The Education Gadfly Weekly

Caroline M. Hoxby, Harvard University and National Bureau of Economic Research September 2004

Harvard economist Caroline M. Hoxby has just issued the most effective rejoinder to the misleading AFT "study" of charter school achievement that's been much in the news of late: she's done a far better study, and it yields far different results. Whereas the AFT relied on a small NAEP sample of charter schools and had no way to compare their students' performance with that of similar youngsters in similar communities, Hoxby obtained 4th grade achievement data for 99 percent of all charter students and compared them with similar test results from the geographically nearest district-operated school or, in some circumstances, a school that's not quite as close in miles but closer in racial composition. Moreover, the measure she used is one that actually counts for children and schools: the percentage of youngsters in a school who attain proficiency on their state's test according to their state's standards (in 2002-3). For the country as a whole, she found, "Charter school students are 3.8 percent more likely to be proficient on their state's reading examination when compared to students in the nearest public school. They are 4.9 percent more likely . . . when compared to students in the nearest public school with a similar racial composition." (The corresponding "charter advantage" in math is 1.6 and 2.8 percent.) On the whole, she concludes, "The average charter school student in the United States benefits from having a charter school alternative." Moreover, charter pupils do better in reading in fifteen states; the same in five; and worse only in North Carolina. (In math, they fare better in eleven jurisdictions; the same in eight, and worse in North Carolina and Texas.) As Hoxby acknowledges, it's not a perfect study; like the NAEP-based analysis, it's a snapshot of student performance at a single point in time and does not show "value added." Nor is it based on randomized assignment of children to charter and district schools. But it's promising enough, as the author remarks, not to blow the whistle on charter schools, as the AFT and their allies would do, but to "make us patient enough to wait for the results of multi-year studies based on random lotteries among charter school applicants." Hoxby herself is working on such studies. In the meantime, you can find this one on the web.

"New data fuel current charter school debate," Education Week, September 8, 2004 (registration required)

"Views on charter schools," with Caroline Hoxby and Howard Nelson, NPR, September 14, 2004

"Charters get high grade on report," by David Andreatta, New York Post, September 8, 2004

Jay Greene and Greg Forster, The Manhattan InstituteSeptember 2004

Jay Greene and Greg Forster of the Manhattan Institute released this study last week, ostensibly an answer to the work of Richard Rothstein and others who argue that social factors make it all but impossible to close achievement gaps via education reform, and that reformers should instead focus on fixing those underlying social problems. To test that theory, the authors create a "teachability index," which uses factors of school readiness, community, race, economics, health, and family to determine how "teachable" students are, then ranks states according to their students' teachability. In addition, using state student achievement data, the authors rank the states according to how efficient their schools are, i.e., how well their students perform when compared with the achievement level predicted by their teachability and state education spending. Interestingly, Greene and Forster find that "States with low scores on the [teachability] index do not inevitably produce low-performing students, and states with high scores do not inevitably produce high-performing students." Thus, they argue, social factors and school spending do not necessarily predetermine the level at which students can achieve, and therefore increasing spending or targeting these social problems will not necessarily yield higher achievement. Critics will quibble with the factors Greene and Forster chose to include in their index and, as the authors note, it's "only a rough indicator of student teachability." One also wonders whether engaging in the "teachability" debate may be counterproductive, given that social inequality should never justify the achievement gaps that too many educators have come to accept. But, the report does provide a counter-argument to those who insist that education reform is impossible so long as social inequality persists - which is to say, in effect, that reform is impossible forever. To check it out, click here.

Sara Mead, Progressive Policy InstituteSeptember 2004

In this concise report, Sara Mead makes a strong case for effective preschool programs. Culling information from multiple studies, she argues that universal access to high-quality preschools could take us a long way toward closing racial and socio-economic achievement gaps. The core of her report is a specific proposal for universal preschool access. This could be pricey, Mead admits, but making it free for poor families and then increasing the burden for others on a sliding scale would be affordable - well, maybe $8.1 billion a year. (A billion here, a billion there....) Although Mead emphasizes the need for state-by-state flexibility in curriculum and other matters, she also thinks this federal money should come with accountability strings attached. Whatever the likelihood of such a program being instituted, Mead has put a coherent model on the table. Policy makers would do well to read her discussions of the preschool programs already implemented in Georgia and Oklahoma. A recent study of the latter, for example, found remarkable gains - 17 and 54 percent, respectively - in the cognitive and language assessment scores of African American and Hispanic children. You can find it here.

Every person in America wants every child in America to have a terrific teacher every year. That much we can assume. Why, then, is it so hard to craft sound policies yielding that universally sought result? Excellent question. My answer is that we've made ten basic mistakes:

1. Out of deference to adult preferences rather than what's best for children, we've opted for quantity rather than quality, for hiring more teachers instead of demanding (and paying for) better ones. NEA data show that, for every 20 additional students enrolled in American K-8 schools over the last 10 years, we hired three more elementary school classroom teachers. Take a longer view and the numbers astound: between 1955 and 2000, the number of K-12 teachers in the U.S. almost tripled while enrollments rose by half. Instead of paying a smaller number of people more money, we opted to pay lots more teachers a more-or-less constant wage. We surely could have found ways to place better educated and better compensated instructors in our classrooms if we hadn't set out to hire so many millions of them. Why did we do that? There's a provocative economist's theory about changing labor markets, particularly for high-ability women (see, for example, Darius Lakdawalla's terrific piece in Education Next.) But there's also a simpler explanation: notwithstanding ambiguous evidence as to its efficacy, both teachers and parents tend to favor smaller classes. So classes keep shrinking - and that means hiring more teachers, able or no.
   
2. We use paper credentials as the gauge of who will be allowed to teach, rather than demanding evidence of subject-matter knowledge and/or the ability to handle a classroom. And we entrust control of entry to universities, many of which do no real screening for individual aptitude, energy, or knowledge and almost all of which have an unquenchable self-interest in accepting large numbers of people into their teacher-credentialing programs. To be sure, any profession demands certain credentials, but they usually stand for some knowledge gained or skill acquired. In K-12 education, we count beans: tallying courses taken and hours spent in class while exhibiting alarming indifference to ability or knowledge. 
   
3. We sorely underestimate the importance of attracting intellectual talent into the profession, excusing this failure by saying there's more to teaching than being smart. That's true to a point, but doesn't mean we oughtn't do our utmost to shape policies that draw exceptional people into the classroom, if only as short-termers. No other profession is as disdainful of academic performance - and what could be more ironic for a profession charged with educating people? Yet because ed schools and teacher prep programs are money makers for universities, we let them enroll weak students who would likely struggle with tougher courses. Is that really who we want teaching our kids? 
   
4. In the name of fostering diversity in the teaching ranks, we avoid noticing the dire quality of many preparation programs that enroll lots of minority students. That oversight perpetuates a devastating cycle of underperformance and keeps the learning gap wide at the very time we should be directing at least as much energy to closing it for teachers as we're now doing for primary-secondary pupils. 
   
5. Given higher education's stranglehold over teacher preparation these many decades, one might suppose that by now the preparation process would rest on a bounteous and sturdy research base. Wrong. Our understanding of the sources and attributes of teachers' classroom effectiveness sits upon a stunningly weak foundation, which in turn cripples efforts to improve teacher identification, preparation, and evaluation. Worse, where we have robust evidence as to what makes teachers effective (e.g. in beginning reading), our campuses often shun or deprecate it. (If medical training were similar, doctors would still swear by leeches and mustard plasters.)
   
6. We pay (and treat) teachers uniformly rather than distinguishing among them on the basis of (among other things) effectiveness, specialty, and work environment. And when we evaluate them at all, we base it on peer and supervisor impressions of their work rather than evidence of the value they add to children or the results they produce in class. 
   
7. We don't do any quality control, except via paper credentials at the point of entry, and we make it extremely difficult to move (much less remove) bad teachers or reward good ones. We confer tenure on teachers prematurely and automatically, linking it to time on the job rather than effectiveness. Then we hold it inviolate.  
   
8. We treat teaching's "short-termers" as system failures rather than human assets. Public education's personnel norms still assume that a person should spend his/her entire career in the classroom, notwithstanding evidence that today's young people in every field are career hoppers. In other words, a system designed for the "lifers" of our parents' generation has not been updated. (Hence we also have no proper career path for long-termers except out of the classroom and into administration.)
   
9. We don't make key teacher hiring, assignment and retention decisions where they should be made - at the building level - even though that's where critical judgments can be made about a person's suitability for teaching specific content to particular children. Nearly everything about a teacher that's important and that cannot be gauged through objective means is best gauged by that teacher's future colleagues and immediate superiors. Yet we deploy teachers to schools via downtown personnel offices that are constrained by union contracts, seniority, tenure, and state credentialing. Worse, school principals who ought to demand full personnel authority often decline to do so, seemingly preferring the impersonal safety of having someone else make these decisions. 
   
10. Instead of better salaries, we spend top dollar for generous benefits but often in dysfunctional ways, such as not letting teacher pensions be portable. Rather than paying teachers better and letting them control their own benefits and shape their own pensions, we allocate huge percentages of school budgets for old-fashioned retirement systems that relatively few teachers ever benefit from. 

Are these errors fixable? I don't know. But I'm sure that the starting place is to recognize mistakes we're making now - and have made for so long that we take them for granted rather than as problems searching for solutions.

Minnesota Governor Tim Pawlenty announced this week that three schools in the Land of 10,000 Lakes will pilot a new teacher pay-for-performance plan that he hopes will pave the way toward more ambitious merit pay schemes down the road. According to the Minneapolis Star-Tribune, the plan includes: $5,000 for teachers who serve as mentors; $8,000 in raises for master teachers, who assist with teacher training, evaluation and student achievement analysis; and $2,500 to $3,000 in bonuses awarded to teachers based on evaluations of their teaching and how much they boost student achievement. Pawlenty, whose more ambitious proposal to pay "super teachers" up to $100,000 a year (read more here) died in the state's 2004 legislative gridlock, is optimistic about this first step. "My view is that we want to change the whole system. But you can't do that overnight. . . . We want to move more toward a performance-pay plan, and this is a step in the right direction."

"State will test teacher merit pay," by Norm Draper, Minneapolis Star-Tribune, September 14, 2004

The unfortunate saga of the California Charter Academy, closed for various improprieties last month (click here and here for more), has something like a happy ending. Of the 3,300 K-12 students left stranded by CCA's closure (far fewer than the 10,000 originally claimed), 2,600 have been enrolled in other local charter schools. This good news comes after a Herculean effort by Caprice Young's California Charter Schools Association, which vowed to find openings for as many of those students as possible. It's heartening to know that some people, amidst the mess and recrimination that followed CCA's implosion, kept their eye on the kids.
 
"It's back to school for 2,600 displaced students," by Erika Hayasaki, Los Angeles Times, September 14, 2004

Ask most parents to rate the American education system compared to the rest of the industrialized world, and they'd no doubt tell you it is second to none. According to the 2004 edition of the Organisation of Economic Cooperation and Development (OECD)'s Education at a Glance report, however, this basic assumption of American educational excellence has grown shaky. (To read our take on last year's results, see "Facing Facts.") The U.S. does still have the world's highest college completion rate among those aged 44+. But among ages 25-44, America now ranks 10th - not because our numbers are decreasing, but because others are catching up. Perhaps more troubling, the U.S. ranks just slightly above average in reading performance, below average in math achievement, and below average in high school completion rates, compared to the rest of the industrialized world. This despite spending more per student on all levels of education. Barry McGraw, OECD director of education, observes that "The best in the U.S. is as good as the best in the world. What drags you down is the worst performers."

Education at a Glance 2004, Organisation for Economic Cooperation and Development

"Report: U.S. slips in education ratings," by Ben Feller, Associated Press, September 14, 2004

"U.S. teens have big hopes, average skills," by Greg Toppo, USA Today, September 14, 2004

The American Enterprise Institute's Frederick M. Hess and our own Checker Finn have three new and overlapping analyses of the No Child Left Behind Act, with particular reference to that law's accountability and choice provisions. The book version, which was edited by Hess and Finn and contains contributions from fifteen other authors, is Leaving No Child Behind? Options for Kids in Failing Schools. If you haven't got time for the full volume, you can read its major conclusions in the September edition of Phi Delta Kappan (click here for more). And, if you've not yet gotten your fill, check out the fall 2004 edition of The Public Interest, where they examine NCLB in broader perspective in "On Leaving No Child Behind." The bottom line: this historic law can work and must work, but there's no point in stubbornly insisting that it will work perfectly as first enacted (or that more money is the answer). Big, complex laws and programs normally need calibrating. In this case, some NCLB provisions should be tightened, others made more flexible. These studies are by no means the last word; they are, in fact, primarily glimpses of and reflections on the law's early implementation.

Leaving No Child Behind? Options for Kids in Failing Schools, Frederick M. Hess and Chester E. Finn, Jr., Palgrave MacMillan, ISBN: 1403965889

"Inflating life rafts of NCLB," Frederick M. Hess and Chester E. Finn, Jr., Phi Delta Kappan, September 2004 (subscription required)

"On Leaving No Child Behind," Chester E. Finn, Jr. and Frederick M. Hess, The Public Interest, Fall 2004 (The Fall 2004 edition is not yet available online.)

Caroline M. Hoxby, Harvard University and National Bureau of Economic Research September 2004

Harvard economist Caroline M. Hoxby has just issued the most effective rejoinder to the misleading AFT "study" of charter school achievement that's been much in the news of late: she's done a far better study, and it yields far different results. Whereas the AFT relied on a small NAEP sample of charter schools and had no way to compare their students' performance with that of similar youngsters in similar communities, Hoxby obtained 4th grade achievement data for 99 percent of all charter students and compared them with similar test results from the geographically nearest district-operated school or, in some circumstances, a school that's not quite as close in miles but closer in racial composition. Moreover, the measure she used is one that actually counts for children and schools: the percentage of youngsters in a school who attain proficiency on their state's test according to their state's standards (in 2002-3). For the country as a whole, she found, "Charter school students are 3.8 percent more likely to be proficient on their state's reading examination when compared to students in the nearest public school. They are 4.9 percent more likely . . . when compared to students in the nearest public school with a similar racial composition." (The corresponding "charter advantage" in math is 1.6 and 2.8 percent.) On the whole, she concludes, "The average charter school student in the United States benefits from having a charter school alternative." Moreover, charter pupils do better in reading in fifteen states; the same in five; and worse only in North Carolina. (In math, they fare better in eleven jurisdictions; the same in eight, and worse in North Carolina and Texas.) As Hoxby acknowledges, it's not a perfect study; like the NAEP-based analysis, it's a snapshot of student performance at a single point in time and does not show "value added." Nor is it based on randomized assignment of children to charter and district schools. But it's promising enough, as the author remarks, not to blow the whistle on charter schools, as the AFT and their allies would do, but to "make us patient enough to wait for the results of multi-year studies based on random lotteries among charter school applicants." Hoxby herself is working on such studies. In the meantime, you can find this one on the web.

"New data fuel current charter school debate," Education Week, September 8, 2004 (registration required)

"Views on charter schools," with Caroline Hoxby and Howard Nelson, NPR, September 14, 2004

"Charters get high grade on report," by David Andreatta, New York Post, September 8, 2004

Jay Greene and Greg Forster, The Manhattan InstituteSeptember 2004

Jay Greene and Greg Forster of the Manhattan Institute released this study last week, ostensibly an answer to the work of Richard Rothstein and others who argue that social factors make it all but impossible to close achievement gaps via education reform, and that reformers should instead focus on fixing those underlying social problems. To test that theory, the authors create a "teachability index," which uses factors of school readiness, community, race, economics, health, and family to determine how "teachable" students are, then ranks states according to their students' teachability. In addition, using state student achievement data, the authors rank the states according to how efficient their schools are, i.e., how well their students perform when compared with the achievement level predicted by their teachability and state education spending. Interestingly, Greene and Forster find that "States with low scores on the [teachability] index do not inevitably produce low-performing students, and states with high scores do not inevitably produce high-performing students." Thus, they argue, social factors and school spending do not necessarily predetermine the level at which students can achieve, and therefore increasing spending or targeting these social problems will not necessarily yield higher achievement. Critics will quibble with the factors Greene and Forster chose to include in their index and, as the authors note, it's "only a rough indicator of student teachability." One also wonders whether engaging in the "teachability" debate may be counterproductive, given that social inequality should never justify the achievement gaps that too many educators have come to accept. But, the report does provide a counter-argument to those who insist that education reform is impossible so long as social inequality persists - which is to say, in effect, that reform is impossible forever. To check it out, click here.

Sara Mead, Progressive Policy InstituteSeptember 2004

In this concise report, Sara Mead makes a strong case for effective preschool programs. Culling information from multiple studies, she argues that universal access to high-quality preschools could take us a long way toward closing racial and socio-economic achievement gaps. The core of her report is a specific proposal for universal preschool access. This could be pricey, Mead admits, but making it free for poor families and then increasing the burden for others on a sliding scale would be affordable - well, maybe $8.1 billion a year. (A billion here, a billion there....) Although Mead emphasizes the need for state-by-state flexibility in curriculum and other matters, she also thinks this federal money should come with accountability strings attached. Whatever the likelihood of such a program being instituted, Mead has put a coherent model on the table. Policy makers would do well to read her discussions of the preschool programs already implemented in Georgia and Oklahoma. A recent study of the latter, for example, found remarkable gains - 17 and 54 percent, respectively - in the cognitive and language assessment scores of African American and Hispanic children. You can find it here.