Karl Alexander's decades-long study shows the long shadow of a poor start
Karl Alexander has spent more time in prisons than most professors. For 25 years, the Johns Hopkins University sociologist and his Johns Hopkins colleague Doris Entwisle followed the lives of 790 children growing up in a variety of Baltimore neighborhoods. The researchers interviewed their subjects almost every year while they were in school and every few years after they became adults. Early on, the participants could usually be found in school or at home. But as they aged, some of them began to land in prison. So Alexander, Entwisle, and their colleagues followed them there.
The researchers discovered that these young people, even if they were in jail for some "pretty nasty things" like attempted murder, usually made great interview subjects. They were polite and respectful and relaxed in conversation. They spoke freely about the most personal aspects of their lives: unfulfilled dreams, time spent using and dealing drugs, relationships and sex. "It's amazing what people will talk to you about, especially in the interviews we did in lockups," Alexander says. The stories the interviewees told could be heart-wrenching, even to a veteran sociologist like Alexander. One subject told the researcher that he had seen his brother hang himself outside his window. "That haunted me. That stayed with me," Alexander says. "You hear something like that, you say, 'Wow. That's really something to grow up with.'" The tales weren't all bleak. One subject had enrolled in his prison's master gardener program. Some hoped to earn their GEDs after they got out.
Alexander and his colleagues recorded every story. They collected a mountain of data: each subject's work history, how far he or she had advanced in school, their past drug use, number and ages of children and other family members, and relationship status. The sociologists combined this information with data from earlier interviews of both study subjects and their parents, along with profiles of the neighborhoods where their subjects grew up, school report cards, and family backgrounds. Pulling these strands together, the researchers wove a rich tapestry from the lives of children growing up in Baltimore from the early 1980s through the mid-2000s.
Many of the middle-class children in the study progressed through life's stages as expected: school, college, work, marriage, parenthood. But for poorer children, the picture was largely bleak. In their book The Long Shadow: Family Background, Disadvantaged Urban Youth, and the Transition to Adulthood (Russell Sage Foundation, 2014), co-authors Alexander, Entwisle, and Linda Olson, an instructor in the School of Education, combine an explication of 25 years of data with powerful anecdotes—stories of murdered friends and siblings, absent fathers, mothers too addicted to drugs or alcohol to provide basic care, dreams deferred. The researchers show how, at each step on the path to adulthood, neighborhood and family and school conspire to pass down advantage and disadvantage from generation to generation. Contrary to the popular American narrative that everyone has equal access to opportunity as long as he or she is willing to work hard, the reality revealed by the study is grim. Education and hard work lift people from the inner city out of poverty only in exceptional cases. The vast majority born poor are almost certain to stay that way.
The 68-year-old Alexander has the tall, lean build of a basketball player and sports a close-cropped white beard and wire-framed glasses. He smiles easily and loves to talk. He grew up in working-class neighborhoods in west and northeast Philadelphia in the years after the Second World War, and he fell in love with sociology in an undergraduate statistical methods course at Temple University. "Most everybody else hated the course, and I really liked it," he recalls. "What really appealed to me was to look at data and evidence and try to discern patterns in data in a way where you could make sense out of those patterns."
Inspired by that course, Alexander chose to major in sociology. "I was kind of floating. I didn't have a clear sense of purpose before I wandered into that class," he says. "It was a transformative moment for me." He went on to earn his PhD in 1972 from the University of North Carolina at Chapel Hill and took an assistant professorship at Johns Hopkins.
The university's Sociology Department was home at the time to two leading scholars with whom Alexander was excited to work: homelessness researcher Peter Rossi and education expert James Coleman. Within two years, however, both Rossi and Coleman left for other universities. But Alexander also met Doris Entwisle, a prominent researcher who studied education and early childhood. (Entwisle died of cancer in November 2013, after the manuscript for The Long Shadow was completed.) When Entwisle was tapped for a term as editor of the journal Sociology in Education, she invited Alexander to be her deputy. The two discovered they worked well together, complementing each other's strengths and avoiding ego battles. When Entwisle's editorship ended, they decided to develop a joint research project. Alexander had spent his early years at Johns Hopkins studying how students' high school experiences affected how they fared in the workforce. But he had come to feel that he was missing important pieces of the picture. By the time children got to high school, Alexander realized, much of their life path had already been set. "You don't start high school with a blank slate," he says. "There's a built-up history there that's relevant." He wanted to better understand that history. Given Entwisle's expertise in early education, a study of how family and neighborhood influence children's early success in school made for a natural collaboration.
The researchers developed a plan to interview children and their parents during first and second grades, and to survey teachers and gather school records. They would then spend a year analyzing data and writing up results. No one had ever tried to get data from interviews with first-graders. "There was skepticism that we really could get useful information from children that young," Alexander says. "We had to craft questions very carefully." For example, rather than ask what children aspired to be when they grew up, the researchers commissioned an artist to draw images of different occupations—say, a cop and a teacher—so they could ask the children to choose which they identified with.
For their study population, the researchers chose at random 790 students from 20 of Baltimore's public elementary schools. Four of the schools served middle-to-upper-class communities. Eight had a middle-income population, and eight served students who were among the poorest in Baltimore. Four of the low-income elementary schools had no white students; five of the middle-income schools had no black students. Alexander and Entwisle discovered hidden pockets of white poverty and realized they could study how both race and income affect school performance.
After three years, Alexander and Entwisle had enough data to show that by second grade, children from poorer families and neighborhoods had fallen behind their wealthier peers. The researchers published those findings but felt there was more to be learned from extending the study another year or two. After all, Alexander says, they had already done a lot of the heavy lifting—selecting the sample, establishing relationships with students, parents, and teachers, and securing cooperation from everyone up to principals and the superintendent's office. Why not follow the students into third and fourth grade, and see whether the income- and race-based educational stratification grew with time?
In the end, they kept the project going not just through elementary school but far beyond. They and their colleagues ended up interviewing study participants and parents nearly every year through the end of high school, occasionally taking a year off to develop a new survey. To compile and crunch their data, the researchers assembled a core staff who remained with them throughout the project, including researcher Linda Olson. The researchers supported their small army of data gatherers and analysts with funding from public agencies and private foundations. Support was wide but not always deep. "There were times when it came very close," Alexander says. "It was like the 11th hour, 59th minute—how are we going to make payroll?" The project's overall cost came out to somewhere between $12 million and $14 million, with no single sustaining grant. Alexander and Entwisle were always patching together relatively small sums of money.
As the project matured and the data piled up, the picture of how children's early circumstances played out in their later school success became clearer. But it took a decade for the project to make waves outside of academic journals and conferences. That opportunity came about because Baltimore City Public Schools, unlike most school systems, tested students in math and reading in both fall and spring. Alexander and Entwisle had collected students' test scores from the beginning, and around five years into their study they realized they could compare learning gains (or losses) during the school year with those from summer. The idea wasn't new, Alexander says; sociologist Barbara Heyns had published a book in 1978 documenting summer learning loss among poor schoolchildren in Atlanta. Heyns' finding had earned considerable attention within academia— both Alexander and Entwisle had read it—but few educators or policymakers were aware of it.
Drawing from multiple years of test scores, the Johns Hopkins researchers found that poor students in Baltimore generally kept up academically with wealthier students during the school year, which Heyns had found to be true among her Atlanta students. But in both cases, poorer students tended to fall behind during the summers. The main reason was that wealthier parents provided more opportunities for learning— trips to museums and libraries, for example. In some cases, these parents simply had access to more resources like books and money for vacations. But even free learning opportunities like a trip to the library, Alexander says, can be out of reach for children of single parents who must work multiple jobs just to get by.
Alexander and Entwisle published their summer learning results in 1992 in the American Sociological Review. Their team's finding got wide attention, sparking policy discussions and innovations in summer school programs around the country. One of the most important outcomes emerged close to home after a Johns Hopkins undergraduate, Matthew Boulay, read about Alexander's work in the magazine Education Week. Boulay had a year earlier launched a project called Teach Baltimore, through which he and fellow students taught summer classes to Baltimore youth in 14 locations around the city. Inspired by Alexander's work, Boulay decided to take his project national. "I was surprised and amazed," Boulay recalls. "[The research] opened up this whole new set of possibilities and need for what we were doing." He sought out Alexander, who became his mentor. Teach Baltimore eventually morphed into the Center for Summer Learning, and later into the National Summer Learning Association, an advocacy organization whose advisory board Boulay now chairs.
Had the Beginning School Study ended there, it still would have been significant. But Alexander and Entwisle's evolving vision gave the study a scope far beyond school. To fully understand how school and career trajectories emerged from early life experiences, the researchers decided to continue following their subjects as they graduated or dropped out of high school, got jobs (or didn't), and began to have children of their own. The team conducted extensive interviews when subjects were 21 or 22 and again at 28 or 29. The interview topics changed over time, reflecting what was going on in participants' lives as well as trends in the kind of data sociologists of the time were collecting from adolescent and young adult populations. The researchers' surveys eventually encompassed not just school and home life but also work and unemployment, drugs, crime, and personal hardship. As early as age 14, study subjects started having children; the researchers found that especially among low-income participants, these children often provided a sense of meaning that the rest of life lacked. Most of the low-income men had run-ins with the law at some point, and around half spent time in prison, which is where Alexander and Entwisle often found them.
Despite the considerable time demanded of them—up to two hours per interview—and the need to answer personal questions, only a handful of the study participants ever left voluntarily. "We think one of the reasons we've been so successful [at retaining participants] is that we've been a presence in their lives as far back as most of them can remember," Alexander says. "Beyond that, and I think this is the deeper reason, many of them really looked forward to talking to us because we were interested in them and wanted to hear their stories. And often they didn't have any adults in their lives who were interested in them and wanted to hear their stories."
Their stories could be both harrowing and uplifting, sometimes both from the same source. Danté Washington, a study participant who has allowed his real name to be used, recalls that he heard gunshots so often growing up that he didn't think much of them until one night he went downstairs and saw that a bullet had come through his home's living room window. "I looked at that and realized there has to be something different," he says. Washington went on to earn a degree in business from Strayer University and now owns a home in a different neighborhood and works at a publishing company.
Through it all, Alexander and his colleagues kept a professional distance, listening sympathetically but never offering material help. (The researchers did sometimes provide lists of resources to help study participants or their families with particular issues.) Striking this balance could be hard, Alexander admits, given their subjects' often intense experiences. "We worried about a lot of them," he says. "But we also rejoiced when we saw positive things happening."
The parents in the study had less personal connection to the researchers but were generally also willing to talk, Alexander says; the interviews he and his colleagues did with them tended to be shorter and less probing. But the subjects moved often and didn't necessarily keep Alexander and Entwisle in the loop. The researchers would seek participants by calling them, searching online, knocking on their doors, knocking on neighbors' doors, talking to people sitting on stoops on the blocks where subjects had lived. They sent birthday cards each year and noted which ones bounced back—a "goddamn genius" strategy, according to study participant John Houser. When a card came back undelivered, the researchers knew they needed to do extra reconnaissance work. "It required dogged perseverance," Alexander says. He and Entwisle learned the hard way that such perseverance can be hard to muster for people who are not personally invested in a project. For their final survey, at the behest of their funders, the study leaders for the first time hired a professional research firm, which got around a 30 percent response rate from past participants. Horrified, Alexander and his colleagues took their survey back and, reprising their time-tested boots-on-the-ground approach, rounded up 80 percent of the original subjects. Of those the researchers couldn't find, a sizable fraction were dead.
Following children for 23 years is unusual; there are only a very small number of examples of studies that do that, says Adam Gamoran, president of the William T. Grant Foundation, which helped fund the Beginning School Study. "Even the best federal datasets follow young people for eight, 10, 12 years." By staying with their population for so long, Alexander and Entwisle set themselves up to do what few, if any, researchers had done before: trace a direct line from children's earliest experiences to their eventual places in the workforce and in society.
Success in the workforce turned out, again, to be largely a tale of haves and have-nots. Nearly all the higher-income students graduated from high school on time or close to it; of those, almost half went on to earn a bachelor's degree by age 28, and around half were working at well-paying managerial, professional, or technical jobs by that age. Education is now almost universally recognized as the ticket to career success, and Alexander and Entwisle found that the ticket worked as promised—for the higher-income segment of their study population. In the book the authors refer to the school-to-work path as an "attainment regime," though Alexander now prefers the more accessible "success narrative." But he and Entwisle found that this narrative was mostly closed off to the low-income participants. A quarter of this population had not graduated from high school or earned a GED diploma by age 28. Only 12 percent had ever enrolled in a four-year college, and only 4 percent had graduated. Alexander mentions that 4 percent figure often. "To see just 4 percent of poor kids in our study complete college is just shocking. That's upsetting to me," Alexander says. "We should be able to do better by our poor children, especially given that we also see that so many want to be more successful at school."
Alexander is referring to the fact that even among those who had not succeeded at school, most seemed to believe the education success narrative could still work for them. In their final survey, 85 percent of study participants who had dropped out of high school said they still hoped to get more education. Study participants' attempts to follow through on those desires were often stymied, however. In one of the book's many illuminating vignettes, the authors tell the story of a young woman, "Tami," who grew up in a blue-collar neighborhood and had a baby during her junior year of high school. Tami managed to graduate on time and near the top of her class, and she tried to pursue a pre-law degree, but parenting and poverty kept getting in her way. At one point she left college to care for her daughter until she was in school full time; marriage and relocation to a different part of town later forced her to quit a second program. By 28, Tami had started and stopped community college three times without ever earning a degree. Speaking of her daughter, she said, "'I don't mean to say she stands in my way, but she has to come first, you know. I'm sure if I didn't have her I would be in law school right now.'"
Subjects who didn't succeed in school faced a tough road, Alexander and Entwisle found. Gone were the days when someone without a high school diploma could walk into a skilled, well-paying unionized job at one of the city's factories or shipyards. By the time the young adults in the Beginning School Study were entering the workforce, Baltimore, like many American cities, had been wracked by decades of deindustrialization. Left were lower-paying, mostly non-union jobs in a shell of the city's once-booming industrial economy.
Still, some study participants found ways to make a go of it. "Todd," for instance, dropped out of college after one semester and went to work for his father's trucking company. By 28, he was making $50,000 a year—a solidly middle-class income. But he was the exception; the average study participant without a college degree was earning well under $30,000 a year when last interviewed.
The most surprising findings showed up when Alexander and Entwisle broke down their data by race. From their interviews, the researchers had learned that many of the racial stereotypes attached to young people—men in particular—did not hold up. For instance, low-income white boys in their study were actually slightly less likely than black boys to graduate from high school. They were also as likely to have used drugs, and more likely to have used a drug other than marijuana. They were only slightly less likely to have an arrest record. That data suggested that, on paper, poor black men were at least as qualified as their white counterparts for whatever decent blue-collar jobs were available. Yet Alexander and Entwisle found that those jobs were going almost entirely to white men. The researchers' 2005–2006 survey revealed that low-income white men were employed full time, on average, 20 percent more during the previous 24 months and earned an average of $8,000 more per year than low-income black men. They also earned far more than both black and white women without high school educations, who got mostly low-wage clerical and service-sector jobs. But even though black and white women earned on average roughly the same (low) salaries, black women still fell victim to inequity. White women in the study did far better economically because they tended to partner with higher-earning white men. At age 28, 42 percent of low-income black women in the study were raising children by themselves on an average income of $20,000 a year. By comparison, only 9 percent of low-income white women with children were single; they earned an average salary of $30,000. Those with partners—64 percent—enjoyed an average annual family income of more than $50,000.
This stark racial disparity had its roots not in qualifications but in privilege. Alexander and Entwisle found that white men were finding their jobs through informal, word-on-the-street hiring networks of which black men were not part. The networks were relics of Baltimore's well-documented Jim Crow past, when blacks were systematically shut out of most skilled trades. Modern anti-discrimination laws ended such practices' official sanction, but white employers continued to hire mostly white workers through family and social connections rather than through formal job postings. Alexander demurs from blaming white men's comparative success in today's workplace on conscious racism. He acknowledges that racist attitudes among employers could influence hiring decisions, but he and Entwisle did not attempt to gauge such attitudes. Rather, he believes that in connecting their children and relatives with available jobs, white adults were simply giving their kids the best chance in life that they could. "You can hardly object to that," he says. But due to the legacy of discrimination and deindustrialization, parents of the black low-income study participants by and large had no such connections to offer. "At age 22, we asked everyone how they found their current or most recent job," Alexander says. "White men were much more likely to say through family or friends, while African-Americans most often said on their own. And 'on your own' is not a good place to be." This second success narrative, of relatively steady, living-wage work available almost exclusively to white men through informal hiring networks, "took us by surprise," Alexander says. "If there's an aha moment in the book, I think it revolves around that."
A belief in social mobility—that anyone can pull him- or herself up the socioeconomic ladder through education and hard work—has long been ingrained in American self-identity. In recent years, however, America's social mobility narrative has lost much of its credibility. Writing in 2013 in The New York Times, Nobel Prize–winning economist Joseph Stiglitz called equal opportunity "our national myth" and argued that children who grow up poor in the United States face more barriers to success than do those growing up poor in other industrialized countries. The Beginning School Study's fine-grained results reinforce this kind of argument. Among 314 children from low-income families, only 30 ended the study in the middle class. If every child in the study truly had an equal chance to reach that socioeconomic stratum, the researchers would have expected this number to be more than twice as high. A starker picture emerges from the outcomes of the 18 black male students who attended the survey's poorest elementary school and were still in the study at age 28. Seventeen of the 18 had been arrested and 16 had been convicted of a crime, seven were interviewed in prison, seven had dropped out of high school, only one had earned a four-year degree, and only five had been continuously employed for the previous two years.
The Long Shadow is not the first work to argue that poor children are getting a raw deal in today's America. But it documents with precision the mechanisms that can transfer advantage and disadvantage from one generation to the next, says Glen Elder, a University of North Carolina sociologist who advised Alexander's master's thesis. "One setback leads to another and they pile up over time," Elder says.
The authors did more than just compile and crunch numbers, says Lenora Fulani, a psychologist and political activist in New York. She applauds Alexander and his colleagues for also putting a human face on urban poverty, which often gets abstracted into statistics that distance both writer and reader from poor people's lived reality. "I was very thrilled with his take on understanding and his interest in exploring what was really happening in poor communities in ways that were much more personal, much more intense . . . than the usual things that are written in psychology and elsewhere," Fulani says. "He spent time there [and] got to know people." She commends Alexander et al. for discussing the white urban poor. By broadening the depiction of urban poverty beyond the usual stereotypes, the authors make it harder for readers to ignore it, she says. "One of the ways that I think this country stays away from poverty is that it's basically projected as black, with a lot of negativity connected to it," she says. "And that's unfortunate because it hurts poor people across the board."
Alexander retired from Johns Hopkins this summer. His study is over and the data now live in an archive at Harvard, where other researchers are free to mine them. For those seeking to help children escape urban poverty, Alexander's career capstone just lays the foundation. "The book sets us up for asking the next question, which is: What can we do about it?" Gamoran says. "We shouldn't think that the long shadow observed in this study is inevitable or is impossible to counteract." Boulay agrees; he thinks the education community is only now beginning to address in earnest the summer learning loss that Alexander and Entwisle revealed in the mid-1990s. "I think we're actually just seeing the very beginning of the implications of Karl's work," he says. "In 20 years, this is going to be even more important." Like the poverty he spent his career studying, it seems that Alexander's work will cast a long shadow.