foundation paid to have built-in cabinets removed, something teachers had wanted for a long time. But there weren’t enough wireless access points.

On the software front, Rogers learned a lot from the first year about what not to do, says Klein. “We asked teachers to learn multiple education tech tools at the same time.” It was too much.

Across the four schools, pilot middle-school teachers used a number of different learning programs, including Achieve3000, i-Ready, Khan Academy, ST Math, iLearn, and iPass, according to an SRI International report on the pilot’s first year (see First-Year Evaluation sidebar). At the elementary level, Google Drive, Achieve3000, Mangahigh, Khan Academy, Digital Passport, and i-Ready were popular.

“Online content providers can look like bright, shiny objects,” says Bryson. And Rogers was paying.

“It was overwhelming,” says Keara Duggan, an Education Elements staffer who worked with teachers.

“Now we say, start with Khan, and then add more as you see the need,” says Klein. The pilot’s new motto: “Go slow to go fast.”

His other conclusion: “Listen to teachers. Be authentically humble.”

“We did way too much too fast,” says Logan.

“This is really, really hard,” teachers told her. “It’s like being a first-year teacher all over again.”

A Leap of Faith

Once teachers gain experience with blended learning, they don’t want to go back, says Logan. “They say, ‘I wouldn’t do it any other way.’”

Technology will “change what teaching looks like”—eventually, she predicts. Right now, however, “it’s a bit of a leap of faith.”

Technology can “make the best use of teacher time, adapt to meet students where they are, and encourage collaboration and creativity,” Logan says. It can “expand the classroom. Students don’t just have access to what a teacher knows or what’s in the science book.” Most of all, she wants students to “take ownership of their learning.”

Jessica Tucker wants that for her students, too. Every Friday, Tucker would give a math quiz to her 6th graders at Madison Park. Half would fail. She’d try to reteach the lesson the following week, while introducing the next concept. On Friday, once again, half would fail.

After making the transition to blended learning, Tucker asked students to signal if they were ready for a quiz. Thumbs-up meant ready, thumbs-down meant no way, and a sideways thumb meant “not sure.”

On one Friday, 15 students put their thumbs up. Those students all passed the quiz. Tucker taught an extra lesson for those who needed it before they took the quiz. “It was effective because they felt they had a choice,” she said in a video on differentiation in a blended classroom. “I did an extension lesson at a higher level” for the ones who took the quiz first.

“I’ve done that every Friday since,” says Tucker. “It has improved mastery scores a lot.”

She uses a program called MasteryConnect to design her own tests. If a student does poorly, “I can say, ‘before you retake test, I want you to go to these four Khan videos.’ The kids who didn’t get it after two lessons with me, they’re obviously not going to get it with another lesson with me.”

Tucker wants to instill the belief that “you can’t not master a concept. I don’t care how many times you have to take it. You can’t leave until you learn.”

Year 2 and Beyond

In Year 2 of the pilot, Rogers added two district middle schools, a charter elementary school, and a charter K–8 school, while serving more grade levels at the original four schools.

ASCEND, a district school turned K–8 charter, has a Chromebook for every student and every teacher. Aspire’s Millsmont Elementary, a charter, has gone in a different direction, by designing a mobile computer lab that groups students in “pods” of four.

Korematsu, one of the original district pilot schools, is blending in nearly every classroom now. EnCompass simplified its program, and the principal hired teachers who are eager to use digital learning.

This year, Elmhurst’s Internet works consistently at a high speed thanks to newly installed access points. Blended learning is starting to work, says Betlach.

“It allows you to gather data on student performance much more quickly,” says the principal. The “exit ticket”—a mini-quiz at the end of each lesson—gives students instant feedback. “Kids say, ‘I really like to know how I did right away.’”

Teachers can track trends and patterns: Which question was problematic? What needs reteaching? “Better data let you target instruction to where kids are,” says Betlach. With students at widely different skill levels, teachers can have “different groups working on different things.”

All the 8th-grade teachers are participating in the blended pilot this year. So are two special-education classes and one 6th-grade math/science class. Teachers who’ve tried it are spreading the word, says Betlach. Nonpilot teachers ask him, “When do I get my Chromebooks?”

Blending More, Spending Less

Rogers is learning how to blend without overspending. The foundation “overpaid” for hardware in Year 1, spending $670 per student, says Klein. Schools bought MacBook Airs, Lenovo PCs, and Windows desktops. This year, the four new schools and the new teachers at Year 1 schools bought Chromebooks for $249 apiece. The price already has fallen to $199, says Klein.

Chromebooks hold up, require less tech support, and have a battery that lasts all day, says Klein. “Google has a cloud-based interface to manage the devices. You can buy 20 Chromebooks and 30 seconds later they’re set up and away you go.”

Teachers are using more free software this year. “You can get a lot of value out of Khan Academy at zero cost,” says Klein. “Then, when you go shopping for premium content, you have a better idea of what you need.”

Because of savings on computers and software, the Year 2 cost is $10,000 per school, half the cost of Year 1, says Klein.

With hardware and software spending going down, Rogers is focusing on helping teachers use technology to improve learning. That’s what really counts, says Klein.

His don’t-go-there example is Los Angeles Unified, which spent $500 million to buy iPads for every student. Implementation was a disaster. Other districts, such as Miami-Dade and Guilford County, North Carolina, have put laptops-for-all programs on hold.

The Rogers pilot is having a ripple effect in Oakland and beyond. Schools that didn’t make the Rogers pilot, both district and charter, are “trying to blend with very little money,” says Klein.

Learning Without Limits (LWL), a district-run elementary that recently converted to a charter, applied for the blended-learning pilot but didn’t make the cut. The school went ahead anyhow. With few computers in the early grades, LWL blends in 4th and 5th grades.

School PTAs and local donors are helping. Some use very low-cost computers refurbished by Oakland Technology Exchange West. Others can afford Chromebooks.

Before the pilot started, only one or two “bleeding edge” charters in Oakland were trying to use blended learning, says Rogers’s Year 2 report. “Just two years later, our eight pilot schools are joined by at least five more district schools, and practically every charter in the city is actively leveraging adaptive online content to personalize instruction.” A majority of Oakland schools have asked the foundation how they could try blended learning.

Oakland Unified is investing $3.5 million—half of its one-time Common Core implementation dollars from the state—to upgrade networks at every school. The district won a $100,000 Gates Foundation planning grant to design a personalized blended learning system at all its schools and is seeking funding for implementation.

Nearby districts have adopted blended learning on their own. Thirty-five miles to the south, next to San Jose, Milpitas Unified is blending instruction in two-thirds of elementary-school classrooms; the district’s teachers designed the plan. At Milpitas High School, the district has provided Chromebooks for students and teachers to use across the curriculum.

Phasing Out the Factory Model

It will take some time to determine how well blended learning works in the Oakland schools. Even though most teachers reported that students were highly engaged when working with digital content and were better able to learn the material, only one-fifth of Rogers pilot teachers said their students did better on benchmark tests the first year (see Figure 1).

Results on the 2013 California Standards Tests were “mixed,” says Klein. Madison Park’s 6th graders outperformed the district average in math and algebra students also showed progress. Korematsu, which blended only in 4th grade, showed gains in reading, where blended learning was concentrated, and smaller gains in math.

Rogers predicts “dramatic changes to student outcomes in years two, three, and beyond” as personalized learning takes hold.

Because there’s no control group, it will be difficult to tell whether higher scores are due to the pilot or other factors. And California’s state testing system will not report scores next year because of the transition to Common Core standards, which will make it even harder to track progress. The district gives the Scholastic Reading Inventory exam three times a year, however, which will enable the pilot schools focusing on English language arts to see what’s working (see Figure 2). In addition, some online content programs have built-in assessments that can be used to evaluate students’ progress.

The foundation is surveying teachers and students to see how attitudes are changing.

In the pilot’s first year, “blended learning…encourage[d] experimentation,” wrote Klein and Douglass in a January 2013 update to the case study. “We see teachers dramatically changing their schedules, grouping structures and habits more frequently and with more excitement and openness than ever before.”

In the second year, “we have a handful (and growing number) of teachers who are truly innovating at each site and are pushing their peers to both play and learn alongside their students,” they wrote in an earlier update.

The foundation’s focus now is on training teachers on programs and tools, connecting them to coaches, and helping teachers collaborate with each other.

Education technology often is used to do the same old thing—only with less paper, says Klein. Technology must change teaching in order to make a difference. That’s the foundation’s goal in Oakland. Yet Klein recognizes that much has stayed the same at the pilot schools. “Our pilots highly disrupt the status quo at schools, but it’s the same teaching staff ratio, grade levels, and bell schedule.”

“We’re not necessarily funding next-generation models,” says Klein. “We’re less funding the future and more funding the breakdown of the past. It’s the beginning of the end of the factory model.”

“Next-generation learning” will go much deeper, Klein predicts. But, first, he wants to show that the blended pilot is helping students learn more.

“In education, we took down the chalkboards and put up the whiteboard, but it probably didn’t make a lick of difference in academics,” says Klein. “Now we’re moving to student-facing devices. We’re hopeful we can help Oakland do it well. How can we transfer teaching and learning in a daily way in core classes?”

Blended learning is “not transformational right now,” he says. He thinks it will be.

Joanne Jacobs, a former San Jose Mercury News editorial writer and columnist, writes about K–12 education and community colleges at joannejacobs.com and ccspotlight.org.

Lost Classroom, Lost Community: Catholic Schools’ Importance in Urban America
By Margaret F. Brinig and Nicole Stelle Garnett
University of Chicago Press, 2014, $45; 224 pages.

As reviewed by Nathan Glazer

Thirty years ago, research by James Coleman, Anthony Bryk, Andrew Greeley, and others showed surprisingly that Catholic inner-city schools were enrolling black students and educating them effectively. Only a small minority of American blacks are Catholic, but black parents were sending their children to Catholic schools, despite tuition fees and instruction in a religion the parents did not themselves follow. These parents made this decision primarily because of the schools’ reputation for effective discipline, and in reaction to the observable disorder of their local public schools.

Meanwhile, as Catholics, along with others, left the central cities for suburbia, Catholic schools were losing their local Catholic parish students. As the costs of maintaining these schools rose, dioceses and archdioceses were forced to close many of them, often against fierce local opposition. Catholic schools also lost their traditional teaching force. Fewer young women were taking up the religious life, and the nuns were replaced by much more expensive lay teachers. Catholic churches are territorially fixed, that is, tied to a defined parish; they do not move along with their constituents, as Protestant churches and Jewish synagogues and temples can and do. Emptying out, the parish schools were forced to either close or adapt to a changing local population, which could not provide the same financial support as those departing.

So, ironically, the high reputation of Catholic schools for educating black children in the inner cities was accompanied by an increasing financial crisis, which radically reduced their number. In Chicago, the major site of the research reported on in Lost Classrooms, Lost Community, 130 Catholic schools closed or were merged with others between 1984 and 2004. The Archdiocese of Chicago educated 300,000 students in 1965; by 2012 that number was reduced to 87,000. Of these, 22 percent were not Catholic, and 30 percent were “racial minorities” (one assumes Latino as well as black). In the major cities of the Northeast, where Catholic immigrants from Europe had settled in large numbers in the late 19th and early 20th centuries, the reduction in the number of Catholic schools was similar. (In Philadelphia, the second site for research reported in this book, enrollment peaked at 250,000 in 1961 and totaled just 68,000 in 2011.)

Brinig and Garnett assume Catholic schools’ effectiveness with lower-income inner-city children and take up a different story: the effect of school closures on the quality of the social life in their surrounding neighborhoods. The authors lean heavily for this analysis on the concept of “social capital,” which for James Coleman explained the educational effectiveness of these schools. “Social capital” is a murky concept, and Brinig and Garnett make heroic efforts to give it substance. It may be considered as a “factor of production,” paralleling the use by economists of physical capital, financial capital, and human capital (the effectiveness of the individual, defined by education) as factors in the production of goods. But whereas financial and human capital can be given specific measures and meaning, this is not so easy with social capital. Social capital is best envisioned as something that helps produce a good society: less crime, less disorder, more trust.

Robert Putnam, in Bowling Alone, defined social capital as “social networks and the norms of reciprocity and trustworthiness that arise from them.” Brinig and Garnett argue that “Catholic schools in Chicago appear to bolster neighborhood social cohesion and suppress crime—findings that, we hypothesize, stem from the fact that these schools generate social capital…. neighborhoods experiencing a Catholic school closure during the relevant time period were more disorderly, and less socially cohesive, than those that did not.” The originality of the book is that this thesis, which could be argued anecdotally from many cases, is based on statistics, relying primarily on survey data collected by the Project on Human Development in Chicago Neighborhoods—on detailed observational data collected by the project, block by block, of the state of order or disorder on the block—and on crime data, at the police-beat level. The study is replicated, insofar as data are available, for Philadelphia, which seems to tell the same story, and for Los Angeles, which does not seem to be a parallel case. But of course Los Angeles’s Catholic population is so different in ethnicity, and in its relation to the Catholic Church, from the European ethnic groups that have made up most of the Catholic population of the large eastern and midwestern cities, that one would not expect the same effects of school closings.

But how is one to make the case? The areas in which Catholic schools are closing are also areas in which overall population may be declining, black and minority population increasing, and poverty rising. Do these not explain the increase in crime and disorder and the decline of trust? The authors make the ingenious argument that they can detect the distinct influence of the closing of a Catholic school because such events are not related only to the increase of poverty and the growth of minority populations. Which schools in such areas close, they argue on the basis of detailed knowledge of how Catholic schools operate, depends on the commitment of the pastor of the parish. Those more committed to the school and parish will fight harder to enable them to remain open. And the authors argue they can measure commitment by a number of factors: how old the pastor is, how long he has served in the parish, whether he has been ordained in the archdiocese or belongs to a religious order and is subject to its authority, whether there is some irregularity—like a sexual abuse charge—involving the pastor, and whether a temporary administrator heads the parish.

So one has a chain of statistical arguments: which schools close depends on the pastor, not only on the social conditions in the parish, and the neighborhoods of schools that close suffer a loss of social capital. Their chain of arguments may not convince everyone, but their hypotheses on the effect of pastoral leadership, independent of the percentage of Latino, black, and in poverty, is borne out by the statistical analysis. The authors write, “The factor that most predicted whether a school would close—more so than income or race—was whether there was something ‘irregular’ about the parish leadership….”

Will other schools that are free of the bureaucratic net of the public school system have such an effect? While Catholic schools were closing, the number of charter schools was increasing, and various states were setting up voucher programs for low-income students to attend (some) private schools. The question is particularly pertinent because in some states Catholic schools have been reorganized as charter schools, educating in the same building, with many of the same students and teachers. The authors tell a number of stories of such transformations, which do “save the school.” The authors consider the effect of the opening of charter schools in Chicago on their neighborhoods and whether they have the same “social capital” effects, but these schools are too various, and the time since their opening generally too short, for any clear conclusion to be drawn. And there is no way for them to disentangle the effect of a charter school, as they could with Catholic schools, by using the commitment of the local pastor as an independent variable. But one suspects that the effect of the Catholic school on its neighborhood is unique, as the commitments between school, teachers, administrators, parents, and students are strengthened by residence in the same neighborhood, as well as the tie of a common religion binding many of them.

If Catholic schools have such effects, one must raise the question, and the authors do, of why these schools cannot get public funds: they do as good a job of educating their students as public schools, perhaps better. But while voucher programs, for example, have overcome legal prohibitions in some states, political resistance to the flow of substantial public funds to schools not under the control of districts remains intense.

Nathan Glazer is professor emeritus of education and sociology at Harvard University.

On January 8, 2014, the Civil Rights Division of the Department of Justice (DOJ) and the Office for Civil Rights in the Department of Education (ED) issued a joint “Dear Colleague Letter” to K–12 schools. The topic discussed in their joint letter is whether administrators are punishing minority children more harshly than white children for the same infractions. Unfortunately, the Dear Colleague letter does not squarely address that issue. Instead, it offers instructions on how schools should strengthen oversight of their disciplinary processes in order to meet their obligations under Title IV and Title VI of the Civil Rights Act, as applied to discrimination on grounds of race. The agencies advise,

The administration of student discipline can result in unlawful discrimination based on race in two ways: first, if a student is subjected to different treatment based on the student’s race, and second, if a policy is neutral on its face—meaning that the policy itself does not mention race—and is administered in an evenhanded manner but has a disparate impact, i.e., a disproportionate and unjustified effect on students of a particular race.

Much of the analysis turns on the word “unjustified.” Disproportionate rates should not be regarded as unjustified merely because they reflect higher rates of improper behavior by minority students than by white students. But this point is never explicitly acknowledged in the ED and DOJ guidance, which was prompted by their joint dissatisfaction with zero-tolerance policies. Such policies, by definition, impose the same penalty for a given infraction, regardless of the student involved. When implemented in some school districts, the policies have led to levels of school suspension and expulsions that differ markedly by race. According to data collected by the Office for Civil Rights and cited in the letter, black students, who make up about 15 percent of the student population nationwide, receive about 35 percent of one-time suspensions and 36 percent of expulsions from school. DOJ and ED offer no estimate of the number of school districts that experience these problems, or the number of schools or students affected. No matter what the aggregate numbers demonstrate, imposing a nationwide policy would certainly compromise the operation of schools that have no taint of any purported civil-rights violation. Nonetheless, the letter applies to all school districts indiscriminately and requires them to put in place expensive compliance systems.

Zero-tolerance policies in school give straightforward discipline guidance to administrators by removing discretion in the enforcement of a rule. They can also, in some instances, lead to punishments that fail to “fit the crime.” Wholly apart from the Civil Rights Act of 1964, it is fair to ask whether zero tolerance makes sense in an educational context. But the DOJ and ED err in arguing against the policies on the basis of “disparate impact,” a phrase with a legal history that, when applied to schools, imputes race-conscious behavior on the part of school administrators. For one thing, zero tolerance has no monopoly on potential violations of the civil rights laws. The switch to a policy that substitutes lesser sanctions for suspension or expulsion raises the same civil-rights problem if the incidence of disciplinary action also captures a fraction of black students that is larger than their proportion of the student body.

In this article, I examine some of the practical and legal issues that informed the preparation of the guidance letter from ED and DOJ, including the zero-tolerance and disparate impact theories that form its foundation. I look at the legal issues that emerge from the aggressive interpretation of Title IV and Title VI of the Civil Rights Act and also at the procedures by which ED and DOJ seek to impose their proposals.

Zero Tolerance

The ED and DOJ guidance letter takes particular aim at zero-tolerance policies as implemented by school districts. The application of particular disciplinary policies in schools surely matters: some 3 million students in grades K–12 were suspended, for example, in 2009.

The notion of zero tolerance has considerable political support. The straightforward link between an infraction and its consequences gives fair warning to students what the consequences will be should they break a particular rule. By limiting administrator discretion, zero-tolerance policies make it harder for administrators to play favorites than under a system that requires them to adjust punishment according to the specific facts and circumstances. Whether such policies improve student behavior or school climate overall is a separate question. Hard-and-fast rules may reduce the rate of infraction so much that it is easier to accept the occasional instance of a punishment that seems far too severe. It may also be that where rules are frequently breached, officials are hesitant to follow through and backtrack on enforcement or just pretend that the minor offenses never occurred in the first place.

Their history shows that zero-tolerance policies work best in those settings where a clear on/off switch triggers the relevant violation, such as a policy that requires suspension of any student who brings a gun or drugs into school. The rule gives all students fair notice and encourages them to behave appropriately. The detection of a weapon will in most cases not present any questions of degree. In practice, the on/off switch required by the rule maps well into the on/off nature of the penalty.

The gains from such a system may be substantial. Removing a difficult student from the classroom may make the environment safer for other students, who now have less to fear from threats of rule breakers. For these cases, there appear to be good reasons for invoking a hard-edged rule. But even in these cases, generalizations could prove hasty in the absence of thorough knowledge of alternative strategies. It is very difficult for any outsider—even ED and DOJ—to make a definitive judgment that would be applicable to all schools.

Zero-tolerance policies look less attractive when applied to offenses that are less easily defined, such as harassment. The phrases “unwelcome advances” and “hostile environment” leave a lot of room for interpretation. The same can be said about rules on hate speech, which covers “bigoted speech, which attacks or disparages a social or ethnic group, or one of its members.” The high probability of borderline cases may benefit from administrators having discretion in the penalties that they impose.

The hard question is, quite simply, why do zero-tolerance policies come under attack under the civil rights laws? One possible explanation is that the application of these rules is “overbroad,” and leads, as the New York Times has argued, to serious and lasting consequences for the students who are suspended or expelled. They can face an added risk “of being held back, dropping out or ending up in the criminal justice system.” Such race-neutral effects of suspension and expulsion may be sufficient to justify a modification or elimination of the policy. But even that judgment cannot be made solely by looking at the impact of these rules on the punished students. One must also factor in the effect that keeping difficult students in school has on other students, many of whom may now be subject to increased probabilities of being held back, dropping out, or ending up in the criminal system. It is the net effect that matters.

This net effect is systematically ignored by the ED and DOJ in promulgating their new civil-rights guidance. The government points only to studies that show the proportion of black students who receive certain penalties to be greater than their proportion of the students in the government’s data set.

At this point, it is a real puzzle to decide what should be done if every single permutation of discipline policy shows differential impact by race. Clearly, it cannot be that all discipline procedures must be put on hold, especially in schools with the greatest discipline problems. What then is the alternative?

Disparate Impact

Something is amiss in the ED-DOJ analysis. The agencies’ joint error rests on the peculiar way in which they apply the standard of disparate impact. The origin of the disparate impact test was the 1971 Supreme Court case of Griggs v. Duke Power Co., decided under Title VII of the Civil Rights Act, which governs employment relations. That decision marked, to say the least, a major expansion of the scope of the 1964 act and was driven in large measure by the fear that using only a disparate-treatment test (which entailed proof of some form of race-conscious behavior) would allow discrimination by private firms with a history of discrimination by race to slip through the cracks. The unanimous Supreme Court decision in Griggs was prepared to tolerate some overenforcement of the civil rights law in employment cases, lest underenforcement allow too many wrongdoers to escape these rules. Accordingly, under a key 1982 decision in Connecticut v. Teal, employers could not administer various educational and achievement tests that had strong predictive value within both white and black populations to rate job candidates across races—or indeed within the group of black applicants.

We are of course in very different circumstances today, for the days of systematic forms of race discrimination defended at the highest level of government are mercifully behind us. The disparate-impact standard therefore is only weakly justified on the grounds that it is necessary to uncover hidden forms of unconscious race bias against minority students. But it is imperative that the ED and DOJ show more than differential punishment rates by race to establish some latent patterns of discrimination in their studies.

One source of the difficulty is that the data ED and DOJ use to justify their guidance are aggregated across schools and therefore provide no insight into the kinds of behavior that are observed inside particular schools. To see the problem, assume for the moment that there are two schools, one with an all-white population and the other with an all-black population. Assume further that the rate of suspension and expulsion in the all-white school is only one-half of what that figure is in the all-black school. The only inference that can be drawn from these statistics is that, on average, the discipline problem is more severe in the all-black school than in the all-white school. One cannot infer that any administrators treat white and black students differently, because all administrators are dealing with students of only one race.

It could be argued that this unrealistic objection proves little because many schools have students of different races. But even if one school has a student population that is 75 percent black and 25 percent white, and the other has those percentages reversed, there is still no evidence of any form of discrimination if the school with the majority black students has a higher overall discipline rate. That is most obviously the case if the black students and white students are both punished at a higher rate in the majority-black school than in the other school. And it is even the case if a close empirical investigation showed in each school that a scrupulous color-blind system produced differential rates. One could press further and ask whether black or white administrators had different rates of issuing suspensions. If they are the same, again, the disparate impact case is undercut. If they are different, the evidence is still very weak unless there is some evidence of bias. The simple proposition here is that so long as like cases are treated alike regardless of race, discrimination is not to blame for disparate discipline levels.

How should a school go about correcting the imbalance? Just what sanction should apply to a school where discipline is imposed on a color-blind standard yet has statistically imperfect outcomes? Should some white students be summarily suspended, expelled, or otherwise sanctioned to make the numbers come out correctly? Or should schools give a pass to black students who have committed serious offenses in order to achieve the same ends?

The point here is bitterly ironic because U.S. secretary of education Arne Duncan has claimed that the guidance package is needed because it “provides resources for creating safe and positive school climates, which are essential for boosting student academic success and closing achievement gaps.” He conceded that “incidents of school violence have decreased overall,” so why should he advocate policies that have as their likely consequence exposing the good kids who have avoided violence to the greater fear from others who engage in it? The sad truth is that this muddled analysis of zero tolerance and disparate impact is likely to reverse the hard-earned gains of earlier years.

Statutory and Administrative Authority

In its Dear Colleague letter, ED and DOJ refer to Title IV and Title VI of the Civil Rights Act without quoting their language or discussing their scope. The letter insists that Title IV and Title VI protect students over the entire course of the disciplinary process, “from behavior management in the classroom, to referral to an authority outside the classroom because of misconduct—a crucial step in the student discipline process—to resolution of the discipline incident.”

The statutory authority is far less clear than this assertion of authority suggests. Title IV is largely intended to provide schools in the process of desegregating with various forms of personnel advice or hiring specialists and to make grants to achieve that end, topics that are distinct from zero-tolerance policies. Enforcement of the new disciplinary policy does not appear covered by these provisions.

The entire case for government control therefore rests on Title VI, which contains the general prohibition that “No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.” It is a very broad reading to take the last clause—subjected to discrimination—as allowing the government to usher in a comprehensive program in an effort to rectify disparate results of legitimate government actions. Indeed, it is hard to see how the disparate impact standard, which the Supreme Court adopted in Griggs v. Duke Power for employer testing, could be carried over to actions under Title VI. In this context, its application is far more intrusive, given that no school has the option of not engaging in some form of discipline. More specifically, the guidance offers no safe harbor for conducting routine discipline, free of constant federal oversight. Under the Dear Colleague letter, ED has virtually unlimited discretion in deciding, for example, whether “[s]elective enforcement of a facially neutral policy against students of one race is also prohibited intentional discrimination.” Schools will be reluctant to take these cases to court for they always face the risk that ED and DOJ will claim that the observed differences in behavior were themselves caused by racially insensitive or inappropriate district policies.

The Dear Colleague does not put any of these concerns to rest. Instead, the letter gives as illustrations single incidents between two persons of different races where, in the absence of an explanation, differential sanctions can give rise to an inference of racial discrimination. Yet virtually every disputed case could give rise to some inference of actual or intended discrimination. In addition, it is highly likely that any such investigation will extend to cover supposedly similar incidents involving nonminority students. This aggressive ruling therefore pushes enforcement to the outer edge of the legal authority under Title VI.

This general pronouncement itself appears to introduce race distinctions into the system. Although stated in race-neutral terms, none of its carefully selected examples indicate that ED or DOJ will ever intervene on behalf of white students claiming to be victims of discrimination, even though Title VI, which starts with the words “no person” was consciously drafted in race-neutral terms. It is likely that in practice, both ED and DOJ will use different standards in different kinds of cases.

Dangerous Use of Administrative Guidance

The difficulties with this aggressive interpretation of the statutes are further compounded by the use of the guidance mechanism to promulgate these rules. These guidance practices are nowhere mentioned by name in the Administrative Procedure Act (APA), which governs how federal agencies may propose and establish new regulations. The act assumes that the implementation of complex regulatory schemes will be done only after a formal rule-making procedure (which is like a full trial) or more truncated notice-and-comment proceedings, which require the government to first announce its position so that others may comment on it. These procedures were intended to offer interested parties a chance to weigh in on proposed rules before they are put into place. Notice and comment takes time, but by the same token the comments often supply useful information that could, and should, influence the shape of the final regulation.

Guidances are an agency invention to circumvent the notice-and-comment procedures. To the government, guidances are just “interpretative rules, general statements of policy, or rules of agency organization procedure or practice,” which need not meet the procedural requirements for regulations under the APA. If this Dear Colleague letter fits within those terms, then so too does any other massive pronouncement by any government agency.

But woe unto any targeted party that wishes to ignore the guidance on the grounds that it does not enjoy the force of law! At every point in the process, the government retains the options to apply these guidance documents to the recalcitrant, just as if they had been formally promulgated, and the parties who resist them will be dealt with harshly because they have been given notice of what to expect in the future. To be sure, there is an outside chance that some portion of these guidances will be rejected in court, but it takes a strong stomach to risk the heavy sanctions that the government will impose if it prevails, as it usually does now that these guidances are part of the basic regulatory toolkit in every area from food-and-drug to environmental law. To be sure, a school district could seek to clarify the guidance before it has the force of law. But at that point, ED and DOJ (which have long memories) will claim that they do not have standing to attack the law prior to the time of its direct enforcement. In the meantime, the failure to take steps to conform to the requirements of the guidance will count heavily against any school district that refuses to play the game in accordance with the new rules. The guidance will likely be treated as binding for all practical purposes.

In sum, the ED and DOJ action forces school districts to comply with a substantive rule of dubious legal validity and practical soundness. Their “guidance” represents the worst in federal policy on K–12 education. At every stage, the government pushes its case to the limit: It fails to inquire whether zero-tolerance policies are defensible on their own terms. It applies dubious measures of disparate impact to deal with alleged civil-rights difficulties. The DOJ and ED then compound the problem by using the guidance technique to avoid the notice-and-comment protections built into the Administrative Procedure Act. Any legal challenge to these regulations faces an uphill battle, yet judicial efforts may be needed to prevent ED and DOJ from using the civil rights laws to federalize all issues of discipline in the nation’s schools.

Our children deserve better.

Richard A. Epstein is professor of law at New York University, senior fellow at the Hoover Institution, and professor of law emeritus and senior lecturer at the University of Chicago.