My month as Slate 's resident Schoolhouse Rocker comes to an end today, and so with this post I hang up my blogger's keyboard. (Not entirely, actually; for the last few weeks I've been maintaining a bare-bones book blog for news about Whatever It Takes , and I'll keep that one going indefinitely.)

I'm grateful to my editors at Slate for giving me this opportunity, and I'm grateful to all the readers who wrote me with tips, suggestions, and stories about life in the classroom.

John McCain and Barack Obama aren't talking too much about education on the campaign trail these days—it probably has something to do with the whole two-wars-plus-collapse-of-capitalism thing. But nonetheless, the next president will take office at a moment of crisis and change in the public education system, and whichever candidate wins, he'll have a chance to push the conversation, and the system itself, in some new directions.

The local developments I've been writing about here—teacher-contract negotiations in Denver and D.C. ; experimental schools in Los Angeles and San Francisco and the Bronx and New Haven —are crucial laboratories and test cases for national policy, and the next president will have the opportunity to expand, refine, and build on those experiments. There are some other useful steps he could take right away. He could make school financing more equitable (a Berkeley law professor named Goodwin Liu has some good ideas on how to do that). He could fund more education R&D. (You could fill volumes with what we don't know about which educational practices actually work.) And he could construct accountability models that are more precise and humane than those found in No Child Left Behind.

Both candidates have some good ideas in their education platforms and some smart advisers. And though neither candidate has moved too far away from his party's traditional approach to education, Obama seems more willing to look for new and innovative solutions, ideas that might stir up the stagnant politics of education.

To me, one of the most significant planks in Obama's education platform isn't in his education platform at all-it's in his poverty platform : his pledge to replicate Geoffrey Canada's Harlem Children's Zone in 20 cities across the United States, as public/private partnerships, with the federal government's share of the bill coming to as much as a few billion dollars. Each city's version of the program would be unique, but presumably each would include a charter school as well as intensive preschool programs, after-school tutoring, and family counseling. And I would hope that each replication would include a program, like the Harlem Children's Zone's Baby College , that would encourage parents to adopt better child-rearing strategies. (I had a piece on This American Life this weekend about Baby College and the complicated process of trying to sway Harlem's parents toward different methods.)

There are a lot of potential obstacles to Obama's pledge becoming a reality (including the fact that he will have to be elected president before he can do much about it). As I wrote earlier this month in the New York Times Magazine ,

A lot of conservatives would oppose a new multibillion-dollar federal program as a Great Society-style giveaway to the poor. And many liberals are wary of any program that tries to change the behavior of inner-city parents; to them, teaching poor parents to behave more like middle-class parents can feel paternalistic. Union leaders will find it hard to support an effort that has nonunion charter schools at its heart. Education reformers often support Canada's work, but his premise-that schools alone are not enough to make a difference in poor children's lives-makes many of them anxious.

There aren't yet airtight data to prove that Canada's model works—though as I wrote earlier this month , the numbers coming out of his elementary schools are very promising. And so rather than simply cloning the Zone and airdropping it into communities around the country, Obama's replication project will work best if each city is encouraged to adapt and innovate, to compete with every other city for the best results. (As Obama said in his speech announcing the plan, "every step these cities take will be evaluated, and if certain plans or programs aren't working, we will stop them and try something else.")

Obama's "Promise Neighborhoods" could challenge the traditional division between education policy and poverty policy—between improving schools and improving the lives of poor families. Geoffrey Canada's argument is that it no longer makes sense to think of each one separately. If we try to fix the schools in a low-income neighborhood without addressing the other needs of students there, it's not a real solution to the neighborhood's problems. And it isn't enough to provide social services to poor children if their neighborhood schools are still giving them a lousy education. A true solution to the problem of underachievement in inner-city public schools is going to require more nurturing families and safer neighborhoods as well as better teachers and more accountable schools. That's the real point of the Harlem Children's Zone, and, I think, it's going to be the next chapter in the debate over schools.

After my post on cheating went up on Monday, a few readers wrote to remind me of a classic work of cheating economics, " Rotten Apples: An Investigation of the Prevalence and Predictors of Teacher Cheating ," by Harvard economist Brian A. Jacob and Chicago economist Steven D. Levitt, which was prominently featured in Freakonomics , the best-selling book by Levitt and Stephen J. Dubner. According to the original paper's abstract , the authors:

estimate that serious cases of teacher or administrator cheating on standardized tests occur in a minimum of 4-5 percent of elementary school classrooms annually. The observed frequency of cheating appears to respond strongly to relatively minor changes in incentives. 

As Levitt and Dubner reported in Freakonomics, Arne Duncan, the CEO of the Chicago public schools, contacted Levitt and Jacob after their paper was published and asked for their help in catching the cheaters. Duncan and the economists chose 120 classrooms—some classrooms with suspicious results, some regular classrooms as a control—and retested them, but with officials from Duncan's office overseeing the test, rather than the teachers themselves. When the test results came back, the scores from the classrooms with suspicious results had dropped sharply. A dozen teachers were fired, and as Dubner and Levitt write in their book, "The final outcome of the Chicago study is further testament to the power of incentives: the following year, cheating by teachers fell more than 30 percent."

Another reader sent a link to this story , by the New York Sun 's education reporter, Elizabeth Green, about an alleged rash of cheating at P.S. 48 in the Bronx. Green uncovered convincing evidence of a principal who pressured teachers to cheat:

Meanwhile, 11 of 12 P.S. 48 graduates interviewed last week said they were coached during the state tests. They said that teachers would look over their shoulders and instruct them to try again and again until they got answers right. ...

"When I was at 48, I never went to class, and I still passed the test," a seventh-grader said. "If you go to graduation, you pass."

If your paycheck depends on your classroom's test scores, there's always going to be a temptation to cheat. But stopping cheaters doesn't seem like rocket science. As another reader wrote : "One obvious solution is to stop allowing teachers to proctor their own student's tests. You don't let the students grade their own tests because of the temptation to cheat; you shouldn't allow the teachers that temptation either."

I've long been impressed by the accomplishments of Achievement First , the charter-school network that grew out of Amistad Academy , a middle school founded in 1999 in New Haven, Conn., by Doug McCurry, an educator, and Dacia Toll, then a recent graduate of Yale Law School.

Amistad has always served a low-income, high-minority population—a demographic that tends to do poorly in school and on standardized tests. And yet Amistad's results are consistently excellent. In 2007, the eighth-grade students at Amistad were neck and neck with the students at Worthington Hooker School for the highest state-test results in New Haven. Hooker is a well-respected public school in the city's East Rock neighborhood; some of its students have parents who are Yale professors or who are otherwise connected to Yale. Very few of Hooker's students—just 16 percent—are from low-income families, while at Amistad, that figure hovers around 80 percent. And yet Amistad's eighth-grade students beat out Hooker's eighth grade for the highest math score in New Haven, and they earned the second-highest reading and writing score in the city, right behind Hooker—a fairly stunning result. ^*

What impresses me just as much as Amistad's sterling test scores is the commitment of the school's founders to the broader issue of equity in education. Achievement First's leaders see their mission as not just to serve the 3,700 students enrolled in their schools in Connecticut and Brooklyn, but to create a national model for school success in low-income communities.

Two years ago, Amistad middle school expanded upward to include a high school as well. And so I was interested, and a little concerned, to read in the New Haven Register last month that students at the high school have been struggling.

Last year, 19 of the 56 students in Amistad's ninth grade—the same cohort that scored so well on the eighth-grade tests in 2007—failed at least one course. And 11 students, 20 percent of the grade, failed two courses, meaning that they did not move on to the tenth grade.

It's not quite clear just what this means. Last year, the 10th-grade students at Amistad once again did very well on the state achievement tests. So by outside measures, achievement at the school remains high. And yet many students are struggling to meet Amistad High's own standards, even after four or five years of Amistad education.

Amistad, to its credit, has launched an internal review of the middle school's methods to see what steps it can take to better prepare its students for high school. As the Register reported :

Amistad is now examining both the academic and non-academic structure of its middle school, adding more non-fiction reading, for example, and evaluating whether the institutional attitude of "we will not let kids fail" has in turn failed to teach students how to learn independently. "Independence is something you teach and develop and cultivate. We need to be more thoughtful and more intentional about doing that," said [Dacia] Toll.

My guess is that whatever difficulties Amistad's high-school students are now facing can be traced back to their experiences before reaching Amistad. As at most high-performing low-income middle schools, students generally have been arriving in fifth grade at Amistad performing well below grade level. Amistad's intensive methods are usually able to propel those students to grade level and above in just a few years. But it may be that the early academic deficits those students experienced have some lingering aftereffects that will make high levels of achievement a continuing challenge for them at every stage.

Given Amistad's track record, and Toll's willingness to face her school's problems directly and publicly, I imagine that she and her staff will figure out how to get most of those students to graduate from high school well-educated and prepared for college, even if it takes an extra year or two.

I would also hazard a guess that in a few years, things will start to get a lot easier at Amistad High. Achievement First has been assembling its own version of Geoffrey Canada's "conveyor belt" ; this year, for the first time, one of the group's charter elementary schools began feeding students into one of their middle schools with four years of Achievement First education under their belts—arriving ahead of grade level instead of well behind it. It seems a safe bet that the experience those students have in ninth grade, a few years from now, will be a lot smoother than the experience of the kids now enrolled at Amistad High.

To me, the struggles at Amistad High point to the same two conclusions as do the attrition issues at the Bay Area KIPP schools: First, that the KIPP/Achievement First middle-school model can achieve truly remarkable levels of success among the kind of students—low-income, behind grade-level, a less-than-ideal home life—who just a few years ago were routinely written off altogether. And second, that if and when these middle schools are linked together with equally well-run elementary schools and high schools  (and then, I hope, with high-quality prekindergartens, parenting programs, and community and family supports), they will be able to achieve much more: better and more consistent results from a broader selection of students. If that happens—and at Amistad, it may begin as soon as this year—what now seems miraculous will, I hope, come to seem downright routine.

^* I'd like to provide a link to the New Haven Register story reporting those figures, but it's not easily available on line. The reference, should anyone want to Nexis it, is: Maria Garriga, "Charter school group ups the ante on scores," New Haven Register , Aug. 7, 2007.

The new union-friendly charter school in the Bronx

is not the only big project that

has taken on this fall. The other is the attempted transformation of Locke High School in the Watts neighborhood of Los Angeles. The school, which currently has about 2,500 students, has long been notorious as one of the worst in the city, with what the

as a "reputation for student fisticuffs and an appallingly high dropout rate."

Green Dot was founded by Steve Barr, a garrulous, outspoken Irish American in his late 40s who helped start Rock the Vote in 1990 and nine years later decided his role in life was to run high schools. His organization now manages 10 of them, mostly in L.A., and his new mission is to transform the way public education works in the city (and then in the rest of the country).

Barr is a bit of a bomb-thrower, and Locke is his most recent incendiary device. Two years ago, with the support of a few hundred dissatisfied parents of Locke students, he persuaded more than half of the Locke faculty to sign a petition calling for the L.A. Unified School District to turn the school over to Green Dot's control, and in September 2007, the school district reluctantly handed over the keys . Barr hopes to show the district that he can run a big urban high school better than it can. And he hopes to show the L.A. teachers union that the bare-bones 33-page contract that Green Dot's unionized teachers sign is a lot better than the 300-page contract that the regular union has with the L.A. school district. In the process, Barr hopes to change the way both institutions do business.

Last year was a transitional one for Locke, and it didn't go very well. Green Dot hadn't yet taken over, but the district had already more or less moved out. The result was a disastrous year for the students, culminating in a schoolwide brawl in May that involved as many as 600 students and brought dozens of police officers to the campus, some in riot gear.

Barr showed me around Locke yesterday morning. I never saw the pre-Green Dot Locke, so I didn't have anything to compare it with, but things looked calm and orderly, students wearing their new polo-shirt uniform tucked into their new regulation khakis. There were no fisticuffs.

The Locke project is in many ways a risky undertaking. It's hard to turn around miseducated ninth graders. And Locke is unlike other charters, in that families don't have to fill out a special application to attend. If they live in the neighborhood, Locke is their high school. As the L.A. Times pointed out recently,

it's one thing to make progress with students who voluntarily sign up for a rigorous academic environment and whose parents actively support the endeavor. Green Dot's experience with Locke's many doubt-filled teens will provide a more realistic measure of what charter schools can do for poor and minority students who typically have lower test scores and higher dropout rates. And if it succeeds, Green Dot will have created a blueprint for public schools.

Of course, what makes the project somewhat less risky is that after last year, the school has nowhere to go but up.

To all San Franciscans: I'll be reading and discussing Whatever It Takes tonight (Tuesday) at 7 p.m. at the Books Inc. in Opera Plaza , an event co-sponsored by 826 Valencia , the drop-in tutoring center for kids who want help with their writing. And tomorrow (Wednesday), I'll be giving a talk at lunch at the University of California-Berkeley journalism school, an event hosted by Berkeley professors Michael Pollan and David Kirp . If you're in the neighborhood, please drop by.

One potential problem with basing teacher compensation in part on test scores is that it gives teachers an incentive not just to "teach to the test," but to game the test completely. Because of strict accountability measures imposed in Texas in the 1990s by then-Gov. George W. Bush and Rod Paige, the Houston school superintendent, test scores rose sharply. A few years later, after the governor had become the president and the superintendent had become the federal secretary of education, the Dallas Morning News , in a series of investigative articles , revealed that at least part of those test-score gains were due to widespread cheating by teachers and administrators.

I don't want to overstate the prevalence or the impact of cheating on standardized tests—I think accountability measures are crucial, and I believe a well-run school system can find ways to all but eliminate cheating. But still, I thought I should share this story, which a reader of this blog, a young teacher in the New York City public schools, e-mailed to me recently:

I'm writing in response to your column on paying teachers more money for raising their students' test scores from the year before. When I first heard this proposal, a couple of years ago, I was excited. That should be a decent measure of a teacher's efficacy, I thought.

Until I got my class last year. As sixth graders, they were all new to my school and came from different elementary schools. And when it came to the test, every single one of them, without fail, had the same story.

During our first standardized test of the year, many hands went in the air. I shook my head at them, because according to the rules, I'm not allowed to talk to them during a test. They were outraged. "Why can't you help us?" they asked.

We had a class discussion after the exam was over. Every single one of them had received help on state standardized tests in their old school. For some of them, the teacher would explain the questions when they didn't understand something. Several students had teachers give them the correct answer. One of my students saw a teacher sit down with another student's test, erase all of his answers, and write all new ones in.

If there are bonuses tied to test scores, even more cheating will take place. What can school districts do to ensure this doesn't happen?

It's a good question, without a simple answer. I do think school systems have the tools to stop cheating. But it's hard to do without first acknowledging that it's a problem. And superintendents (and mayors) have the same accountability pressures that principals and teachers and students do. Whatever level of the bureaucracy you inhabit, when your success depends on rising scores, it's hard to take steps that will serve only to lower those scores—whether that means blowing the whistle on a fellow teacher or launching an investigation of the whole system.

Speaking  of KIPP:

In Whatever It Takes , in one of the chapters on the Promise Academy  middle school, I describe the impact of the KIPP  schools in the Bronx and Harlem on the Promise Academy’s leaders and staff. This was during the first few years of the Harlem Children Zone’s middle school, which were a struggle, and those KIPP schools, which had very good test results, were for the Promise Academy administrators both a standard to be aspired to and a frustrating reminder that their own students weren’t performing at the same high level as KIPP’s students.

Terri Grey, the Promise Academy principal at the time, believed the attrition issue was part of what was holding her school back. As she put it to me in one conversation, "At most charter schools, if the school is not a good fit for their child, the school finds a way to counsel parents out"—to firmly suggest, in other words, that their child might be happier elsewhere. "Whereas Promise Academy is taking the most disengaged families and students and saying, 'No, we want you, and we’re trying to keep you here, and we don’t want to counsel you out." That policy made it impossible, she believed, for the Promise Academy to achieve KIPP-like results.

I’m not entirely convinced that that was the real problem at Promise Academy—or that the KIPP schools in New York were actually "counseling out" a significant number of students. But I do think it’s true that Geoffrey Canada’s guiding ethic has always been to go out of his way to attract and retain the most troubled parents and students. And that makes running a school, or any program, more difficult, even if it makes the mission purer and, in the end, more important.

To me, the solution to the attrition issue, whether it’s at a KIPP middle school or the Promise Academy middle school, is the Harlem Children’s Zone’s "conveyor belt" model , which provides continuous, high-quality early-childhood and elementary education to precisely those "disengaged families and students," so that when those children arrive in middle school, they won’t have the kind of difficulty doing demanding work as did the kids who left the Bay Area KIPP schools or who underperformed at the Promise Academy middle school in its first few years.

As Geoffrey Canada put it in one conversation I quote in my book , "The question is, can you build a system where kids in middle school won’t need these kinds of interventions in order to be successful? And my bet—I could be wrong, but this is my bet—is if we start with kids very early, and we provide them with the kind of intense and continuous academic rigor and support that they need, then when they get to the middle school and high school level, we’re not going to need those superhuman strategies at all."

The good news, from my point of view, is that a few KIPP schools are now beginning to follow a similar model. KIPP Houston , the flagship KIPP school, is creating its own version of a conveyor-belt system, one that starts with prekindergarten for 3-year-olds and goes right through high school. To me, that’s a very promising development. And if the model spreads to other KIPP schools, I think this whole attrition debate could before long be a thing of the past.

In Wednesday's post , I mentioned a new study of the KIPP middle schools in the San Francisco Bay Area. I still haven't read the whole study (it's 140 pages), but I've read a good chunk, along with a report from the San Francisco Chronicle and one from the San Francisco Examiner . The study shows mixed results, and they're mixed in an intriguing way.

First, the five KIPP schools studied do, on the whole, a very good job of raising the achievement level of the students enrolled in them. Eighty percent of the students progressed faster than normal, and at the end of fifth grade, KIPP students outperformed their "matched counterparts"—demographically similar students at other schools who began fifth grade at the same academic level—by anywhere from five to 33 points on a percentile ranking.

Second, the KIPP schools didn't attract higher-scoring students than other schools in their districts. This is important, and it was somewhat surprising to me. In my 2006 article in the Times Magazine about the achievement gap , I wrote about the fact that at the KIPP Academy in the Bronx, one of KIPP's two flagship schools, incoming students were above average for the neighborhood:

Even though almost every student at the KIPP Academy ... is from a low-income family, and all but a few are either black or Hispanic, and most enter below grade level, they are still a step above other kids in the neighborhood; on their math tests in the fourth grade (the year before they arrived at KIPP), KIPP students in the Bronx scored well above the average for the district, and on their fourth-grade reading tests they often scored above the average for the entire city.

The new study shows that even though that phenomenon may be a reality in the Bronx, it's not happening in the Bay Area. The report says that "students with lower prior achievement in [English] or mathematics were more likely to attend each of three KIPP schools, compared with other same-grade students in the same neighborhood." That makes KIPP's accomplishment in raising the scores of their students even more impressive.

But the third big piece of news in the report, on attrition, is less impressive. Sixty percent of the students who enrolled in fifth grade at the three KIPP schools studied had left their school by the end of eighth grade. Some of those students left simply because their family moved out of the district. (KIPP school leaders estimated that family moves accounted for 42 percent of the leavers.) But many of the rest of the students left, it seems, because KIPP was too demanding or difficult for them. One school leader is quoted as saying, "I think for a cohort of students and families, it was harder than they thought it was going to be. Our expectations were more than they had anticipated."

If that's true of KIPP schools across the country, it means that they're accomplishing an important but somewhat limited mission: providing an excellent education to that group of low-performing, low-income students who are able to keep up with the schools' intense demands.

In my post last Friday on the Harlem Children's Zone (the subject of my new book ), I referred briefly to the difficulties that the organization's charter middle school, the Promise Academy, faced while I was reporting in Harlem. On Sunday, I received an e-mail from a reader, a teacher named Derek, who had just finished the book and asked if he could "see more reporting if possible on the middle school at the HCZ; the book seemed to end with it hitting a real rocky road."

Then Diane Ravitch, a respected education historian who is a part of the Broader, Bolder Approach group I mentioned last week, wrote about Whatever It Takes on her blog for Education Week. She wrote that she found the book as a whole "very hopeful"—except for the middle-school chapters, which she found "depressing."

All of which made me think that there might be some gaps that I should fill in here regarding the Promise Academy middle school. To summarize those chapters briefly: The Promise Academy middle school opened in Harlem in the fall of 2004 with a class of 100 sixth-grade students. That first year was difficult, and the school's results on the citywide tests in the spring of 2005 were quite poor despite a lot of arduous preparation. Over the next two years, things improved gradually, but not quickly enough to prevent Geoffrey Canada from making some major changes to the school.

Each one of those first three years in the life of the school gets a chapter in my book. What I wasn't able to report there, because of my book deadline, was that in the fourth year (2007-2008) things improved significantly at the school. Test scores were up, and, from what I'm told, there were fewer behavioral problems and less antagonism between students and teachers. And after a year's hiatus in admissions—Canada decided that the middle school wouldn't bring in a new sixth-grade class in the fall of 2007—a new grade was admitted last month.

This week, the New York City Department of Education released their annual "report cards" for every elementary and middle school in the city, and the Promise Academy received an A , the highest grade, for the second year in a row. (Once you click on that link , you have to choose "Manhattan" and "District 84" to find the grades for charter schools.)

To me, though, the real point of those chapters wasn't to judge the success of Geoffrey Canada's middle school. It was to give readers a chance to ponder some of the most difficult questions in urban education today.

The first has to do with age. One of the main conclusions I reached in my book was that educational and social supports that start early in a child's life are much more effective than those that start in middle school. (Those are the "very hopeful" chapters Ravitch referred to.) Early interventions work better and faster, they involve less stress for both students and teachers, and they have the potential, I believe, to propel students to a higher level of success.

So, if we want to improve outcomes for poor children and eliminate the achievement gap, should we devote more of our resources toward preschool and elementary school, where interventions are arguably more efficient? Or should we devote more of our resources toward middle schools, even if it takes more work and more money, because that might be our last chance to rescue failing students? (My preference is to spend more money earlier, though, of course, I think you need a balance.)

The second question is one that Ravitch raised in her blog post yesterday:

Do poor black and Hispanic kids really need to be in "no excuses" schools that insist on rote learning and rote behavior? That take control of their lives and change their culture? Should this be the model for education for children of color in big cities?

Ravitch is referring to middle schools run by charter organizations like KIPP and Achievement First , which emphasize not only an intensive academic curriculum, but also "character" education, often establishing an elaborate system of rules, rewards, slogans, and punishments intended to better prepare middle-school kids to learn. She's not a big fan of those schools, and I wouldn't necessarily agree with some of her language above—I wouldn't say those schools take "control" of students' lives, and I don't think they insist on "rote behavior." But I know what she means.

And I think it's a big question. I wrote about those schools in the Times Magazine back in 2006, and since that article came out, I've continued to visit KIPP schools and schools modeled after KIPP in cities across the country. Though I try to be skeptical, I'm always impressed by the atmosphere of the schools, by the engagement of their students, and by their results. (Coincidentally, a giant report just came out evaluating the KIPP schools in the Bay Area. I've only read a bit of it, but Eduwonk says it's "overall good news.")

In fact, I think one of the big reasons for the early problems at the Promise Academy middle school is that Geoff Canada and the school's administrators and teachers weren't able to do what every new KIPP school tries hard to do, which is to settle very early on a coherent school culture and then stick to it and reinforce it at every turn. That school culture doesn't need to include KIPP-type chants and slogans, and it definitely doesn't need to involve "rote behavior." But it does need to go beyond the classroom.

I think students from low-income families in blighted neighborhoods who enter middle school way behind grade level need something more than just extended hours and expert teaching (though they need that, too). They also need adults around them who believe in them and care about them and who can guide them toward the behaviors and the mental habits that will help them succeed in school and in life. I'm not sure if I'd call that "changing their culture." But I'd certainly call it changing their minds.

The divide between the teachers unions and the charter-school crowd often seems unbridgeable. At an education-reform forum in Denver last month, on the eve of the Democratic Convention, Cory Booker , the mayor of Newark and a member of the pro-charter Education Equality Project , spoke out against the unions, noting how "vicious" they can be:

Ten years ago, when I started talking about school choice, I was tarred and feathered. ... I literally was brought into a room by a [teachers] union [representative] ... and threatened that I would never win in office if I kept talking about school choice, if I kept talking about charter schools."

When collaborations—or even kind words—between the two sides do come along these days, they can seem all the more startling. Take the new Green Dot school in the Bronx.

Green Dot Public Schools is a charter-management organization that runs 12 charter high schools in Los Angeles. It's different from other successful charter groups, like KIPP , Achievement First , and Uncommon Schools , in three important (and challenging) ways: Green Dot runs high schools instead of middle schools. The group specializes in "transformations" of existing failing public schools instead of starting their own from scratch. And their schools work with unionized teachers, unlike most charter schools.

Two weeks ago, Green Dot opened its first school outside of L.A., a charter in the South Bronx. What's unusual about the school is that it is run in a partnership with the United Federation of Teachers, the New York City teachers union. The ceremony marking the school's debut was a lion-lying-down-with-the-lamb moment: Joel Klein, New York's school superintendent, stood next to his frequent adversary, Randi Weingarten, the president of the UFT, as they cut the ribbon.

It's hard to know if the new school will be the beginning of a trend toward unionized charters—for now, there doesn't seem to be much enthusiasm for this kind of arrangement on the part of any charter provider but Green Dot. But at the very least it's a model for a new type of relationship between old enemies. And if the reform folks really want to scale up their no-excuses model to a point where it's big enough to change things for a significant number of kids across the country, it couldn't hurt to have the unions on board.

The reason I think the Harlem Children's Zone is so important—the reason I wrote a whole book about the program—is that I think it's the closest thing we have to a model for the kind of collaboration I was referring to yesterday .

What Geoffrey Canada has constructed in Harlem is a comprehensive set of integrated programs that currently serve 8,000 kids in a 97-block neighborhood, starting at birth and going all the way through college. It is based on two innovative ideas. The first is what Canada calls the Conveyor Belt—a system that reaches kids early and then moves them through a seamless series of programs that try to re-create the invisible cocoon of support that surrounds middle-class and upper-middle-class kids throughout their childhoods. The Conveyor Belt starts with Baby College , a nine-week program that provides expecting parents and parents of young children with new information about effective parenting strategies. The next stop is an all-day language-focused pre-kindergarten for 200 4-year-olds, who then graduate into a K-12 charter school that has an extended day and an extended year and employs some of the intensive academic practices developed in the KIPP schools. Throughout their academic careers, students at the school have access to social supports: after-school tutoring, a teen arts center, family counseling, and a health clinic.

The second idea is a tipping-point notion—what Canada refers to as contamination. His theory is that in a low-income, high-crime neighborhood, if you offer social and educational supports to just a few of the kids who live there, their participation will always seem a bit oddball, and they won't have much of an effect on their peers. But if you get participation rates up to 40 percent or 50 percent or 60 percent, then taking part will come to seem normal, and some of the behaviors that used to seem commonplace in Harlem—teenage pregnancy, drug use, dropping out of school—will start to seem like the oddball path. The engaged kids will "contaminate" their friends with their behaviors and attitudes.

Canada's system isn't easy. It requires a lot of hard work just to keep it in motion. And in the years that I spent reporting in Harlem, Canada and his staff made lots of wrong turns and hit plenty of dead ends. In the book, I followed one class of parents through Baby College, and some of them, it seemed, faced such big deficits and such huge obstacles in life—they couldn't read, they had had other children taken away by Child Services, they had spent a couple of years in jail—that it seemed hard to believe they would ever be truly effective parents. In the middle school, the first couple of years were quite rocky, as Canada struggled to combine the ethos of a community organization with the accountability of a no-excuses charter school.

By the time I finished my reporting, though, the middle school was starting to find its footing, and the elementary schools, where some of the students had been with the Harlem Children's Zone since Baby College, were truly thriving. The third-grade test scores last spring were good—at one charter school, 97 percent of the third-grade class was on grade level in math, and in the other, 100 percent were. (The English scores were lower, but they were still quite good. At one school, they were 10 points above the state average, and at the other they were just a point or two below the state average.)

And perhaps more importantly, the elementary schools and the kids in them felt somehow ... normal. When I spent time in the classrooms, I got the strong feeling that when these kids got to middle school, they weren't going to need the kind of heroic interventions that Promise Academy and most charter middle schools need to employ today. They wouldn't need remediation and advanced character-building and constant test prep—they would just be competent, engaged students for the rest of their school careers. And these are kids who, for the most part, came from low-income, often difficult backgrounds, with a fair number of teenage parents and parents who didn't complete high school.

They were exactly the same kind of kids, in other words, who arrived in the sixth grade in the first year of Promise Academy middle school, the ones who showed up reading three and four years behind grade level , and whose subsequent middle school careers were a constant struggle. This new generation of kids had the good fortune to find a place on the Conveyor Belt, and that meant they faced a very different kind of future than most kids growing up in Harlem.