Alyssa mayer was four months pregnant the day a police officer showed up at her motel room in Kingston, New York. It was late afternoon in August 2013, the sun dragging toward the Catskills on the west side of town. Earlier that week, her boyfriend, who’d been sleeping at her place since he found out about the baby, had missed a curfew check. Both of them had recently gotten out of prison on parole, and weren’t supposed to be around anyone else with a criminal record. With the authorities looking for him, they could both get in trouble. So they’d packed some clothes and driven to a Super 8 and hoped for some idea of what to do next. Mayer was going out to pick up a pizza when she ran into the officer in the hallway.

She and her boyfriend had grown up together around Kingston. The area had been a manufacturing center for IBM until the company started laying off workers in the early 1990s, around the time Mayer was born, leaving not much more than strip malls and fast-food joints, along with rising crime rates, in stretches of the Hudson Valley. After Mayer’s parents split up, when she was a toddler, her mother worked two jobs and would return home seeming distant. Mayer spent a lot of time at her grandmother’s house and, later, on the streets in the rough part of town. In high school, she moved in with a cocaine dealer she met one day at a gas station. He bought her new clothes, manicures, anything she wanted. By the time the relationship ended, she was making sales of her own.

In 2009, when Mayer was 18, she fronted six grams to a friend who had just gotten out of prison. He told her he was broke and needed to make a quick deal. As it turned out, he had already made one with the local narcotics team. Some time later, the cops kicked in the front door of her apartment, and she ended up with a three-year felony sentence.

When Mayer learned she was pregnant, in the summer of 2013, she had already returned to prison twice for parole violations. She called a clinic to make an appointment for an abortion. She knew she wasn’t in the best position to be a parent—she had started a new job and believed she could turn her life around, but she wasn’t sure that her boyfriend wanted to do the same. She didn’t want her child to be raised without a father, like she had been. Once her boyfriend found out, though, he swore to her that they would work things out. So she didn’t show up for the appointment, and instead got a tattoo across her collarbone that read Blessed. Not long after that, they went on the run.

The officer who handcuffed Mayer in the motel didn’t seem to care when she told him she was pregnant. Neither did the parole judge, who charged her with fraternizing with another parolee and skipping curfew and ordered her back to prison. As she stripped down at the intake facility and stepped forward to be searched, she faced the question that thousands of American women do each year: What happens to a baby born in detention?

Over the past four decades, as the inmate population in the United States has grown into the largest in the world, the number of children with a parent in custody has risen to nearly 3 million. For corrections officials and policy makers, those relationships can fade into the background. But not when a child is born on the inside.

For as long as women have been doing time, prisons have had to contend with the children they carry. In 1825, a pregnant inmate named Rachel Welch received a whipping so severe that it was suspected of causing her death not long after she gave birth. Nearly 200 years later, the clashes are less violent but perhaps no less consequential: the vast majority of women who give birth while incarcerated in the United States must hand over their baby within a few hours of delivery, to family, friends, or the foster-care system. For some mothers—even those with short sentences—these separations turn out to be permanent. And with a nearly 800 percent increase in the number of women in custody since the late 1970s, the births are happening on a scale that is hard to ignore. An estimated one in 25 female inmates is pregnant when the prison doors lock behind her.

In recent years, the flood of women into the correctional system has prompted a growing number of states to create programs known as prison nurseries, which allow women to keep their newborn children with them behind bars. Inmates who qualify can raise their babies for a limited time—ranging from one month to three years, but in most states 18 months—in separate housing units on prison grounds. Eight states now offer prison nurseries, all but one of which have opened in the past two decades; Wyoming recently finished constructing a facility that will bring the total to nine.

Research associating participation in the programs with lower recidivism rates among mothers has helped make nurseries a rare shared cause for prisoner advocates and officials looking to manage costs. The idea, though, is more than 100 years old. First popularized around the turn of the 20th century, nurseries flourished for a time, but started to close about 50 years ago, as correctional attitudes became more punitive and prison administrators began to question the costs and the effects on children.

Stephaine Reis and Major, 10 weeks old. Reis was serving time for criminal possession of a controlled substance. (Wayne Lawrence)

Today, as nurseries return to prisons teeming with an unprecedented number of inmates, the questions are even more pressing. Should institutions that limit so many basic rights allow inmates to be active parents? Most important, what does spending the first years of life in prison mean for a child?

Inside the barbed-wire enclosure of Bedford Hills Correctional Facility, a maximum-security women’s prison an hour north of New York City, about a dozen of the newest residents played within the confines of a three-foot-high baby gate. The morning program was under way in the prison’s Infant Development Center, where sunlight slanted through flowered curtains. A small boy with a pacifier banged a drum. Staff in smocks and stocking feet circulated, some rocking babies, while a toddler sat in pajamas and surveyed her options: a row of dolls on a shelf, piles of board books, crates of balls and squishy blocks. A menagerie painted on the back wall—a lion, a koala, a monkey swinging from a banana tree—stood out brightly against the cinder block.

Bedford Hills is home to the country’s longest-running prison nursery, which opened with the rest of the facility in 1901. Set amid the colonial estates and horse pastures of Westchester County, the brick buildings sit on a rise surrounded by maple and oak trees, whose leaves were just turning when I visited in October.

The prison is the reception center for all female inmates in New York, so Mayer had learned about the nursery when she landed in Bedford Hills the first time, before she was transferred to a lower-security facility upstate. While at Bedford Hills, she could sometimes see mothers and babies in the yard during their recreation period, or a row of strollers parked outside the Infant Development Center. But as she waited in the county jail a few years later—facing just over a year of additional time, and entering her third trimester—she didn’t know whether she wanted to keep her own baby in prison. “I didn’t want my son to experience what I did,” she told me. “Being locked up all the time.”

Working with an advocate she met through her lawyer, Mayer looked into community programs that would offer an alternative to prison, but none would agree to take her while she was pregnant. She ran through the list of who could take custody while she was gone. Her boyfriend had ended up with extra charges for a gun the officer had found at the motel, and was going to be locked up for another seven years. She didn’t want to ask her family, either: she and her mother still weren’t close, and she didn’t want to burden her grandmother, who had already raised several children and grandchildren and was now caring for her aging husband. So when Mayer arrived again at Bedford Hills, in December 2013, she filled out an application for the nursery. Two months later, she gave birth to her son at the local hospital. She named him DeVanté, after his father. They rode back to the grounds together in a prison van.

I first met Mayer outside the Infant Development Center, where she was picking up her son, who had just turned eight months old, at the end of her morning shift sorting packages and cleaning in the visitor-reception area. Now 24, she wore a pink T-shirt over her prison-issue pants, and her curly brown hair hung loose over the tattoo on her collarbone. DeVanté was propped on her hip, a diaper poking out of his elastic-waist jeans, sucking down a bottle.

The two of them were living with 12 other mothers and their babies in the nursery’s housing unit, one floor in a building set apart from the general population. Although Bedford Hills is a maximum-security facility, most inmates in the nursery program are less serious offenders—the screening process tends to eliminate women with a history of violent crime or involvement with the child-welfare system—and the unit looks more like a college dormitory than a cellblock. Mothers with newborns live along a corridor of double rooms, moving into singles once their babies are four months old. (The age limit for children at Bedford Hills is one year, but women who will be out before their babies turn 18 months old can apply for an extension so they can leave prison with their child.) Mayer and DeVanté shared a small room with pastel walls and a window looking out on the trees beyond the prison fence. Her narrow bed stood a few feet from his crib, photos of her boyfriend taped to a metal locker between them.

After Mayer put DeVanté down for a nap, we sat on couches in the unit’s rec room. Light filtered in from an attached sun porch, where decorations for an up-coming Halloween party were spread across the floor. The mothers spend all their time in the self-contained nursery, except while they are attending their daily programs—GED classes, substance-abuse treatment, career training—when their children are watched in the Infant Development Center. The unit has its own dining room, and a kitchen where the women can cook. They go outside for recreation in a private yard. In the evenings, they play together or watch Netflix in the rec room. DeVanté liked to settle in with a book. “He just wants to sit on my lap,” Mayer said. “He’s a mommy’s boy.”

Despite the toys and bright paint, the nursery is recognizably a prison—a fact made clear by the corrections officer stationed just inside the entrance. The seclusion makes for a sense of community—the women trade advice and babysit for one another when someone wants to go to the gym or the library—but also isolation. And the sleep deprivation that every new mother endures gets worse when all of your neighbors also have newborns crying at night. But Mayer believes that the experience has created a special bond between her and her son. “Nothing has made me want to change before,” she said. “Kids make you want to change.”

They don’t, of course, guarantee that you can. Many nursery participants have older children back home. But administrators point out that the program provides support and structure that women might not have had on the outside. “We’ve had mothers say, ‘I have two other kids, and I didn’t know the color of their eyes,’ ” Jane Silfen, the nursery director, told me. “They can connect with their babies here. If they were on the outside, they’d be doing everything but that.”

“The long-term goal is that women leave better off than they came in,” Karen Graff, the nursery manager, told me. In addition to doing administrative work—ordering baby wipes, coordinating visits from a lactation specialist and a pediatrician, overseeing clothing donations from Westchester residents—Graff, who is a trained social worker, helps mothers with daily challenges that range from soothing a baby who won’t stop crying to navigating tensions with corrections officers. “A lot of my job is just listening,” she said. “So many women have a long history of extreme trauma.” She tries to get them to reflect: How did you get here? How do you want to parent your children while you’re here? What happens when you go home?

The program seems to be working: research has suggested that women who participate in the nursery at Bedford Hills are significantly less likely to return to prison than inmates in the general population. Results like these have drawn interest from other states. A few weeks before my visit, a group of legislators and corrections administrators from Connecticut came to tour the nursery. Members of the state’s general assembly had raised the possibility of starting a similar program at the Connecticut women’s prison, York Correctional Institution, and the delegation had traveled to Bedford Hills to talk with administrators and inmates.

Eric Coleman, a co-chair of the Connecticut legislature’s judiciary committee, told me that he first learned about prison nurseries a few years ago, from a legislative clerk. The clerk had been translating for a group of prosecutors visiting from Russia. When the conversation turned to corrections, the prosecutors expressed surprise at the American policy of separating mothers from their babies. In their country, they told the clerk, children born to inmates could stay right there with them. Why didn’t prisons in the United States allow the same?

Much of the rest of the world manages to uphold public safety without routinely taking newborns from their incarcerated mothers—some with accommodations that would be unthinkable in an American prison. At the Preungesheim Prison in Frankfurt, Germany, women can keep their children on the grounds until they are old enough to go to school. Mothers with older children at home are allowed to spend days with their family as a kind of work release—cooking and cleaning and tucking their kids into bed before checking back into prison for the night.

According to a comprehensive survey from 1987—the low point for American prison nurseries—the U.S. was one of only five responding United Nations member countries (along with the Bahamas, Canada, Liberia, and Suriname) that did not generally provide accommodations for a baby born during a woman’s prison term.

Amanda Losurdo and Keegan, 12 weeks old. Losurdo was serving time for a second DWI offense.(Wayne Lawrence)

This was not always the case. The country’s first prisons exclusively for female inmates opened after the Civil War, built on the idea that specialized attention, rather than warehousing in the attics of male penitentiaries, would be more likely to successfully reintegrate law-breaking women into society. By the 1900s, a new model of detention for women, the reformatory, had cropped up in some 20 states. Whereas the penitentiary model focused on restricting freedoms, reformatories—which mostly held women for moral offenses, like prostitution and “manifest danger of falling into vice”—made it their mission to correct behavior, instructing inmates in everything from physical fitness to table manners to vocational trades.

Reformatory administrators focused on rehabilitating the women in their charge. “We must guard against institutionalizing them,” the board of directors at the Connecticut State Farm for Women declared shortly after the facility opened in 1918. “Our training here must fit them for the work they are to do when they go out.” That training often included child-rearing. Many of these early women’s prisons provided separate facilities where young children could stay with their incarcerated mothers.

Estelle Freedman, a historian at Stanford, told me that prison nurseries had originally been guided by an ideology of maternalism, the belief that innate virtues accompany motherhood. The presence of children in prison, the thinking went, could have a virtuous effect on “fallen women.” But as decades passed, that optimism waned. Drug use increased, as did the population of black inmates in the Northeast and Midwest, where the reformatory movement had concentrated, and Progressive-era reformers gave way to a generation of “corrections officials,” whose attitude toward incarcerated women was fast becoming, as Freedman put it: “There’s nothing we can do about them.”

In the 1960s, a pair of social workers who visited a nursery in West Virginia—where a prominent activist once called the presence of children “a pleasant humanizing influence”—signaled what would soon become the new correctional mind-set: “Prison is no place for a child.”

Over the next few decades, as lawmakers answered Richard Nixon’s call for a war on drugs with zero-tolerance policies and mandatory sentencing minimums, prison terms got longer, and judges were given less discretion about how to dole them out. Women—particularly women of color—counted high among the casualties. Since the 1970s, the female incarceration rate has increased twice as fast as the male rate. At the same time that incarceration became the main answer to a slate of the country’s social problems, the states that still had nurseries stopped operating the programs and repealed the laws that governed them. Through the ’70s and into the early ’80s, every facility except Bedford Hills closed; administrators cited concerns about security, insurance costs, management problems, and child welfare.

As nurseries disappeared, the prison explosion of the 1980s flung families even farther apart. Farming and manufacturing jobs were drying up across the country, and small towns and rural areas competed for prison-construction contracts and the employment opportunities they would create. New facilities were built far from the urban centers where many offenders lived, so inmates who were parents usually ended up more than 100 miles from their families—and because there were so few women’s prisons, many mothers were even farther away. Most did not see their children until they were released. And those reunions, in many cases, were brief: by the early ’90s, the rate of inmates, male and female, re‑arrested within three years of release had reached nearly 70 percent. More than half would return to prison.

Corrections officials were unprepared for the influx of women, many of whom were unmarried mothers of young children. In 1992, the National Institute of Corrections held a training to address the growing population of female prisoners. The superintendent of Bedford Hills stood up to speak about the nursery program, catching the attention of an audience member named Larry Wayne.

Wayne was then the superintendent of the Nebraska Correctional Center for Women, which had a visitation program that allowed children to stay with their mothers a few nights each month. The program not only provided an incentive for good behavior but also had what Wayne called “a therapeutic effect” on the whole population. A nursery seemed to promise even more benefits. Two years later, using Bedford Hills as a model, Nebraska’s corrections department opened a nursery of its own.

Administrators in Nebraska invited Joseph Carlson, a new hire in the criminal-justice department at the University of Nebraska at Kearney, to evaluate their program. His first results, published in 1998, showed a 13 percent drop in misconduct reports among women who joined the nursery. He also found, based on early data, that only about a third as many nursery participants returned to the prison compared with inmates who had been separated from their infants before the program started. “The potential for rehabilitating and training the mother inmate far exceeds the costs to the state and taxpayer,” Carlson wrote. He calculated that nursery supplies, staff salaries, and medical expenses would total about 40 percent less each year than foster care for the babies who would otherwise end up there, and predicted more-significant savings from a decline in recidivism. “If this trend keeps up, the program would pay for itself over time.”

Other corrections departments soon followed Nebraska’s lead. South Dakota opened a nursery the same year that Carlson published his report, and Washington State followed in 1999. When Ohio opened a nursery a few years later, prison administrators cited the promising results in Nebraska. New York released its own data in 2002, reporting that the recidivism rate for participants was half that of the general population. In 2009, Carlson published the 10-year results of his study, which showed that while 50 percent of mothers who had been separated from their newborns had returned to custody, only 17 percent of nursery participants had. By that time, nurseries had opened in Illinois, Indiana, and West Virginia.

Policy makers were interested not only in reducing the number of women in prison but also in improving outcomes for their children. Some research suggested that children of incarcerated parents were at elevated risk for academic, behavioral, and emotional problems, as well as future involvement with the criminal-justice system. More than half of the mothers in Nebraska’s nursery program reported to Carlson that their own mothers had been incarcerated. “The cycle has to be broken,” he wrote, “and education of the mother is one of the first places to begin.”
When Wyoming passed a nursery-funding proposal in 2012, the warden of the women’s prison at the time, a former employee of the Nebraska prison, told a local newspaper that he saw the impact of a nursery reaching down generations. “We want [the mothers] to be successful at raising those children,” he said, “so those children don’t repeat the sins of the parents.”

The claim that nurseries could benefit children as well as their mothers has a radical extension: children not only should be allowed in prison but might be better off there. That idea is, unsurprisingly, controversial.

“I don’t think any children should be in prison,” James Dwyer told me last year, as legislators in Connecticut considered a proposal for a nursery. “Period.”

Dwyer, a family-law professor at the College of William & Mary and the country’s most outspoken critic of prison nurseries, disputes the idea that advocates of the programs have child welfare in mind. Screening inmates for fitness as parents based on a history of child abuse or violence, he told me, is missing a larger point: incarceration itself is a marker of unfitness. In a paper published last year in the Utah Law Review, Dwyer further argued that allowing mothers who have broken the law to keep their children in prison is not only unwise but unconstitutional:

There would likely be widespread public outrage if any state began putting mentally disabled or senile adults in prisons with incarcerated relatives in the hope that this would reduce recidivism and provide some benefits to those incompetent adults.

Objections to putting innocent children in prison go back to the heyday of nurseries. “If we were more than three degrees removed from the level of the chimpanzee,” a writer for the Newspaper Enterprise Association declared in 1930, “the bare announcement that thre [sic] was even one baby in prison, anywhere in the land, would stir us to a yell of protest that would rock that prison to its foundations.”

The shortcomings of raising a baby in prison are probably most obvious to those actually doing it. DeVanté was an easy infant, Alyssa Mayer told me, even taking naps when they were closed in their room for the twice-daily attendance count. But now he wanted to crawl around and explore. He had started scooting up and down the corridor outside their room and lurching around the rec room, holding on to couches for support. He would be 14 months old when she was up for release, and she was already thinking about how much catching up they had to do: he had never seen the ocean, never been on a swing. “Sometimes I think I’m selfish for keeping him here, even though he doesn’t know what’s happening,” she said. “If he was home, there’s so much more he would experience.”

Those who advocate on behalf of incarcerated mothers are also quick to point out the drawbacks to parenthood in prison. In February, the Women in Prison Project at the Correctional Association of New York released a report finding that pregnant inmates were routinely shackled during labor and recovery—sometimes with waist chains after a C-section delivery—despite a 2009 law restricting the practice. Other problems are more subtle. Gail Smith, the founder of Chicago Legal Advocacy for Incarcerated Mothers, served on an advisory committee for the Illinois prison nursery a decade ago and recalls the “control-oriented thinking” that permeated the early planning process. “Staff members were discussing the ‘parameters’ of breast-feeding and when mothers would and would not be permitted to feed their babies,” she told me. “I was appalled that these administrators could think … that it was appropriate to deny a hungry infant sustenance until the scheduled time convenient for corrections officers.” Advocates argue that funding could be better invested in community-based alternatives to incarceration, where women can parent their newborns without all the restrictions inherent to the prison environment.

Such alternatives, though, remain scarce for pregnant women—and many have no better places for their newborns to go. Most incarcerated mothers, unlike incarcerated fathers, were primary caregivers for their children before getting arrested, and family members or others who take custody are in many cases poor, sick, or overburdened. Researchers don’t know exactly why children of inmates might be at elevated risk for behavioral problems, but evidence suggests that the disruption of family life could play a significant role. For Dwyer, this is a reason for prison officials to encourage adoption.

But short of that extreme, prison nurseries may actually be the most stable environment for babies of incarcerated mothers. New York implemented a legal standard in 1930 for nursery admission matching the one that guides custody decisions outside prison: the best interests of the child. In 1973, an inmate in a New York jail named Kathleen Apgar, who had given birth while awaiting trial for murder, brought a suit against the local sheriff for taking her newborn son from her at the hospital. The state supreme court, ruling in Apgar’s favor, wrote that in addition to adequate food, shelter, and medical care, a child’s best interests included “the constant care and attention of its natural mother”—even if the mother was an accused murderer. That notion, which is at the heart of the disagreement between nursery advocates and critics like Dwyer, is only now being researched in depth for children starting life inside prison.

In 1945, an austrian-born psychoanalyst named René Spitz conducted a seminal study of childhood in incarceration. He used a 16-mm camera to film two groups of babies and toddlers—one being raised by their mothers in the nursery of a penal institution for delinquent girls, and the other by the staff of a “foundling home,” a shelter for abandoned youth. His findings revealed developmental gaps. Even the oldest children in the foundling home, who were between 18 and 30 months old, were incontinent. Few could walk, talk, or eat without assistance. Even though the facility was kept clean and a physician visited every day, more than a quarter of the children died from a disease outbreak.

Which makes what Spitz found in the nursery especially striking: Children who were less than a year old could already speak a few words. They were so mobile that without close supervision, they would shimmy up the bars of their cribs and dive onto the floor. The biggest challenge, Spitz reported, was “how to tame the healthy toddlers’ curiosity and enterprise.”

Spitz searched for an explanation for the contrast. Food and housing conditions in the two institutions were similar, and the children in the foundling home came from more-favorable family backgrounds. The most significant difference? The “nursery provides each child with a mother to the nth degree,” he concluded, “a mother who gives the child everything a good mother does and, beyond that, everything else she has.”

Seventy years later, Spitz’s proposition has gained support from the first longitudinal study of prison-nursery outcomes. Starting in 2003, a team of researchers led by Mary Byrne, a professor at the Columbia University School of Nursing, followed 100 children and their mothers as they went through the nursery program in New York and reentered their communities. (The study participants were drawn from Bedford Hills and a neighboring medium-security facility, where the New York corrections department had opened a second nursery program in 1990. The two programs consolidated a few years ago.)

Byrne’s research is based on attachment theory—a line of thought that surfaced about a decade after Spitz’s study, holding that children develop a secure sense of themselves and others through the stability and attentiveness of caregivers in the first stages of life. The theory suggests that early caregiving can have profound implications on everything from brain development to the quality of future relationships.

For a paper published in 2010, Byrne’s team interviewed nursery mothers and found that only a third had formed secure attachments to their own parents. So what the researchers discovered when these mothers’ babies reached their first birthday was surprising: 60 percent showed signs of secure attachment, on par with a comparison group of children growing up in stable middle-class families outside prison, and a significantly higher rate than that of sample groups of at-risk children. “Their children should be in trouble,” Byrne told me. “But they’re not.”

Looking more closely at the results, the researchers found that children who stayed the longest in the nursery had the best outcomes. About half of the mothers had less than a year left on their sentence when their baby was born, and had returned home by the time of the assessment. The rate of secure attachment among those children, while still not significantly different from the rate for the comparison group of middle-class children, was lower than among their peers who had stayed in the nursery for a full year. Byrne hypothesized that rather than being harmed by the correctional setting, the babies actually benefitted from the structure the prison provided—particularly the restriction of drugs and alcohol, as well as the parenting support their mothers got from staff and other inmates. (The longitudinal study included parenting guidance from a nurse practitioner, which Byrne believes also contributed to the outcomes.)

Crystal Degnitz and Aliviana, 10 weeks old. Degnitz was serving time for attempted burglary. (Wayne Lawrence)

James Dwyer points out that the attachment findings might be optimistic if extrapolated to nursery participants as a whole. The results included only children who were with their mothers at the time of assessment. As Byrne documented in a subsequent paper, more than 40 percent of pairs in the longitudinal study were separated before the mother left prison, in most cases because the baby reached the age limit or because of disciplinary action against the woman. Byrne noted that the misbehavior in those cases did not seem to pose any obvious threat to the children. (At Bedford Hills, the kind of mistakes any sleep-deprived new mother might make—leaving an extra blanket in the crib, drifting off with your baby on your chest—can become grounds for losing custody. The safety and well-being of the babies is the program’s primary concern, administrators told me, and behavior that puts them at even slightly elevated risk cannot be tolerated.)

Although separation in the first year can be damaging, experts say that babies who form secure attachments to their mother early on may be better off even if they are later split up. A study led by a member of Byrne’s team and published last year compared a group of 3-to-5-year-olds who had spent between one and 18 months in a prison nursery with a group of children the same age who, as infants or toddlers, had been separated from their incarcerated mothers. Most of the children were living with their mothers at the time of the study, but some in each group were with alternate caregivers. They faced comparable amounts of trouble at home, measured by the adults’ drinking and drug use, reliance on public assistance, and harsh treatment. But the preschoolers who had lived with their mothers in the nursery displayed significantly lower levels of depressed, anxious, or withdrawn behavior. The study concluded that participation in a nursery program may be a “buffer” against environmental risks when children leave the prison.

Byrne is now starting to analyze how the children in the longitudinal study fare as they go through grade school. What her team has found so far, she told me, is that children raised in the nursery perform no differently from other kids across a number of measures. The study design is limiting; for example, her team couldn’t randomly assign women or children to the nursery. But Byrne’s research suggests that prison nurseries could provide children of incarcerated mothers a better starting place than any existing alternative.

Alyssa Mayer and DeVanté left Bedford Hills at the end of April. Her mother—now the closest family she has in the area, since her grandmother moved out of state—came to pick them up the day they were released. It had been an emotional morning: saying goodbye to people who had become like family to her and her son, and not knowing what would come next. DeVanté had never ridden in a car without bars on the windows. They stopped at a grocery store on the way home, and he gaped as they moved through the aisles, picking out fresh fruits and vegetables. After dinner, she curled up in bed with him to watch TV—for the first time, just the two of them.

Three weeks later, when I visited Mayer at her mother’s house—a tidy split-level about half an hour from Kingston that she bought several years ago—DeVanté seemed to have settled into life on the outside. He swiped on an iPad and babbled at Siri, toddled between rooms playing peekaboo, helped himself to a bowl of candy. His hair had grown out in thick curls, and he had a gap between his front teeth that showed when he smiled. Mayer lifted him onto the kitchen counter and pulled up a Barney sing-along on YouTube. He bobbed his head and pumped his small hands toward the ceiling. She laughed. At Bedford Hills, she’d had a radio that she would play so he could dance, but only a couple of stations came through. “That’s what happens when he listens to hip-hop.”

Mayer told me DeVanté had brought her closer to her mother. “His bond with her is keeping my bond with her,” she said. And she knew she was lucky to have a place to go. Still, she looked forward to getting a job and moving into her own place in the city. She had always wanted to be a nurse, but knew that her record could keep her from getting a license. For now, she was open to anything that would pay the bills. At Bedford Hills, she hadn’t had to worry about things like food and shelter, diapers and child care. Leaving the program, she knew her choices mattered for both of them.

“It’s not like I can just get up and decide, Tonight I’m going to go to the bar,” Mayer said. “He gives me that second thought I should have had a long time ago.” That weekend, her mother had offered to babysit so she could go out with friends, for the first time since she’d come home. They were planning to go to a restaurant in the next town: she wanted to stay away from the nightlife in Kingston. She had broken things off with DeVanté’s father, who was still in prison upstate, because she’d heard he was keeping contacts in the streets. “You can’t be in the middle of picking yourself up and pick somebody else up at the same time,” she said. “I feel like I have more-important things to put my effort into.”

Before I left, she picked up a potted plant from the kitchen window, a ruby globe with spiny ridges on a corrugated green stalk. “It’s a moon cactus,” she said. “It was originally just a regular green cactus, but this happens”—she pointed to the globe—“when it lacks chlorophyll.” The mutation that gives the moon cactus its bright color also keeps it from thriving on its own, so the seedlings have to be grafted onto another succulent so they can grow. She and DeVanté had bought the plant for Mother’s Day. She set it back on the windowsill, where it could soak up the light outside.