The Harlem Project

By PAUL TOUGH

Published: June 20, 2004

 

ack in 1990, Geoffrey Canada was just your average do-gooder. That year, he became the president of a nonprofit charitable organization based in Harlem called the Rheedlen Centers for Children and Families, and he set out trying to improve the world, one poor child at a time. It was a bad moment to be poor in New York City. Harlem, especially, was suffering under the simultaneous plagues of crack cocaine, cheap guns and rampant homelessness, and Canada's main goal at Rheedlen, in those years, was to keep the children in his programs alive. The organization had an annual budget of $3 million, which it spent on a predictable array of services in Upper Manhattan: after-school programs, truancy prevention, anti-violence training for teenagers. The programs seemed to do a lot of good for the children who were enrolled in them, at least in part because of Canada's own level of devotion. He was obsessed with his job, personally invested in the lives of the children he was helping and devastated when they ended up in prison or on drugs or shot dead on the street.

But after he ran these programs for a few years, day in and day out, his ideas about poverty started to change. The catalyst was surprisingly simple: a waiting list. One Rheedlen after-school program had more children who wanted to enroll than it was able to admit. So Canada chose the obvious remedy: he drew up a waiting list, and it quickly filled with the names of children who needed his help and couldn't get it. That bothered him, and it kept bothering him, and before long it had him thinking differently about his entire organization. Sure, the 500 children who were lucky enough to be participating in one of his programs were getting help, but why those 500 and not the 500 on the waiting list? Or why not another 500 altogether? For that matter, why 500 and not 5,000? If all he was doing was picking some kids to save and letting the rest fail, what was the point?

At around the same time, he was invited by Marian Wright Edelman, the president of the Children's Defense Fund, to join a group she had recently founded called the Black Community Crusade for Children. Once a year, she brought together two dozen leaders from across the country who were trying to solve the problems of poor black children. They met down at a farm in Tennessee that had once been owned by Alex Haley, the author of ''Roots,'' and they spent a few days comparing notes on the crisis in America's poor neighborhoods. For Canada, the good news at these discussions was that he wasn't alone -- but that was the bad news, too. All across the country, in big cities and in small towns, well-meaning nonprofits were finding the same thing: they were helping a few kids, getting them out of the ghettos and off the streets and sometimes even into college, but for the masses of poor children, and especially those who were black, nothing was changing; those children were still falling behind in school, scoring below average on reading tests and staying poor.

Most of the men and women who were meeting in Tennessee were from Canada's generation -- he is 52 -- and they had come of age in the hopeful period following the passage of the Civil Rights Act, when things seemed as if they were about to improve for poor black Americans. But as Edelman convened her group, life in the ghettos was getting worse, not better. To Canada, it felt as if he and his peers were losing more children more quickly than they had ever lost before. For the first time, he felt a sense of hopelessness, and he found himself thinking that the kids he was seeing in kindergarten in Harlem were already doomed, destined to spend the rest of their lives stuck at the bottom.

Canada knew there were success stories out there. There were always reports in the newspapers about ''special'' kids who ''overcame the odds.'' Some brilliant teacher or charity or millionaire went into the ghetto and found 100 kids and educated them and turned their lives around. But those stories seemed counterproductive to Canada. Instead of helping some kids beat the odds, he thought, why don't we just change the odds? When he looked around, though, he couldn't find anyone who knew how to do that. Experts in his field had figured out how to educate one disadvantaged child, or one classroom full of kids, but the benefits were localized, and usually temporary. And no one had any idea how to change a whole school system or a whole housing project, or for that matter a whole neighborhood. So, in the middle of the 1990's, that's what Geoffrey Canada decided he would do. And now, 10 years later, he has become a very different kind of do-gooder, one with a mission both radically ambitious and startlingly simple. He wants to prove that poor children, and especially poor black children, can succeed -- that is, achieve good reading scores, good grades and good graduation rates -- and not just the smartest or the most motivated or the ones with the most attentive parents, but all of them, in big numbers. Three years ago, he chose as his laboratory a 24-block zone of central Harlem, now expanded to 60 blocks -- an area with about 6,500 children, more than 60 percent of whom live below the poverty line and three-quarters of whom score below grade level on statewide reading and math tests -- and he named it the Harlem Children's Zone.

After welfare reform passed in 1996, the national debate on poverty seemed simply to shut down. Most conservatives explain poverty by looking to culture and behavior: bad parenting, high out-of-wedlock birth rates, teenagers who don't know the value of an honest day's work. To most liberals, the real problems are economic: underfinanced public schools and a dearth of well-paying semiskilled jobs, which make it nearly impossible for families to pull themselves out of poverty. Canada says he believes that both assumptions are true. He agrees that the economy is stacked against poor people no matter how hard they work, but he also thinks that poor parents aren't doing a good enough job of rearing their children. What makes Canada's project unique is that it addresses both problems at once. He keeps the liberals happy by pouring money into schools and day-care centers and after-school programs, and he satisfies the conservatives by directly taking on the problems of inadequate parenting and the cultural disadvantages of a ghetto home life. It's not just that he's trying to work both sides of the ideological street. It's that Canada has concluded that neither approach has a chance of working alone. Fix the schools without fixing the families and the community, and children will fail; but they will also fail if you improve the surrounding community without fixing the schools.

Canada's new program combines educational, social and medical services. It starts at birth and follows children to college. It meshes those services into an interlocking web, and then it drops that web over an entire neighborhood. It operates on the principle that each child will do better if all the children around him are doing better. So instead of waiting for residents to find out about the services on their own, the organization's recruiters go door-to-door to find participants, sometimes offering prizes and raffles and free groceries to parents who enroll their children in the group's programs. What results is a remarkable level of ''market penetration,'' as the organization describes it. Eighty-eight percent of the roughly 3,400 children under 18 in the 24-block core neighborhood are already served by at least one program, and this year Canada began to extend his programs to the larger 60-block zone. The objective is to create a safety net woven so tightly that children in the neighborhood just can't slip through.

At a moment when each new attempt to solve the problem of poverty seems to fall apart, one after the next, what is going on in central Harlem is one of the biggest social experiments of our time. Social scientists and poverty advocates are watching carefully to see if Canada can pull it off. Many are skeptical; they have seen too many ambitious anti-poverty programs collapse because of budget overruns or administrative hubris, and Canada acknowledges that his work has just begun. But the sheer scale of Canada's project has created a palpable excitement among foundation officials, poverty scholars and business leaders. Marian Wright Edelman said that though there are a few other good neighborhood-based programs around the country, ''none are as comprehensive as the Harlem Children's Zone, and none of them hold as much promise.''

David Saltzman, executive director of the Robin Hood Foundation, concurred: ''If it works, it'll be the best thing that's happened in a long time. Man, if Geoff can make this thing work, it's huge.''

The programs that the Harlem Children's Zone offers all seem carefully planned and well run, but none of them, on their own, are particularly revolutionary. It is only when they are considered together, as a network, that they seem so new. The organization employs more than 650 people in more than 20 programs; on a recent afternoon, I spent some time walking around Harlem, dropping in on one program after another. At Harlem Gems, a program for 40 prekindergarten students at a public school on 118th Street, Keith, who had just turned 5 and was missing a front tooth, sat at a computer working away at ''Hooked on Phonics,'' while Luis, a 19-year-old tutor, gave him one-on-one instruction. A few blocks up Lenox Avenue, at the Employment and Technology Center, 30 teenagers in T-shirts and basketball jerseys, all part of the organization's new investment club, were gathered around a conference table, listening to an executive from Lehman Brothers explain the difference between the Dow Jones and the Nasdaq. At P.S. 76 on West 121st Street, fifth-grade students in an after-school program were standing in front of their peers, reading aloud the autobiographies they had written that afternoon. And over at Truce, the after-school center for teenagers, a tutor named Carl was helping Trevis, a student in the eighth grade, with a research project for his social studies class, an eight-page paper on the life of Frederick Douglass. In a nearby housing project, a counselor from the Family Support Center was paying a home visit to a woman who had just been granted legal custody of her two grandchildren; in other apartments in the neighborhood, outreach workers from Baby College, a class for new parents, were making home visits of their own, helping teach better parenting techniques. A few blocks away, at the corner of Madison Avenue and 125th Street, construction was under way on the organization's new headquarters, a six-story, $44 million building that will also house the Promise Academy, a new charter school that Canada is opening in the fall.

While the new building is going up, Canada works on Park Avenue between 130th Street and 131st Street, in a small office in a six-story building that always seems to be under renovation. When I visited him on an icy afternoon in February, the radiator in his office was hissing constantly; when the room got too hot, Canada propped his window open with a book about community revitalization. That cooled things off, but it also created a new distraction. Directly outside Canada's second-floor window, no more than 20 feet away, are the elevated tracks that carry every Metro-North train heading out of Grand Central Terminal toward Connecticut and Westchester County. Each time a train passed, full of commuters on their way back to the suburbs, a rumbling filled the room, and Canada leaned a little closer so that I could make out what he was saying.

He is a tall, lean, athletic man with rounded shoulders and long limbs, and on this afternoon he was wearing a dark suit and a light blue shirt with his monogram sewn over the breast pocket. His graying hair was cropped close to his scalp. His office is spare -- a desk with a phone and a computer and a few piles of mail. There's a coat rack for his suit jacket, a bookshelf and a small round table with four chairs where he holds meetings. On his desk is a big picture of his 5-year-old son, Geoffrey Jr., from his second marriage. On the wall behind his desk are photographs -- Canada with President Clinton, Canada with Mayor Bloomberg -- as well as a portrait of a dozen or so of the young people he has trained in tae kwon do, which he has been teaching two nights a week for 21 years. A framed citation on the opposite wall certifies him as a third-degree black belt.

Although Canada likes to say that he is sick of against-the-odds success stories, he is one himself. He grew up on Union Avenue in the South Bronx. His father left when he was 4, and his mother reared him and his three brothers herself, sometimes supporting them with wages from menial jobs and sometimes relying on welfare and food from local charities. In his memoir ''Fist Stick Knife Gun,'' Canada describes the rituals and codes of violence that governed life for children like him, growing up in the inner city in the 50's and 60's. As a teenager, he drank and fought and smoked pot and carried a knife, but he also stayed in school, worked in a factory in the evenings and won a scholarship to Bowdoin College in Maine, and from there went on to earn a degree from the Harvard Graduate School of Education.

Bowdoin doesn't have many black alumni, and so when a promising black high-school senior is applying to the college from New York, the admissions office often sends the student to Canada for an alumni interview. When I arrived at his office, Canada was firing questions at Julian, a 17-year-old from Brooklyn who attends a magnet school. Canada made notes on a clipboard as Julian talked, answered a few of his questions about Bowdoin and Maine and then shook his hand and showed him out.

''He's a good kid, and I think he'll be terrific at Bowdoin,'' Canada said as he sat back down. ''But he's not my kid.'' Not a Harlem Children's Zone kid, he meant. ''He comes from a family, they've got it together, his parents are both educators. That's not us. We want those other kids, the ones who don't have two parents, whose parents haven't gone to college, who haven't got a chance statistically of making it.''

This distinction -- between children who are ''his kids'' and children who aren't -- is one that Canada draws all the time, and it goes back a long way in his own personal history. In ''Fist Stick Knife Gun,'' published in 1995, he describes a summer night when he was 14. He and half a dozen friends were sitting on someone's stoop drinking Rheingold when a car came screeching up and a man they didn't know got out and challenged one of them to a fight. The fight started, and at first it went according to the rules of Union Avenue, meaning no weapons and no one else gets involved. But then the stranger pulled out a gun, which was almost unheard of in 1966. Canada was ready to run, but instead he and his friends slowly converged on the older, bigger, better-armed man and scared him off the block through the strength of their numbers. What he learned that day, he wrote, was that he and his friends ''loved one another enough to be willing to die.''

When Canada talks about ''my kids'' now, he means the 6,500 children who live in the Harlem Children's Zone. But even more than that, he means the least successful children in the neighborhood. When he talks about them, it feels personal, as if his real objective is to work with children just like the cold, tough, frustrated boys he grew up with on Union Avenue, most of whom are now dead or in prison -- to make amends for the past by saving Harlem's children today.

Still, Canada's approach to poverty is not sentimental. When he considers any given poor family living in the Harlem Children's Zone, he divides their problems up in his mind into the ones he needs to solve and the ones he doesn't need to solve. The ones he needs to solve are the ones that are keeping the child from succeeding in school. Everything else, he has decided, he can leave alone. ''Do we think that it would be better for our parents to be married?'' he said, tilting back in his chair. ''Absolutely. Why? Because two-parent families have more income. Children tend to do better when they have two parents in the house.'' In fact, he knows that the great majority of births in Harlem are to single mothers and that most of the parents whose children he serves are unmarried. ''But what ability do we have to make an impact on that?'' he asked. ''None. Right? If we tried to do that, we'd spend all our time just doing that.''

Canada admitted that he is engaged in a kind of triage. He described for me an imaginary Harlem parent. ''You can be 20 years old, with a job that doesn't pay you enough money to survive on,'' he said. ''You're underemployed. You've got a kid. The kid's not doing well in school. You've got no place for the kid to be after school. Well, we'll provide services for that child. But we're not going to solve the problem of you being underemployed. That's not going to go away.''

He is, in other words, sidestepping the macroeconomic solutions that some advocates insist are the only way to solve the problem of poverty -- better wages, a national jobs program, a bigger earned-income tax credit -- in favor of programs that in one way or another directly affect the performance of the neighborhood's poor children.

Canada's educational philosophy emphasizes accountability and testing, and in that way it is similar to the dominant idea in public education today. The doctrine of accountability -- the idea that if students do poorly on standardized tests, schools should lose their financing and teachers should lose their jobs -- first emerged in the late 80's and early 90's in the Houston public schools. It then moved to the White House as the basis of the No Child Left Behind law when Rod Paige, the superintendent of the Houston schools, became the education secretary under George W. Bush. In the past year, though, news reports and lawsuits have revealed that when schools are compelled to meet certain numbers -- graduation rates, standardized-test scores -- their administrators often succumb to the urge to cheat. In Houston and New York, principals have shoved troubled students out of school, often under an administrative sleight of hand, in order to keep their schools' numbers artificially high. Canada has set the same rigorous goals for his own organization, but for him, the urge is the opposite: not to push the worst kids aside, but to recruit them even harder.

On this afternoon, Canada was worried about a set of internal statistics he had just uncovered: some of his students seemed to be doing too well. Last fall, his organization started a new program in four Harlem public schools called the Fifth Grade Institute, an after-school program for 160 fifth-grade students designed to begin catching them up to grade level before the charter school opens in September. Canada wanted to calculate how much the program was improving the reading ability of these students, so he asked to see their scores on the previous year's citywide fourth-grade reading test for comparison. Reading scores in New York City public schools are delivered in four categories, the higher the better. A 4 means the child is reading above grade level; a 3 means the child is reading at grade level; a 2 means below grade level; and a 1 means significantly below grade level. In most of the city, 2's and 3's predominate, with some 4's thrown in. In schools in Harlem, though, about three-quarters of the students score either a 1 or a 2; there are a few 3's, and 4's are rare.

But when Canada looked at the scores for the children in the Fifth Grade Institute, he found a lot of 3's -- more than a random sampling of Harlem students would have drawn. And on the day I visited, he was worried that the process of recruiting students for the Fifth Grade Institute had somehow been selective. He was sure it wasn't conscious on the part of his administrators, and, in fact, when he later received more detailed scores, they seemed more in line with the neighborhood patterns. Still, it was only natural, he knew, that parents who would bother to sign their children up for an ambitious after-school program would tend to be the better-organized, better-educated ones, and so it wouldn't be surprising if their children were better readers. Maybe his kids, the 1's and 2's, hadn't heard about the program, or maybe their parents hadn't managed to get it together in time to sign them up.

o what do you do? If you offer a new program, the best students will naturally enroll first, but you want the worst students. How do you get those parents to apply? Sometimes, in Canada's experience, it happens by accident. In 2001, the first year the Harlem Children's Zone offered Harlem Gems, its Head Start-like program for 4-year-olds, the organizers were behind schedule and didn't manage to send out fliers and start recruiting until August, just a few weeks before the program began. All the well-organized parents had already made their child-care plans for the year, and the last-minute, overburdened parents were the ones who signed up. That gave Canada the demographic he wanted, and he was able to get concrete results. When the 4-year-olds started Harlem Gems in 2001, 53 percent were scoring ''delayed'' or ''very delayed'' on the Bracken Basic Concept Scale for school readiness. At the end of the yearlong program, 26 percent were delayed.

In that case, the search for the most delayed kids worked because the organizers got lucky. Usually, though, it involves a lot of knocking on doors. I went out one morning with two outreach workers, Francesca Silfa and Mark Frazier, as they searched through central Harlem for new parents to enroll in Baby College. They knocked on the door of every apartment in a 21-story building on 118th Street, looking for parents with children under 4, leaving fliers under the doors if no one answered. They didn't meet with any resistance or hostility, although they did get a few skeptical looks. Mostly the process seemed slow and painstaking. Whenever they ran into Baby College graduates in the halls, they stopped to chat, prodding them for suggestions of good candidates in the building or elsewhere on the block. In addition to the door-to-door approach, recruiters visit laundromats, supermarkets and check-cashing outlets to look for new mothers. On the day I was with them, Silfa approached one woman pushing a stroller down 118th Street and managed to sign her and her daughter up on the spot.

For most parents, the attractions of a program like Baby College are obvious: useful information, the relief of spending Saturday morning getting to know other mothers instead of being stuck in a cramped apartment and free child care during class time. But if those aren't enough of a pull, there are raffles at the end of each class for $50 Old Navy gift certificates and a big drawing at the end of the course in which one parent wins a month's rent.

This combination of door-knocking, cajoling and offering incentives seems to work, and each nine-week session of Baby College draws a class of more than 100 parents, all of whom receive weekly follow-up visits at home. More than 95 percent of the parents -- 677 so far -- have successfully completed the course. The real test of this approach, Canada said when we spoke in February, will come when the Promise Academy opens its doors. By law, the charter school will be open to students from anywhere in the city's five boroughs, which means a lottery to choose the incoming classes from a broader pool of applicants. But the only students Canada really wants in his school are the ones from central Harlem, and especially the lowest-performing ones, exactly those whose parents, he said, are least inclined to apply to send their children to a special school.

''I will make sure that every single poor-performing school and parent in Harlem knows about this program,'' Canada said back in February. ''And if we don't get more 1's and 2's than 3's, then we haven't done our job properly.'' What that will require, Canada said, is persistent recruiting. ''We're going to have to go and force some of these parents to come and fill out the application,'' he said. Even once he finds the 1's and 2's, he added, ''their parents may say no, which means we have to go back and figure out a way to bribe them and get them to say yes. And I hate to put it like that, but that's what we'll end up doing. We'll end up saying, 'If you get your kid to apply, you'll get free movie passes.'''

The point of all of this intensive recruiting is to amass evidence -- indisputable data that show exactly what it will take to level the playing field and get poor children performing on the same level as the city's middle-class children. ''If we just end up saving a bunch of kids in Harlem, that will be good for them, but it won't mean an awful lot to me in the long run,'' Canada said. ''We want to be able to say that thousands of poor children can learn at high levels and perform at rates that are the same as middle-class children if they are given the opportunity to do so. But I want to be clear when I say we've got an answer that we really have an answer.''

The answer that Canada wants to provide is in fact very much in demand. There are plenty of examples of programs that didn't work: in 2000, for example, the Heinz Endowments abandoned as a failure a $59 million, five-year program to provide early-childhood care and education for 7,600 low-income children in and around Pittsburgh. The program's costs soared, and four years in, when Heinz pulled the plug, only 680 children were being served. What is most startling about the current study of poverty is how little conclusive evidence there is about which cures do work. There are no more than a dozen studies in the field that track how successful various interventions are over the long term, and the evidence from those studies tends to be spotty and subject to debate.

To William Julius Wilson, the Harvard sociologist who is one of the country's leading thinkers on urban poverty, this is precisely the significance of the Harlem Children's Zone. ''It is very, very important for policy makers to be able to cite examples of how you can improve the life chances of disadvantaged kids,'' he said. ''There are so many people who feel that whatever you do, it's not going to work. They want to say, 'Well, there's just a culture of poverty out there, and you can't really change it.''' If Canada's project is successful, though, Wilson said, it will ''provide the ammunition to policy makers who want to do something to address the problems of poverty.'' It will allow them to say: ''Here are kids who would ordinarily end up as permanent economic proletarians, and here is a program that has been able to overcome the cumulative disadvantages of chronic subordination. So why not commit ourselves across the nation to try to duplicate what he's done?''

Canada has a vivid picture in his mind of a judgment day to come -- 8, 10, 12 years down the road. Children who are now entering the system as infants will be taking their third-grade citywide math and reading tests, and they'll be scoring at or above grade level. Children who entered the system as Harlem Gems will be graduating from high school -- he expects that 90 percent of his students will graduate on time. It is only at that point, he explained, that he will be able to say to the rest of the country: ''This isn't an abstract conversation anymore. If you want poor children to do as well as middle-class children'' -- not necessarily to be superachievers but to become what he calls ''typical Americans,'' able to compete for jobs -- ''we now know how to do it.'' If he's right, the services he will provide will cost about $1,400 a year per student, on top of existing public-school funds. The country will finally know, he said, what the real price tag is for poor children to succeed.

Canada first came up with the idea for the Harlem Children's Zone in the mid-90's, but it wasn't until a few years later that he was ready to propose it officially to the board of the Rheedlen Centers (as the organization was then known). Crime had dropped sharply in Harlem, as it had everywhere in New York City, and Canada was no longer overwhelmed by the daily anxiety of trying to prevent the children in his programs from being killed. The national economy was booming, housing prices in the neighborhood were climbing and the first signs of gentrification were appearing, but for children in Harlem, the situation hadn't improved much. The unemployment rate there was still very high. (Even now, after the arrival of new middle-class residents, it stands at 18.5 percent.) It was in this context that Canada went before the Rheedlen board in 1998 and said that he wanted to remake the organization completely, to set up a kind of project that had never been tried before.

Canada had recently brought on a new board member, a fellow Bowdoin alumnus named Stan Druckenmiller, who, while running George Soros's Quantum Fund, became one of the most successful hedge-fund managers in the history of the stock market, amassing a personal fortune estimated at more than $1 billion. After Canada laid out his proposal to the board, Druckenmiller took him aside and told him that in his opinion he had the right plan but the wrong board. Canada agreed, and the two men politely deposed the chairman and replaced him with Druckenmiller, who set about raising money and recruiting new board members from the higher echelons of Wall Street.

The organization now has a lot more money than it did a few years ago. Druckenmiller paid for about a third of the cost of the new headquarters himself, and board members contribute about a third of the annual operating budget. (The rest comes from foundations, the government and private donors.) In April, the organization held a glittering fund-raising dinner at Cipriani 42nd Street, a cavernous former bank building converted into a restaurant, and raised $2.8 million in a single night, mostly from bankers and stockbrokers. (The 2003 fund-raiser, by comparison, pulled in $1.5 million.)

Financiers and C.E.O.'s are drawn to the Harlem Children's Zone not just because of its mission but also because of the way it is run. In fact, the relationship the organization has with its donors is as unusual, arguably, as the program itself. In the late 90's, Allen Grossman, then the president of Outward Bound U.S.A. and now a management professor at the Harvard Business School, began studying the nonprofit sector from a business perspective. It was a mess, he concluded: in an article in The Harvard Business Review in 1997, he described a hopelessly dysfunctional relationship between foundations and nonprofit organizations, in which foundations made short-term grants to pet projects, nonprofits spent all their time chasing money and each side had a vested interest in maintaining the reassuring fiction that failing programs were actually succeeding. Grossman proposed that philanthropists start thinking more like venture capitalists: searching out nonprofits with innovative long-term ideas, financing them early, insisting on transparency and frequent evaluation and nurturing them along the way with expert advice and continuing infusions of capital.

One of the first foundations to take Grossman's ideas seriously was the Edna McConnell Clark Foundation, which had been financing Rheedlen for years. In 1999, Nancy Roob, the foundation's grant manager, approached Canada and offered him $250,000, plus the dedicated services of a new management consultancy for nonprofits called the Bridgespan Group, to write a new business plan. This was unheard of: when philanthropists give away a quarter of a million dollars, they generally want the money to go directly to poor people, with as little overhead as possible. Clark, by contrast, was inviting Canada to spend the foundation's money -- first the $250,000, and eventually another $1 million -- entirely on overhead: on developing the business plan; creating a new, more hierarchical management structure; buying new technology for an internal communication system; and constructing a rigorous and continuing process of self-evaluation.

The business plan that Canada's team came up with proposed a steady increase in the annual budget over nine years, from $6 million to $10 million to $46 million. (This year, four years in, it is $24 million.) The plan reads more like a corporate strategy document than a charity prospectus. It refers to ''market-penetration targets'' and ''new information technology applications,'' including a ''performance-tracking system.'' In practice, too, the organization feels more like a business than a nonprofit, which offers comforting visuals to donors: everyone at the headquarters wears a suit, every meeting starts on time and there is a constant flow of evaluations, reports and budgets. ''Geoff could be a C.E.O. at any S.&P. 500 company,'' Druckenmiller said, and he meant it as a compliment.

uring the months I spent visiting Canada, the issue of education took up more and more of his time and attention. When we first spoke, almost a year ago, he described all of the organization's programs -- an initiative to combat asthma, an organizing campaign for tenants -- as equally important. But as the months wore on, it seemed, all he could think about were the problems in the schools. He was pouring more and more resources into Harlem's public schools, paying for in-class tutors and after-school reading programs, and scores had barely budged. He was beginning to see it as a systemic problem, he said, something that couldn't be solved by the kind of supplementary services he was offering. ''We've got to really do something radically different if we're going to save these kids,'' he told me in the fall. ''If we keep fooling around on the fringes, I know 10 years will go by, and instead of 75 percent of the kids in Harlem scoring below grade level on their reading scores, maybe it will be 70 percent, or maybe it will be 65 percent. People will say, 'Oh, we're making progress.' But that to me is not progress. This is much more urgent than that.''

So at the same time that he has been working inside the school system with a greater intensity than before, he has also begun to try to opt out of it, by establishing charter schools. Beginning in September, the Harlem Children's Zone plans to start educating its own students at two locations. The Promise Academy, which will eventually be a kindergarten-through-12th-grade charter school, will start with a kindergarten and sixth-grade class in September. The sixth grade will be housed in the organization's new headquarters on 125th Street; the kindergarten is expected to open inside an existing public school. The academy will expand over the course of seven years into a school of 1,300 students. The curriculum will be intense: classes will run from 8 a.m. to 4 p.m., five days a week -- an hour and a half longer than regular city schools. After-school programs will run until 6 p.m., and the school year will continue well into July.

Although charter schools are regulated by the state, and are outside the city's control, Joel I. Klein, the city's schools chancellor, has supported Canada's charter-school efforts. Canada said he sees Bloomberg and Klein as his allies, which is a strange feeling, he said, because for so long he felt as if he were at war with the school system. Soon after Klein was appointed, in the fall of 2002, he called Canada and set up a meeting. Canada said that the proposals Klein laid out for the entire school system -- greater accountability, more charter schools, more involvement for outside groups -- were exactly what Canada had been waiting to hear. In the summer of 2003, Klein appointed Lucille Swarns to be the regional superintendent for Upper Manhattan, and when Canada met her, he was once again pleasantly surprised.

''When I first came here in the early 1980's,'' Canada said, ''we felt that District 3'' -- which stretches from the Upper West Side into central Harlem -- ''ran a system almost of apartheid, where below 96th Street, the schools were doing great, and the schools we cared about were doing lousy.'' Even after his organization began spending a million dollars a year in the district's schools, he said, he never once got a call from the district superintendent. With Swarns things were different right away.

The most surprising aspect of the new collaboration between Klein and Canada is that the chancellor is encouraging the Harlem Children's Zone's plans to convert existing public schools in Harlem into charter schools. Canada has received conditional approval from Klein to start a slowly expanding charter school, beginning this fall with the Promise Academy's kindergarten class. As those children move through the school, the organization will take over control of an additional grade each year, sharing the building with the public school while gradually supplanting it. In six years, Canada explained, ''we're going to put the existing school out of business'' -- and he said he hopes that that school is only the beginning.

All of which makes skeptics, especially those in the teachers' union, wonder about Canada's motivation. Randi Weingarten, the president of the United Federation of Teachers, said that she used to be friendly with Canada, even attending his fund-raisers, until Bloomberg and Klein came into office and started making threatening noises toward her union. ''Since the mayor became mayor, Geoff Canada has stopped having any relationship with us,'' she said.

In her opinion, Canada's new attitude comes down to politics: ''I think what's happened is that they've decided that they'll work with the mayor, and that they won't work with the public-school system except through the mayor and the chancellor'' -- meaning that they won't work with the teachers' union. Canada often speaks of the opposition that his charter-school plan is going to face from the teachers' union and what he calls the educational establishment, but to Weingarten, it's the other way around: it's Canada who is picking this fight, demonizing the teachers' union in order to score political points with the mayor. ''They are working very secretly with the Board of Education and the Bloomberg administration'' on the charter-school plan, she said, a strategy that she said was shortsighted, not least because it is far from certain that Bloomberg will be re-elected. ''If you truly want schools to succeed,'' she said, ''you work with the people who represent the teachers.''

The battle between the teachers' union and City Hall has been going on for decades, but it has reached an unusually high pitch under Bloomberg. Canada clearly feels a genuine ideological kinship with the mayor, and with Klein, but there's also an immediate advantage to his alliance with them: his charter schools will use nonunion teachers; they will be paid more than public-school teachers, he said, but they will also work longer days, and for 12 months a year. Canada also wanted a free hand to fire teachers who weren't performing up to his expectations -- authority Canada said he felt sure the union would not give him. At his new school, he will have it.

n the days and weeks leading up to the charter-school lottery, the organization's outreach workers did exactly what Geoffrey Canada said they would do: they went door-to-door in housing projects, they tracked down recalcitrant mothers and fathers, they solicited applications from the parents of children in every one of their programs. And by the evening of the lottery, a rainy Tuesday in April, they had received 359 applications for just 180 slots -- 90 in the kindergarten class and 90 in the sixth grade. Legally, they could have held the lottery behind closed doors and simply mailed out acceptance letters, but Canada wanted it to be a real event, so he arranged to hold the lottery in public, in the auditorium of P.S. 242 on West 122nd Street.

By the time the lottery began, the place was packed. Every seat was filled with a hopeful parent or a prospective student or a patient sibling, and late arrivals were standing in the aisles or cramming themselves into the back of the room. Multicolored helium balloons were tied to the end of each row of seats, which gave the room a festive air, but more than anything the place felt nervous.

Canada took the stage. On a long table next to him, a gold drum held a jumble of index cards, each one printed with the name of a child. ''We are calling our school Promise Academy because we are making a promise to all of our parents,'' Canada said from behind a lectern. ''If your child is in our school, we will guarantee that child succeeds. There will be no excuses. We're not going to say, 'The child failed because they came from a home with only one parent.' We're not going to say, 'The child failed because they're new immigrants into the country.' If your child gets into our school, that child is going to succeed. If you work with us as parents, we are going to do everything -- and I mean everything -- to see that your child gets a good education.''

And then the drawing began, starting with the kindergarten class. Doreen Land, the charter schools' newly hired superintendent, read the first name into a microphone: ''Dijon Brinnard.'' A whoop went up from the back of the auditorium, and a jubilant mother started edging her way out of her row, proudly clutching the hand of her 4-year-old son. Land smiled and took the next card: ''Kasim-Seann Cisse.'' Another whoop, some applause and then another name.

At the front of the auditorium, Canada greeted each mother (or, occasionally, father) and child. Proud parents shook his hand and introduced their children, beaming on their way back to their seats. In the second row, a woman in her 40's, wearing an ''I Love New York'' T-shirt and a red nylon jacket, sat with her head bowed and her eyes shut tight, her lips moving in an anxious prayer. And then Land called the name ''Jaylene Fonseca,'' and the woman's eyes flew open. She made her way to the front, shook Canada's hand and then told me that her name was Wilma Jure and that Jaylene was her niece, a 4-year-old already enrolled in Harlem Gems. As her eyes filled with tears, she explained that Jaylene's mother, Jure's sister, was living in the city's shelter system for homeless families. Most nights, Jaylene slept with her mother in a shelter on 41st Street, then spent the day in Harlem Gems. ''It's an amazing program for people like us who really don't have the means to send our kids to get such a good education,'' Jure said. ''I mean, she's learning French,'' she added, a little incredulous. Now Jaylene was guaranteed a spot in the Promise Academy from kindergarten through the end of high school. It was hard not to feel that her life had just changed for the better, with the draw of a card.

As the evening wore on, though, the mood in the auditorium started to shift. The kindergarten lottery ended, the chosen students trooped out to the cafeteria for a group photo and the sixth-grade lottery began. In the front row, Virainia Utley sat with her daughter Janiqua, listening to the names being called. Utley is something of a model parent in the Harlem Children's Zone. Janiqua is in the Fifth Grade Institute at P.S. 242, and her three younger siblings -- Jaquan, Janisha and John -- are all enrolled in the computer-assisted after-school reading program. Utley is the vice president of her tenants' association, which was organized by Community Pride, the community-organizing division, and she is a regular presence at Zone events. When I first met her, months earlier, she was already talking about Janiqua going to the Promise Academy. But now the lottery numbers were rising -- 54, 55, 56 -- and her name hadn't yet been called.

After Land read out the 90th name, Canada took the stage again and explained to the remaining parents that it wasn't likely that there would be room for their children in the sixth grade. Land would read out the rest of the names and put them on a waiting list, he said, but this part wouldn't be much fun. He encouraged everyone to go home. Land went back to reading names, and Utley and Janiqua sat and listened, still in their seats, as the waiting list grew and the number of cards in the drum dwindled. By the time Land got to the 80th place on the waiting list, Utley told me, she and Janiqua were just waiting to make sure her name was called. Maybe her card got lost, she said, or stuck to another card. The room was thinning out, and the only remaining parents were angry ones. One by one they were letting Canada know how they felt. A line of parents came up to him to find out what could be done to get their children into the school, and he had to tell each one the same thing: nothing could be done. One woman, who was too angry to give me her name, spat out her complaint. ''It's not fair,'' she shouted. ''And I don't like it.'' Why drag everyone out on a rainy night, she wanted to know, just to sit in a public-school auditorium and feel like losers?

Finally, at No. 111 on the waiting list, Janiqua Utley's name was called, and her mother rose, took her by the hand and started up the aisle to the backdoor. As workers began sweeping up coffee cups and popped balloons, I sat down next to Canada. He looked exhausted, overwhelmed not only by the evening but also by the enormity of the task ahead of him. His eyes were watery, and as we talked, he dabbed periodically at his nose with his folded-up handkerchief. ''I was trying to get folks to leave and not to hang around to be the last kid called,'' he said. ''This is very hard for me to see. It's very, very sad. These parents feel, Well, there go my child's chances.''

It was a waiting list, I reminded him, that had started him on the path toward the Harlem Children's Zone more than a decade ago -- and now, despite all the millions of dollars, and the staff of 650 and the backing of the mayor, he is still setting up waiting lists. He nodded. ''We've got to do more,'' he said. ''We've got to do better.'' He sighed and looked up at the stage, where Land had just reached No. 150.