The Education Gadfly Weekly

Brian Gill, Ron Zimmer, Jolley Christman, Suzanne Blanc
RAND Corporation
February 2007

In 2001, Gadfly asked, "Will Edison be able to turn around Philadelphia's schools?" More than five years later, the RAND Corporation has an answer--sort of. Using longitudinal student-level testing data (district reading and math scores from 2002-2006) RAND sought to determine whether Philadelphia's so-called "diverse provider" plan--which enlisted Edison, among other private managers--has benefited that city's students. To recount: in 2001, Pennsylvania took over the School District of Philadelphia (SDP) and invited seven private and nonprofit organizations to manage 45 troubled schools. Another 21 schools were restructured but remained under district management, and a final 16 were only given increased funding. RAND crunched test data for all three groups and found that only the restructured, district-controlled schools showed a statistically significant differential effect on student achievement. RAND also found little evidence that competition from the privately managed schools boosted overall district gains. Accordingly, the authors conclude, "we find no evidence of differential academic benefits that would support the additional expenditures on private managers." But caveats abound. First, "the Philadelphia model was characterized by little competition among providers and by the absence of parental choice among the educational models offered," while "continued district involvement in provider schools... constrained provider autonomy" (see here). Second, "the private managers may be producing other benefits that are not measurable in terms of student achievement" (see here, for instance). Further, former Pennsylvania schools chief Charles Zogby argues in another forum that private managers were not, in fact, awash in extra cash as the RAND report suggests. It seems that more time and deeper changes will be required to reach a true verdict on the diverse provider model. Read the RAND report here. 

Jay P. Greene and Marcus A. Winters
Manhattan Institute's Center for Civic Innovation
January 2007

Jay Greene and Marcus Winters want to recast the way the nation looks at teacher pay: i.e., that it's more important to focus on how teachers are paid than how much they're paid. They do this by making two points. First, that teachers aren't paid too shabbily. The authors look at information from the U.S. Bureau of Labor Statistics regarding the hourly pay of public school teachers. Teachers make, on average, 36 percent more than the average non-sales white-collar worker--architects make 11 percent less than teachers, psychologists 9 percent less, and editors and reporters (gulp) 24 percent less. Their second point is that no correlation exists between higher teacher pay and higher student achievement (measured by high school graduation rates). Correlations between merit bonuses and higher student achievement have been observed, though. Thus, if raising student achievement is schools' top priority, teacher bonuses can and should be distributed with that priority in mind. The authors do not insinuate that good teachers shouldn't be well-compensated, or that the best teachers shouldn't make big bucks, but the emphasis needs to be on rewarding with bonuses those whose students are making classroom gains. Across-the-board pay raises are well and good, and we certainly have lots of them, but they do little to help students learn more. Find the report here.

Merit pay is controversial, especially when tied to student test scores. But if you really want to engulf the teacher lounges in acrimony, make the list of individual bonus winners (and losers) public for all the world to see. That's what Houston did, and now districts in Florida are pondering whether, due to a combination of the state's new STAR plan and its longstanding sunshine laws, they will need to do the same. A blunt-speaking union official made this prediction about what will happen if they do: "On Day 1, you're going to publish the 2,100 people in Pinellas County who got the money." And on day two, "the excrement will hit the rotary." Parents, on the other hand, will get valuable information about the performance of their child's teachers--information that no doubt will place the most effective instructors in high demand (how to deal with that demand is another issue). Some teachers who can't stand the heat will get out of the kitchen; the kitchen will be better for it. But it's unclear how the state's elaborate but imperfect bonus system will weather the harsh glare of public attention, and whether the new governor will support it (initial signs are promising). Of course, there's another way to reward great teachers without all the combustion: give principals the discretion to quietly award bonuses to their top performers, based mostly on student test-score gains, as most managers in America can do. Same result, less drama. 

"How good is teacher? Bonus plan may tell," by Ron Matus, St. Petersburg Times, February 3, 2007

This past week, as the temperature in Chicago dipped below zero, the Chicago Tribune's editorial board warmed itself by the ed reform fire. In an inspired three-part series, the editors explained how Illinois can make educators "more accountable for their good or bad performance, more transparent about how they spend our dollars, and more willing to embrace rigorous, research-driven strategies for teaching our kids." To improve teacher quality, they recommend: 1) stepping up mentoring programs and 2) introducing "private sector standards" into the teaching profession--e.g., merit pay, pension reform, and tenure reform. To fix the classroom they would 3) reduce class sizes, 4) extend the school day and year, 5) provide more tutoring for struggling students, and 6) embrace testing for diagnostic purposes. Saving the best for last, they call for the state to 7) lift its charter cap (currently at 60) and fund charters and traditional public schools at equal levels. To top it all off, schools would receive extra state funds for each of these seven reforms they adopted, and they would have to produce results to keep that funding. The Tribune editors have it right. Do the policymakers?

"In return for more money," Chicago Tribune, February 2, 2007

"Classroom ideas that work," Chicago Tribune, February 4, 2007

"5 more great classroom ideas," Chicago Tribune, February 7, 2007

Supporters of traditional math instruction were dealt a blow recently when the What Works Clearinghouse released its evaluation of elementary math curricula. Of the four programs that WWC examined, Everyday Math, a popular course-of-study with progressives, turned out to be the only one that had "potentially positive effects" on student achievement. Saxon Math, the darling of traditionalists, was said to have "no discernible effects." (For a quick introduction to how Everyday Math and Saxon Math differ in their approaches to teaching basic algorithms, see here.)

The findings came at a bad moment, from the traditionalist perspective, because the National Math Panel (NMP), established last April by President Bush, is currently trying to sort out whether the traditional or progressive approach to math instruction is more effective. As with the National Reading Panel a decade earlier, the NMP is supposed to base its conclusions on the "best available scientific evidence."

Problem is, there's very little decent evidence in this field.  And less if you have high standards, like the WWC when deciding which studies deserve consideration. WWC found just four studies upon which to base its assessment of Everyday Math and just one for Saxon Math. And even these have their problems.

That there is little good research on math curricula that work is no news to the NMP, which at its first meeting acknowledged this fact and the challenge that it posed in drawing conclusions. To deal with this, traditionalists have encouraged the panel to consider studies from other countries, too.

At least one ex officio member of the panel, Daniel B. "Dan" Berch, has urged caution in referencing these studies. He remarked at the initial meeting that "simply mapping data taken from other countries onto practices used in the United States [is problematic], since certain procedures can be appropriate in one context, and inappropriate in another."

Yet most panel members appear open to considering international studies. For example, there has been a spirited discussion at NMP events about Singapore Math, a traditional approach that stresses memorization of basic algorithms and developing computational competence at an early age.

The Singapore program is a good model to examine, but it's not the only one. In fact, there's a world of relevant research on math education beyond the Pacific Rim, and much of it points to the effectiveness of traditional math instruction.

Consider Norway. Between 1995 and 2003, its performance on TIMSS (and PISA) plummeted. That produced a lot of hand-wringing, and several interesting reports. Two of these were released in English in abridged form last February (find them here).

While both studies recognize that Norway's problems are multifaceted, both also place significant blame for that nation's math decline squarely on L 97-new national standards that Norway adopted in 1997 that did away with traditional math teaching and fully embraced progressive techniques.

L 97 stresses "learning by doing." And this, the reports' authors write, is the problem, because "‘doing' [becomes] ‘confusion' instead of ‘learning.'"

Other nations to consider are England, Cyprus, and Slovenia, each of which scored a major jump in math performance on TIMSS between 1995 and 2003.

The NMP deserves a lot of credit for pursuing the very best evidence about math education, no matter how far from our nation's shores it must travel to find it. Think globally, act locally.

California Assemblyman Joe Coto apparently believes that the Golden State doesn't have enough problems getting kids to graduate from high school; he wants to tack on more requirements for the diploma. That's fine when the requirements are academic. But Coto's proposal is simply Big Brotherism; he would require high school students to register to vote before they receive a diploma. (Individuals could opt out of the provision by submitting a written request to do so.) We've got nothing against voting, and Coto's heart is in the right place. But charging schools with solving all society's problems is a flawed approach. As long as we're mandating virtuous impulses, though, Gadfly has a few proposals of his own. In order to receive high school diplomas, students must: Provide proof that they floss daily (this may be in the form of a signed dentist's note, or by turning in hundreds of used strings of floss); write a one-page essay about 1) the virtues of temperance, or 2) why it's important to respect one's elders (essays may not be sarcastic); and sign pledges that they will abstain from 1) sex, 2) drugs and alcohol, and 3) wishing to emulate Lindsey Lohan. Or, in lieu of the above, students may opt to do none of it at all--but only if they put their requests in writing.

"Link diploma to voting, says S.J. legislator," by Edwin Garcia, San Jose Mercury News, February 1, 2007

And you thought trying to compute high school graduation rates was complicated. Try figuring out the percentage of students who need "remedial" work once they enter the hallowed halls of higher education. Estimates range from a low of 42 percent of students in two-year institutions and 20 percent in four-year schools, to as many as 60 percent of community college students and 60 percent of university students, too (according to U.S. Education Department and university studies). Why the discrepancy? Probably because there's no definition of "remedial." Moreover, schools employ a legion of tests to decide which students need extra work (California's community colleges, for example, use more than 100 different tests), which is costly, too. So a student at, say, the University of Wisconsin who is deemed ready for credit-bearing college-level courses may be tagged as a remedial case at the University of Minnesota. What's the answer? Stanford's Michael Kirst has a good thought: "secondary and postsecondary education systems need to create a process to define and measure remediation based on curriculum content and assessment standards for specific subjects." Sounds like exactly what the American Diploma Project creates the possibility for, and why Achieve is working with states to adopt it.

"Who Needs It?: Identifying the proportion of students who require postsecondary remedial education is virtually impossible," by Michael Kirst, National Crosstalk, Winter 2007

School reforms come and go. But educational determinism, it appears, goes on forever. By which I mean the view that schools are essentially powerless to accomplish much by way of learning gains, no matter what is done to or about them. That is because--take your pick--they don't have enough money/time/experienced teachers; or students face so many problems in their lives that it's folly to expect schools to do more with them; or kids lack the innate ability to acquire more skills or deeper knowledge, regardless of how their schools may change.

That's three different arguments, of course. The first is most apt to come from educators and experts who contend--forget four decades of post-Coleman evidence to the contrary--that there's some sort of linear relationship between what goes into schooling by way of resources and what comes out by way of learning. Hence if we crave more of the latter we had best cough up more of the former.

The second--I've long thought of it as the "Gee, Officer Krupke" argument--is typically heard from well-meaning liberals (e.g., Richard Rothstein) who earnestly want income to be redistributed, health care provided, families propped up, racial barriers eased, and so on. They see kids, especially disadvantaged kids, facing non-school challenges that swamp what schools can accomplish. Solve those larger societal problems, they say, and educational achievement will flourish. (The latest Quality Counts from the publishers of Education Week contains a whiff of this. And one often hears something like it from educators who contend that any achievement shortfalls are really the parents' fault.)

The third form of determinism, most prominently associated with Charles Murray, holds that IQ is destiny--and is immutable. With half of everyone having below-average intellect, he writes in the Wall Street Journal and elsewhere (see here), schools are fruitlessly trying (as the schoolteacher wife of a prominent education-reform governor once said to me) to cram a quart of learning into a pint pot.

Chocolate, vanilla, or strawberry, it scarcely matters which flavor of fatalism you select. All send the same signal: that standards-based reform in general, and NCLB in particular, are doomed. That school choice can't accomplish much, either. Indeed, that nothing within the realm of education policy and practice, or within the control of schools and those who work in them, is really capable of producing significant achievement gains.

I beg to differ. Education reform is all about making schools more effective and productive so that kids do in fact learn more. That's no fantasy, even if the determinists want us to think it is. Rather, it's the reality in hundreds of high-performing, high-poverty schools across the country, be they district, charter, or private. They all start with more or less the same sorts of kids with the same sorts of problems and with similar cognitive capacities. Yet they produce markedly better results. They may not get to 100 percent proficiency, but they get a heckuva lot closer than your typical school serving disadvantaged youngsters.

No, we don't have nearly enough super-schools today. That's the education reform challenge. But it doesn't take many such schools to belie the claim that it can't be done. As Kant wrote, the actual proves the possible.

High-performance schools have been studied for at least twenty years and we know, more or less, what sets them apart. They have a clear mission, team spirit, a coherent curriculum and pedagogy (though these take many forms), talented teachers, and values that embrace success. They also benefit from strong leaders,

Many of them also muster more time on task: longer school days, weeks, and years, and teachers who are accessible during off hours. Which isn't to say every "extended day" program yields higher achievement, only that high-achieving schools typically occupy a larger fraction of kids' lives.

That's both because these schools see the need for more teaching-and-learning time and because they want to keep their students off the streets. High-performance schools are reshaping poor kids' aspirations, priorities, and peer groups while imparting cognitive skills and knowledge. 

If we want more such schools, we'll likely have to spend more, or at least drive more of our dollars to our neediest schools. (Yes, "inputs" can make a difference--properly used.) But just as important is quelling the forces in American life that push against more time-on-task: summer employers who don't want the school year longer; union contracts that confer on teachers the right to go home at 2:43 p.m.; the entertainment industry with its appetite for kids' attention during evenings and weekends; and, of late, the dumbest idea of all: the push for less homework.

Homework is the most economical way for a school to stretch kids' learning time, albeit not necessarily the most effective (nor best at keeping them off the streets). But several wrong-headed recent books assert that there's too much of it, that it's mindless and formulaic and doesn't leave youngsters enough time "to be kids" (see here and here).

As one might suspect, the schools first influenced by this nonsense are elite institutions attended mainly by upper-middle class kids. Those are the kids fortunate enough to have salubrious places to go and things to do, and people to look after them once school lets out. They are the kids most apt to have shelves of books and parents who read--and who limit their TV access. These kids may do fine with limited school-prescribed homework. But the normal trajectory of education ideas is for them to trickle down from schools serving the prosperous middle class to schools that serve mostly the poor. (The most destructive example being classroom constructivism, which may work okay for kids with structure, discipline, and systems at home but, as E.D. Hirsch has shown, is a dire blunder for youngsters who depend on school for such things.)  

Most policymakers, employers, and college professors, and plenty of students themselves, understand that U.S. children can and should be learning more than they are today. Whatever their innate capacity--and it's far more elastic than the Murrays of the world realize--they have plenty of underutilized brain cells that could absorb lots more skills and knowledge. The key premise of NCLB is that this is possible and desirable and that the most direct path to it involves taking steps to change low-performing schools.

Even if 100 percent proficiency by 2014 is dreamy, what a different country this would be--how much better, stronger, and prosperous in so many ways--if we moved from today's 30 percent proficiency (using NAEP standards) to, say, 70 or 80 percent. And if poor and minority kids, in particular, were doing lots better and those vexing gaps were shrinking.

Combatting the determinists and fatalists is a multi-front war. But one well worth fighting--and winning.

Brian Gill, Ron Zimmer, Jolley Christman, Suzanne Blanc
RAND Corporation
February 2007

In 2001, Gadfly asked, "Will Edison be able to turn around Philadelphia's schools?" More than five years later, the RAND Corporation has an answer--sort of. Using longitudinal student-level testing data (district reading and math scores from 2002-2006) RAND sought to determine whether Philadelphia's so-called "diverse provider" plan--which enlisted Edison, among other private managers--has benefited that city's students. To recount: in 2001, Pennsylvania took over the School District of Philadelphia (SDP) and invited seven private and nonprofit organizations to manage 45 troubled schools. Another 21 schools were restructured but remained under district management, and a final 16 were only given increased funding. RAND crunched test data for all three groups and found that only the restructured, district-controlled schools showed a statistically significant differential effect on student achievement. RAND also found little evidence that competition from the privately managed schools boosted overall district gains. Accordingly, the authors conclude, "we find no evidence of differential academic benefits that would support the additional expenditures on private managers." But caveats abound. First, "the Philadelphia model was characterized by little competition among providers and by the absence of parental choice among the educational models offered," while "continued district involvement in provider schools... constrained provider autonomy" (see here). Second, "the private managers may be producing other benefits that are not measurable in terms of student achievement" (see here, for instance). Further, former Pennsylvania schools chief Charles Zogby argues in another forum that private managers were not, in fact, awash in extra cash as the RAND report suggests. It seems that more time and deeper changes will be required to reach a true verdict on the diverse provider model. Read the RAND report here. 

Jay P. Greene and Marcus A. Winters
Manhattan Institute's Center for Civic Innovation
January 2007

Jay Greene and Marcus Winters want to recast the way the nation looks at teacher pay: i.e., that it's more important to focus on how teachers are paid than how much they're paid. They do this by making two points. First, that teachers aren't paid too shabbily. The authors look at information from the U.S. Bureau of Labor Statistics regarding the hourly pay of public school teachers. Teachers make, on average, 36 percent more than the average non-sales white-collar worker--architects make 11 percent less than teachers, psychologists 9 percent less, and editors and reporters (gulp) 24 percent less. Their second point is that no correlation exists between higher teacher pay and higher student achievement (measured by high school graduation rates). Correlations between merit bonuses and higher student achievement have been observed, though. Thus, if raising student achievement is schools' top priority, teacher bonuses can and should be distributed with that priority in mind. The authors do not insinuate that good teachers shouldn't be well-compensated, or that the best teachers shouldn't make big bucks, but the emphasis needs to be on rewarding with bonuses those whose students are making classroom gains. Across-the-board pay raises are well and good, and we certainly have lots of them, but they do little to help students learn more. Find the report here.