My husband, on the other hand, was never referred to as an addict, even though he was actually being accused of using one more illicit substance than me — methamphetamine in addition to heroin. It may be hard to understand why something like this matters. After all, don’t people use the word “addict” all the time?

Get Talk Poverty In Your Inbox

There is an ongoing debate among the addiction treatment and harm reduction communities about which terms should be used when referring to drug use and addiction, and in what settings. In 2017, the Associated Press Stylebook updated their recommendations for reporting on addiction and drug use to exclude potentially stigmatizing terms such as “addict,” “alcoholic,” and “drug abuser,” except in the form of direct quotes. Instead, they recommend using person-first language, such as “person with a substance use disorder” or “people who are addicted to opioids.” These changes aligned with updates to the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders, as well as recommendations by the Office of National Drug Control Policy.

Others followed suit: a 2018 update to the National Institute on Drug Abuse’s blog specified that they no longer use terms such as “addict” or “substance abuser” (though the institute’s name remains unchanged), and in 2020 dictionary.com updated their website to replace all uses of the word “addict” as a noun with terminology such as “person with an addiction” or “habitual user.” They also updated their definition of the word “addict” to note that some might consider it offensive, and added a lengthy sensitivity note explaining: “addiction is the complicated result of genetic predisposition intersecting with dysfunctional behavior, neurochemical modification, environmental factors, and social influences,” and that many members of the treatment and recovery communities advocate against its use.

Controversy around terms such as “addict,” “alcoholic,” and “substance abuse” has been boiling for years. Many members of the medical, recovery, and harm reduction communities are happy to see changes like this implemented. For example, Olivia Pennelle, founder of Liv’s Recovery Kitchen and a journalist in long-term addiction recovery who covered the language debate for The Fix in 2018, said the dictionary.com changes are “important,” adding: “Only 10 percent of people [with a substance use disorder] get access to treatment…[and] a major factor is stigma. If we can do anything to change that, we should.”

On the other hand, Amy Dresner, also a journalist in long-term addiction recovery who authored an autobiography titled My Fair Junkie: A Memoir of Getting Dirty and Staying Clean, said she considers the use of language such as “junkie” and “addict” to be empowering when reclaimed as self-identifiers. And she says terms such as “person with a substance use disorder” fail to “convey the horror” of what she experienced during her active addiction.

“All that PC language feels like putting lipstick on a pig and hiding it more…[the changed terminology] sounds better, it sounds like you have empathy but does it really change somebody’s opinion?” she asked.

Recent research backs the moves by dictionary.com and the Associated Press. A series of studies published in the Journal of Drug and Alcohol Dependence in 2018 found that several terms, including the word “addict,” were associated with negative social perceptions. It also found that these terms produced negative biases significant enough to impact people’s access to healthcare; for example, both treatment professionals and members of the general public were more inclined to recommend incarceration or other punitive measures to those labeled an “addict” or “substance abuser” while a “person with a substance use disorder” was more likely to be deemed in need of medical care.

Robert Ashford, the lead researcher in the language studies, explained a potential cause in a story I wrote for Filter Mag. “[Language is] the primary way we communicate…The cliché ‘words have power’ is the truth.” Regarding language employed in court settings specifically, he added: “It’s not that those [terms such as ‘person with substance use disorder’] are inherently positive; it’s that they are less negative than the pejorative terms that have been created over time. These things have a really strong emotional reaction to most people. We don’t need to do [courtroom actors] any favors by…using language that has come to mean something biased.”

“How you speak about the person you’re representing is how other people will see them and so using the correct language is extremely important,” added Dinah Ortiz, a harm reduction-oriented parent advocate located in New York City. “Language is like a stepping stone, then comes the harder stuff, but it starts with language…if we don’t care about calling a person a ‘junkie,’ a ‘dope fiend,’ a ‘crackhead,’ then we don’t care about that person.”

In my Florida State child services case, my husband and I both received the same charges — neglect and imminent risk of harm — and we both had our parental rights terminated in early 2020. On paper, we shared the same nightmare outcome at the end of a case riddled with misrepresentations of fact, blatant bias, and government overreach at its darkest. But there was a palpable difference in the courtroom between his treatment and mine. While my every word was interrogated with suspicion — I was never even counted as having income despite clearly being able to pay my bills and child support — my husband was rarely questioned. I was criticized for circumstances he and I shared, such as not having a car and relying on his parents for rides to our supervised visits, while he usually escaped mention.

At the disposition for our initial trial, when the judge determined that our daughters were unsafe in our care and should remain with their paternal grandparents while we completed a slew of tasks to try to regain custody, the judge cited as her reasons for issuing these charges against me: “The mother is an extraordinarily educated and gifted individual. You have a gift for language, both oral and written. Unfortunately, the Court finds that you could probably sell ice to an Eskimo.”

Although she issued the same charges against my husband, the judge stated that she “found the father’s testimony essentially credible before this Court,” and mentioned that he acknowledged having a “substance abuse history,” something I likewise acknowledged about myself (though I clarified it as a substance use disorder, as per the DSM IV and V).

While it is impossible to pinpoint my classification as an “addict” versus my husband’s as a “person who uses heroin and meth” as the reason for our differences in courtroom treatment, it can’t be ignored that this experience aligns near perfectly with the outcomes of Ashford’s experiments.

And, as explained by Sheila Vakharia, a former social worker and the current deputy director of the Department of Research and Academic Engagement for the Drug Policy Alliance: “When you refer to someone as an ‘addict,’ and you make salient one person’s single relationship with a drug or several drugs, what happens is you then start to see that person through that lens of that one characteristic or trait, and it can make it hard to see the complexity of a person’s identity.”

Still, it’s hard to know how much, if at all, my case would have changed if I’d not been labeled an “addict” at the outset. Would my daughters be home today? Or would I merely have had a slightly more comfortable courtroom hanging?

In 2020, Allegheny will begin applying a similar algorithm to every family that gives birth in the county, with the goal of linking families in need to supportive services before a maltreatment case is opened. But some critics insist that it will be just another way for government to police the poor.

The new program is called “Hello Baby.” The plan is to eventually apply it across the county, but the January launch will begin in only a select few hospitals. Like the Allegheny Family Screening Tool, the Hello Baby algorithm analyzes family data to apply an individual family score.

Emily Putnam-Hornstein, who helped design both programs, told TalkPoverty that Hello Baby uses slightly different data than the child maltreatment algorithm, which was criticized for targeting poor families because much of the data used was available only for people who used public services.

Get Talk Poverty In Your Inbox

“This is a universal program,” explained Putnam-Hornstein. “In the [child services] model the county was being forced to make a decision after an allegation had been received; in this case we’re taking about more proactively using data … so we wanted that to be built around universally available data.”

But these exclusions don’t guarantee that the data will not end up targeting low-income families again. “They rely on data where the county has the potential to have records for every family,” said Richard Wexler, the executive director of the National Coalition for Child Protection Reform. “The county acknowledges they will probably use data from [Child Protective Services], homeless services, and the criminal justice system, so yes, theoretically everyone can be in that, but we know who’s really going to be in it.”

An overview provided by the county online cites “birth records, child welfare records, homelessness, jail/juvenile probation records” as some of the “available service data” incorporated into the predictive risk algorithm, indicating that Wexler’s assessment was absolutely correct. Although that data is potentially available about anyone, several of these systems are known to disproportionately involve low-income people and people of color.

Putnam-Hornstein said via email that the Hello Baby process is “truly voluntary from start to finish.” A family can choose to drop-out of the program or discontinue services at any time.

The option to drop out will be presented at the hospital, when families are first told about the program. A second notification, and chance to opt-out, will then be made by postcard. If a family doesn’t respond to the postcard, they are automatically included in the next phase of the program, which involves running available data through the system to determine how much social support each family needs.

According to Putnam-Hornstein, scores will be generated about four to six weeks after birth for families that do not choose to opt out (or who are too busy to realize they want to). Once a family is scored, what happens next varies based on which of three tiers they fall into.

Under the “universal” tier, the most basic approach, families receive mail notifications about resources available throughout the county. Families grouped in the second, “family support,” tier will receive a visit from a community outreach provider and an invitation to join one of 28 Family Support Centers located around the city of Pittsburgh.

The “priority” tier engages families with a two-person team made up of a peer-support specialist and a social worker who will work closely with the families to identify their needs and partner them with appropriate providers. It is designed to be an individualized program that grants families access to the full range of support services available on a case-by-case basis. That could mean helping a parent navigate the complexities of applying for housing assistance or ensuring timely placement in a substance use treatment program. The county said in its promotional material — which was reinforced by Putnam-Hornstein over the phone and by email — that choosing not to engage with any aspect of the program will not lead to any kind of punitive action.

But parents who need supportive services still have reasons to fear intervention from child services. The reality is that any program putting families in contact with social service and medical providers means, by default, also putting those families at greater risk of being reported to child services by placing them in more frequent contact with mandated reporters.

A mandated reporter is someone who is legally required to report any suspicions of child maltreatment they encounter. The intention is to ensure timely detection of as much child abuse and neglect as possible, but data have not shown that an uptick in mandatory reporting equates to more child safety.

In Pennsylvania, nearly anyone who regularly interacts with children in a professional or semi-professional capacity is legally considered a mandated reporter. An unfortunate side-effect of the mandated reporter system is that even though a referral program like Hello Baby is not directly involved with child services, participating families will always be haunted by the possibility of coming under investigation.

Putnam-Hornstein assured that family’s scores will not be retained or shared with child services, even for families under investigation — but noted that “it is possible that child welfare workers could infer the level of risk if the family has voluntarily agreed to participate in Hello Baby Priority services and a child welfare worker learned that when gathering family history.”

It’s clear that the new program is not designed to get families involved with child services, although it is spearheaded by the Department of Human Services, which oversees the Office of Children, Youth, and Families that conducts child maltreatment investigations and responses. Rather, Hello Baby was created with the goal of offering a more equitable way to expedite service referrals for families with new children who need them.

“Universalizing the assessment of social needs at birth is the only way to avoid discrimination,” said Mishka Terplan, an obstetrician and addiction medicine physician, who was not talking specifically about the Hello Baby program. He observed that patients with obvious social needs, such as those suffering from acute addiction, were often screened and referred for other issues, like postpartum depression or housing assistance, while other parents’ needs were going undetected and unaddressed. “That seemed unfair,” he lamented. Terplan believes that universal screening programs would eliminate both the disparity between services rendered, and reduce the stigma attached to needing behavioral health treatment and other social supports.

Hello Baby’s creators hope that offering families these programs before there is a child maltreatment complaint can help keep them out of the system altogether. But by using imbalanced data points like child welfare history, homeless services, and county prison history to auto-generate scores, it assumes poverty as the main basis for family need. While poverty does generate certain needs, it is not the only indicator for the whole range of unique social supports that new parents require, such as mental health screening or child care assistance.

A system that continues to embed data that target the poor may only end up automating the social inequities that already exist, while placing vulnerable families under increased scrutiny by mandated reporters for the child welfare system — even if it intends to serve as a universal screening process that helps families avoid punitive interventions.

“As long as the system confuses poverty for neglect, any form of such screening is extremely dangerous,” said Wexler.

I learned this early on in my case, which began in April 2018. The services required for me to reunify with my daughters included trauma-based individual therapy, a psychological assessment, substance use treatment, and parenting classes. I remember going to the first of these, my psychological assessment, and spending 20 minutes in the office arguing over paperwork.

In order to complete the mandatory assessment, I had to sign a consent form that would allow the assessors to send their findings to child services. But when I asked what their “findings” included, it was not simply a diagnosis or treatment recommendations. Instead, it could be the full readout of the evaluation.

Get Talk Poverty In Your Inbox

Everything I said during this invasive and highly personal evaluation would be sent to my caseworker, his boss, the state attorney, the guardian ad litem and her attorney, my attorney, my husband’s attorney (and by proxy my husband), and the judge. Any time my caseworker was changed, which had happened once already and would happen two more times, the new stranger would also be privy to the contents of my psychological evaluation.

But the service was mandated. Without this evaluation, I was not allowed to engage in therapy, substance use treatment, or parenting classes. So I had to do it.

This is standard fare for families involved with the child welfare system. It focuses on the collection of parents’ information, and control over what those parents do and how they behave, not, as proponents of the system fervently claim, on ensuring the safety of children. Like probation and parole, child welfare involvement becomes one more way for the government to keep tabs on people.

As it would turn out, the report would not be a transcript of my evaluation. Rather, it was an interpretation, in which the evaluator handpicked which details to include. She described me as eccentrically dressed. (I had worn a floral dress and flats, but have visible tattoos and a facial piercing that can’t be removed.) She detailed that sometimes I answered questions right away, and sometimes I paused to answer them, which sounds normal enough but seemed sinister and awkward when inserted in the context of the psychological evaluation. She wrote that I suffered from insomnia, even though I’d repeatedly told her I was tired on that day specifically because I was worried about my husband, who had been hospitalized earlier that week. She generalized my history with drug use to make it appear that I had been addicted to heroin since I was a teenager, which is not the case.

In the end, the recommendations didn’t include anything besides the talk therapy I’d already been mandated to take — but the inclusion of all her other highly subjective details handed my opposition a slew of quotes they could use to describe me as eccentric, erratic, and ill-equipped to handle the daily realities of parenting.

It was an evaluation I had no choice but to attend, which should have been a doorway to resources and help for my PTSD. Instead, it served as an intelligence-gathering exercise for the people separating me from my two young daughters.

Other parents have experienced similar issues with the services that are supposedly in place to help them. Kim, a mother in Alabama who asked that her last name not be shared, has been involved with child services since January 2019. Kim’s case was triggered by her arrest when she failed to appear for a court date, but she says the crime she is accused of was actually committed by her abusive partner, who forced her to take the blame by threatening her life.

When her caseworker learned Kim was experiencing domestic violence, she told Kim to move into a shelter. Which she did, but only for about a month. She said the shelter had stringent rules, which included nightly curfews and that she report her whereabouts when she left the grounds.

Kim was never told by the shelter that this information was shared with her caseworker, but she figured it out when her caseworker suddenly knew details only the shelter had. “She knew my comings and goings there. Knew when I met with the therapist, left for work; all of it,” Kim said.

“The sheer fact and status of having a child places you in a situation where you can no longer openly and honestly express what’s happening in your life to mandated reporters because you’re facing family dissolution, or, at minimum, family surveillance,” said Erin Miles-Cloud, a former parent defense attorney in New York who is now the co-founder for Movement for Family Power, a parent advocacy group. Exactly who falls under the category “mandated reporter” varies by state, but they are typically frontline workers such as nurses, doctors, therapists, and teachers, who are required by law to report any suspicions of child maltreatment. She specifically cites shelters, hospitals, and schools as some of the “biggest offenders” when it comes to reporting parents who are seeking care to child services.

While speaking with me for a story I wrote for Filter Mag about the way child services targets parents who use drugs, a nurse named Tracy Longbreak told me about her experience with the “mandatory” aspect of mandated reporting. When a mother came into her emergency department with her baby while smelling of marijuana but appearing prepared, competent, and tidy, Longbreak was told by her superiors that she had to call in the report or risk her job.

Ultimately, the best she could do was include her perceptions of the mother in her report and hope that her positive remarks would offset the accusation of neglect via marijuana intoxication (which was not yet legal in the state of Oregon, but is now).

“In North Carolina, the mandatory reporting law is around any abuse that may have taken place by the caregiver,” said Julie Owens, a survivor of domestic violence who now consults with organizations around the country advocating better practices for people who have experienced violence in the home. “The protective parents who go into domestic violence shelters—primarily mothers—are not the abuser, but unfortunately they are reported as or regarded as abusers because they haven’t reported the abuse that their children have been experiencing, and they often end up being punished or deprived of their children as a result of this.”

Put together, this all means that service providers can be forced to act as eyes and ears for child services, even when they don’t want to. But more reports doesn’t equal more child safety. In Philadelphia, for example, mandated reporting laws were drastically expanded after the Penn State child sex abuse scandal. In an article published in Pediatrics in 2017, Mical Raz wrote that “there is no indication that the increase in reporting has improved the safety of Philadelphia’s children, and there is reason to believe it may detract.” Some of these detractions included increased hotline calls resulting in an overburdened system less able to make accurate safety assessments, and a heightened risk of family separation for low-income families. Later Raz noted that “fear of reporting may prevent families from seeking help, whereas assurance of confidentiality has been shown to increase help-seeking behaviors.”

The majority of substantiated maltreatment charges in the child welfare system are for neglect, which typically means issues like lack of food, child care, or weather-appropriate clothing – things that could be fixed with better social supports or a little more money. Creating a system that encourages families to seek help should be the goal for any agency in pursuit of children’s health and safety.

But forcing more and more providers and even laypeople to report on parents whether they want to or not achieves the exact opposite of that goal. Instead, it creates a cyclical, hypocritical system in which parents are afraid to seek assistance for fear of being punished because of the issue for which they need help, then punished for not seeking that help on their own. It also harbors distrust in therapeutic situations, which renders impossible any kind of genuine recovery.

The network of surveillance that child welfare-involved parents become trapped inside will continue to harm families like mine until it is lifted and parents are allowed to seek help and engage with services without simultaneously leaking the most intimate details of their lives.

These numbers are staggering. When you consider the impact of childhood trauma — which tells us that kids who experience or witness abuse are more likely to develop a slew of physical and mental illnesses as adults — those numbers are infuriating. And baffling. Domestic violence can be hard to escape, especially for those who have been in the mire of it for years, but once kids become involved, shouldn’t that be enough motivation to leave?

It’s this question, and the assumed answer, which drives “failure to protect” laws in child welfare programs across the United States. Essentially, failure to protect laws charge a parent with not doing enough to shield their child from witnessing or experiencing abuse. Virtually every state pursues some form of failure to protect charges within the civil child welfare system. These laws are aimed at the non-abusive parent living in an abusive household. Usually, the parent has been subject to intimate partner violence. But the laws can also be used in households in which the child is the victim of one parent but not the other.

Get Talk Poverty In Your Inbox

While these laws were written with the intention of penalizing a parent who neglects the safety and/or well-being of their children, they all too often make unsafe environments even less safe by penalizing non-abusive parents living in an abusive household, and can become the basis for temporarily or permanently removing children from the home. They rarely leave room to consider the complexities of intimate partner violence, instead relying on assumptions and stereotypes that are incapable of capturing the nuanced reality of family bonds.

In January of 1999, Sharwline Nicholson decided to end her relationship with the father of her infant daughter. He lived in South Carolina, and had been crossing state lines each month to visit Nicholson and their daughter in New York. But when she ended the relationship, he responded with violence.

She called 911 and made arrangements with a friend for the care of her two children while she stayed overnight at the hospital. The next day, Nicholson was notified by the New York Administration of Children’s Services (ACS) that both of her children had been temporarily removed from her care on the basis that she had failed to protect them from witnessing the violence that had been inflicted upon her by her former partner. At the time, this was considered a form of neglect.

Nicholson would eventually win back custody of her children, but would be placed on a child maltreatment registry. This action kickstarted a lawsuit that would eventually lead the New York Court of Appeals to rule in 2004 that a parent’s inability to prevent a child from witnessing abuse could not be a sole factor for removing a child. Child welfare reform activists celebrated the decision.

“What Nicholson actually did was not just to change the attitude toward victims of domestic violence,” said David Lansner, a civil rights and family law attorney who represented the plaintiffs in the Nicholson case. “Neglect had to be shown as a serious matter; you had to show that … there was imminent danger of serious harm and not just the possibility of harm. … [Child services] and the court had to balance the harm that would result from removal against the risk of leaving the child at home, so you couldn’t just ‘take the safer course’ because removal was harmful to kids and shouldn’t be done unless it was really necessary.”

Unfortunately, New York is relatively unique in that respect; other states, lacking a case like Nicholson, still remove children for the possibility of harm that caseworkers and judges interpret by a parent’s “failure to protect” her children from being in a household where abuse takes place.

Erin Miles-Cloud, who formerly worked as a parent attorney in New York and is currently one of the co-founders of the advocacy group Movement for Family Power, explained the ways in which some of the better-resourced, urban systems can still fail families, even today. “Because New York has this middle ground of family shelters, ACS sees it as an unreasonable option to stay in a home where intimate partner violence is occurring,” she said.

What many people don’t realize is that — in New York City — parents who access a shelter as the result of domestic abuse will automatically be moved to a different borough, meaning a change in school district for their children, not to mention the loss of access to support networks, such as friends and family or trusted child care, lack of which can easily become another maltreatment charge. There’s also no guarantee as to what type of housing the family will receive or for how long. In some cases this could mean dormitory-style living for a year or longer, with no access to even a personal refrigerator.

But even the most comfortable, “home-like” shelters remain government-funded institutions — which means they come with restrictive rules, such as nightly curfews and rigid limits on how many days a parent can be away from the shelter, even to visit family. They are also a source of constant surveillance for the families housed inside. Miles-Clouds calls shelters and hospitals among the “largest offenders” when it comes to calling in new maltreatment reports, and notes that New York ACS often uses shelters as “second or third eyes on a family” when arguing a related case in court.

Because child welfare agencies self-report their data, and failure to protect  is not an independent maltreatment category in itself (these cases typically fall under the “neglect” umbrella), it is difficult to know exactly how many non-abusive parents end up being investigated because they were victims of abuse who sought help, or because their children reported being harmed by someone else in the household. But we do know that most states do not have even the mild protections enjoyed by families in New York. That means a child can be removed if the state convinces a judge they have been or will likely be psychologically harmed by witnessing the abuse.

Some states will also pursue criminal charges against victims of intimate partner violence who have children in the home. In six states – Oklahoma, Missouri, Nebraska, Nevada, South Carolina and West Virginia – non-offending parents face potential life sentences for failure to protect charges, and in Texas the maximum penalty is 99 years. Last year, the Associated Press reported on the case of Tondalao Hall, a mother whose boyfriend was sentenced to two-years time served (meaning he had already completed his jail time while waiting to be sentenced) for beating her children, including a three-month-old infant. Hall, who was never accused of harming her kids, is currently serving 30 years in prison for not calling the authorities on her boyfriend.

Latagia Copeland-Tyronce, a parental rights advocate and the founder/executive director of the National African American Families First and Preservation Association who spoke to TalkPoverty about her experience, knows first-hand how devastating it can be to be accused of not protecting children from another person’s abuse. She first faced the traumatic confusion of a failure to protect charge in Toledo, Ohio, in 2013.

She was 26 years old and had been involved in an abusive relationship for 10 years. What she did not know, however, was that her daughters’ father was also sexually abusing her three eldest girls. When one of Copeland-Tyronce’s daughters finally disclosed the abuse to her sister, she promptly contacted child services.

Copeland-Tyronce immediately left her children’s father. She also cooperated with the criminal case that would ultimately land him a 30-year prison sentence. But this was not enough for Lucas County child protective services. They claimed she had known about the abuse and had failed to protect her children both from witnessing the violence perpetrated against her, and from the sexual abuse which they had experienced.

“My children never said that I knew anything or that I was involved in the abuse and I was never charged with a crime related to the case,” countered Copeland-Tyronce.

Less than a year after the initial removal, her parental rights were terminated and all six of her daughters were adopted to other families. When she gave birth to a son in 2014, by a different father and with stable housing in place, he was also removed from her custody.

“Because I had a [termination of parental rights] TPR, failure to protect, with my daughters. No other reason,” she said. At the time, the first TPR was still under appeal.

Candis Cassioppi, a mother based in Athens, Georgia, had her youngest child removed from her in the hospital after giving birth, she told TalkPoverty. The removal was prompted by an incident of assault by her child’s father perpetrated against her during her pregnancy.

Although she initially called the police and sought medical attention — causing those injuries to become part of her medical record — she ultimately declined to press charges or testify against her abuser. After her son’s birth, this incident became a reason to claim she was failing to protect her children from harm. Now, she is court-ordered to participate in a slew of activities, including domestic violence groups and parenting classes, in the hopes of regaining custody of her infant.

Like mandated reporting laws, which require certain professionals and institutions to report suspected child maltreatment, failure to protect laws and policies are in place, purportedly, to ensure that child maltreatment does not go unreported. “If a child dies in the home because there was a batterer who was so dangerous that the victim-partner couldn’t protect [the kids] … we’re still liable to make sure that the child stays safe,” explained Mary Nichols, a now-retired administrator at Los Angeles County’s Department of Children and Family Services (DCFS), to the California Health Report in 2015.

But she also admitted in the same article that the laws are confusingly vague: “If you look up California Welfare and Institutions Code 300 and just read the definitions of ‘failure to protect,’ you can see how broad they are … [If] somebody would like to craft legislation to make it more workable, in terms of protections for domestic violence [victims], that would be great. It’s a pretty raw tool that we have.”

As the cases detailed in this article demonstrate, the reality of domestic abuse is far too complex to address with vague, generalized laws. Instead of protecting families, these blanket laws mean that parents who experience domestic violence may end up burdened by a fear of reprisal for reporting that violence. Take Cassioppi’s case, for example. Her baby was born healthy; had she not called the police and sought medical attention after being assaulted during her pregnancy, she likely would have walked out of the hospital with her newborn in arms. And Copeland-Tyronce now asserts that if she were to ever encounter intimate partner violence again, she would “not at all” feel safe calling the police for help.

Lansner said implementation and caseworker attitudes are major problems with the way domestic violence cases are handled within the child welfare system. “The caseworkers just don’t get it,” he said, adding, “the caseworker might go to the home, find the guy there in violation of a protection order and then remove the children instead of calling the police and having him arrested, which is what [the caseworker] should do.”

Parents who experience intimate partner violence also face a number of other complexities that caseworkers and judges don’t always take into consideration when charging these parents as culpable for traumatizing their kids by proxy. For example, one study found that 99 percent of domestic violence survivors had also been subject to economic abuse, a form of financial control that can leave them stranded without the resources necessary to secure independent housing or provide for their children’s basic needs. Because lack of appropriate shelter, clothing, and food also fall under the child services maltreatment category of “neglect,” this leaves many non-abusive partners trapped between the crosshairs of a failure to protect and a failure to provide charge. Either way, they’re ending up on the maltreatment registry for neglect.

By necessitating that caseworkers identify concrete harm toward a child before removing her from the home, New York has found a way to slightly balance a system designed to punish parents simply for being unfortunate enough to experience abuse. Although their system is far from perfect — as Miles-Cloud noted, it funnels parents into a less-than-ideal shelter system, and the law still does not address the caseworker bias that concerned Lansner — it provides a template which other states could use to begin the process of clarifying these laws.

Ideally, however, survivors of domestic violence should be met with compassion and provided with services that help their families heal and thrive intact. It seems, instead, that as long as failure to protect charges exist, the child welfare system will continue to promote a culture of secrecy surrounding intimate partner violence, thus validating the very abuse it claims to condemn.

That year, the city applied for a $1.34 million U.S. Centers for Disease Control and Prevention grant that would have allowed the city to hire more health department staff focused on assisting the city’s ongoing efforts in preventing childhood lead poisoning. The grant would have funded city officials to test more young children for lead poisoning and collect better data that would allow them to identify the most at-risk kids.

Just months after applying, the city was denied. But the reason had nothing to do with public health. As the CDC explained, the 2010 U.S. Census counted Detroit’s population at 713,777, which was shy of the grant’s 750,000 minimum population requirement. The CDC said in a statement that it does not advance grant applications that don’t meet eligibility criteria requirements for further review.

The lost opportunity underscores the importance of having an accurate count of all people living in the United States during the constitutionally-mandated decennial Census. The count factors into how billions of federal dollars are distributed throughout the country. The number of people in your city can determine eligibility for resources needed to address lead, fix up roads, or improve schools.

Get Talk Poverty In Your Inbox

It is unclear whether Detroit’s 2010 population was undercounted by exactly 36,223 people, the number of residents by which the city fell short of the lead prevention grant’s threshold. But there is a lot of evidence that Detroit’s Census population in 2010 was less than the number of people actually living in the city, and it’s probable that it would have reached 750,000 with a more accurate count. Undercounts are typical for large cities with a large number of hard-to-count populations such as renters or immigrants.

In Detroit, only 64 percent of households responded to the Census, according to Victoria Kovari, the executive director of the city’s 2020 Census campaign. In total, about 220,000 people did not send in the forms. The Census Bureau was able to track down information about some of those households after workers spoke to residents at their doors, as well as landlords, neighbors, or even the mailman.

But, according to the Census Bureau, 26,585 people were never counted, and instead represented an estimated number of people living in uncounted units, which the federal agency calculated based on a formula that includes comparable household sizes for the specific neighborhood. It is likely that the Census Bureau was off on its estimates and that the actual number was higher.

The populations in Detroit that the Census was unable to collect any information for and forced to guess about include people living in gated communities or renters such as young people and small, low-income families living in multifamily apartment buildings, Kovari said.

Kovari said it was too tough to tell whether there was an undercount, but based on the high number of people that the Census Bureau had to make a guess about, the count was likely not accurate. “It’s clear that renters in multi-family housing were not counted,” Kovari said. “I would go as far as to say we did not get an accurate count in those areas.”

For a city like Detroit, which filed for municipal bankruptcy just six years ago, those federal funds that were denied because of a likely undercount could have been critical, said Lyke Thompson, director of Wayne State University’s Center for Urban Studies, who studies lead poisoning in Michigan.

While childhood lead poisoning in Detroit has improved in recent years, its rates still surpass those in nearby Flint. In 2016, city officials found that 8.8 percent of tested kids under the age of six were positive for lead poisoning, compared to 1.8 percent of kids in Genesee County, which encompasses Flint, according to the Detroit News. The elevated levels were higher in the city’s poorer neighborhoods, including one zip code that encompasses the Atkinson Avenue Historic District and Yates Park, in which 22 percent of 686 kids tested positive.

A lot of the city’s childhood lead poisoning problems stem from aging infrastructure that makes the water undrinkable and the city’s aging housing stock, often located in poorer neighborhoods, with lead paint-covered interior and exterior walls. Children in those neighborhoods are exposed to chippings and dust that come from the walls and breathe in exposed lead after nearby homes are demolished without following environmental remediation standards.

“$1.3 million would go a long way for [city officials] to get to the houses, to measure the blood levels in those houses and to provide case management and other services to those families. They simply lose that through this process,” Thompson said. “Detroit has some of the highest percentages of children with lead poisoning of any major city in the country so they really do need the support.”

Other cities likely experienced similar lost opportunities. The U.S. Department of Health and Human Services relies on population data when distributing nearly $3 billion each year in funding and reimbursements of five of its grant programs, including Medicaid, the Children’s Health Insurance Program, a foster care program, an adoption assistance program, and a child care and development fund program, a 2018 report from George Washington University’s Institute of Public Policy found.

Those researchers identified 37 states that may have lost out on millions of dollars in federal funding in fiscal year 2015 if their populations were undercounted by 1 percent during the 2010 Census. This includes Texas by $291.9 million, Pennsylvania by $221.7 million, Florida by $177.8 million, Ohio by $139 million, Illinois by $122.2 million, and Michigan by $94.2 million.

In most cases, it is impossible to tell which communities may have lost out on federal funds because of a Census undercount due to the fact that there are many overlapping programs with different complex funding formulas that take into account statistics beyond population size, such as the age and income of an area, according to another recent report from George Washington University’s Institute of Public Policy.

But what is clear is that undercounts do occur throughout the United States, disproportionately impacting the black population.

According to the Census Bureau’s own 2014 analysis, nearly 1 million children — 4.6 percent of all kids under the age of five in the U.S. — were not represented in the 2010 count. Children who are Latinx or black were undercounted at higher rates than white children. Such undercounts are due to children who have complex living situations, such as splitting time living between parents who do not live together, or who come from families that are considered hard-to-count, such as those who live in high-poverty neighborhoods or rental housing, according to the website FiveThirtyEight.

“The undercount of children under age five in the decennial census, and in surveys like the American Community Survey (ACS), is real and growing,” the 2014 Census Bureau report read. “This is not a new problem and has been present in decennial censuses for many decades. The differential undercount of this population across geography and demographics makes this a larger problem for some racial and ethnic groups and some parts of the country.”

It is reasonable to conclude that Detroit’s undercount was larger than the national average. The city’s population of children under five is higher than the national average and, according to research conducted by the City University of New York, several of its neighborhoods are considered among the hardest to count in the country.

In fact, the city’s population meets the very definition of hard-to-count: Areas in which less than 73 percent of its residents responded to the bureau’s first attempt to reach them.

Hard-to-count communities often include young children, racial and ethnic minorities, non-English speakers, low-income people, people who are disabled, people who are experiencing homelessness, and people who do not live in traditional housing, according to Ron Jarmin, deputy director of the U.S. Census Bureau.

Detroit has a poverty rate of 37.9 percent, 85 percent of its population are considered ethnic minorities, more than 10 percent of its population uses a language other than English at home, and 20 percent of its population is disabled, according to Census Bureau data.

To complicate matters, one in five Detroiters is evicted each year, a problem which, according to Pulitzer Prize winning author Matthew Desmond, disproportionately impacts black women, which would also lead to an undercount.

Lastly, the 2008 economic recession, which crashed the city’s economy, may have also played a part, according to Kurt Metzger, a demographer and Michigan mayor who started the local data organization, Data Driven Detroit. In 2010, city leaders, he said, were trying to address Detroit’s high unemployment rate, foreclosure crisis, and plummeting housing values as residents were underwater on mortgages and land contracts, so they were not thinking about the Census.

Metzger expected an undercount, but the end result was much worse than he anticipated, he said.

“While I have no exact undercount in mind, I was floored when I heard the 2010 count. I knew there was going to be a significant pop loss even without an undercount, but was expecting something closer to 775,000,” Metzger said in an email.

“The undercount was the reason for not qualifying for the grant. Many Detroiters had no interest in being counted and the city never worked to convince them otherwise,” he added.

The Trump administration is going to make this bad situation worse. It tried to include a citizenship question in the Census, a move that would have caused an undercount of at least 9 million people, since non-citizens and households or families with non-citizen members would fear retribution from the government if they answered. The Supreme Court ruled that the Trump administration could not include the question unless it changed its justification for adding it, which they claimed was to better enforce the Voting Rights Act.

The Trump administration shortly after dropped the question, but is still providing an inadequate supply of resources needed to ensure an accurate count. The NAACP filed a lawsuit last year against the Census Bureau and the Trump administration, claiming that their lack of preparedness for the 2020 Census violated the U.S. Constitution, since the government is required to conduct a full head count of everyone living in the country.

The civil rights organization claimed the Census Bureau was under-funded and under-prepared, hiring fewer people to knock on doors and count people that did not self-respond, and opening half the number of field offices throughout the country. Those cuts are being made while the Census Bureau rolls out, for the first time, an Internet-based survey response system.

There are widespread cybersecurity concerns related to allowing people to respond to the survey digitally, and such techniques could affect responses from communities with limited Internet access, which are often areas with a high population of people of color who are considered hard-to-count.

The Census Bureau in a statement defended its 2020 count efforts. According to the bureau, the agency is planning the most robust marketing and outreach plan in the agency’s history: It will spend $500 million on marketing, up from $376 million in 2010, advertise in “many different languages,” and is designing a “robust” outreach plan and hiring locally to engage with communities and reach hard-to-count populations.

The bureau also said that households in areas where Internet is unreliable will receive a paper questionnaire on the first mailing and all households that do not respond, regardless of the area, will receive a paper questionnaire on the fourth mailing. It added that people can respond in 12 different languages other than English over the phone or through the Internet, and enumerators will have 59 different non-English language guides among other ways of reaching out to non-English speakers.

But such threats to the accuracy of the count are real, according to Kelly Percival, a counsel at the Brennan Center for Justice’s Democracy Program.

“The 2020 Census is facing a lot of threats. A lot we have seen in past Censuses and a lot is unique for 2020,” said Percival.

“These are having a snowball effect and they could lead to an undercount in certain communities,” Percival added. “This will translate into less political power and less funding for those that need it… I think it’s an attempt to politicize the census which is not what the census is about.”

A relatively small lead prevention grant can go a long way and help a lot of children. According to Detroit officials, the 2017 grant would have enabled the city to increase the number of children under six years old who are tested for lead by 20 percent, allowed the city to collect better data so it could identify higher-risk populations, improved lead exposure outreach and education for those higher-risk populations, and better identified kids who have been exposed so they could be connected with services. It would have also provided new training for public health professionals, the lead prevention workforce, and other stakeholders who are on the front lines of the fight.

Ask the city, though, and losing out on the grant was no big deal. While, “Federal dollars will certainly assist the Department in coordinating lead related activities,” the city is doing just fine addressing the problem without it, according to city spokesperson Tamekia Nixon.

“After we didn’t receive the 2017 grant, the Detroit Health Department pursued other funding streams to allow us to provide the same scope of service intended in the grant, albeit to a somewhat lesser degree. However, at this time we are not able to quantify the exact difference in numbers,” Nixon wrote in a statement.

Last week, the city received a $9.7 million grant from the U.S. Department of Housing and Urban Development to assess 120 housing units and address lead hazards in 450 homes throughout the city for low-income families with young children, among other functions.

However, the primary function of the grant is for lead abatement, not surveillance of lead poisoning, like the CDC grant would have provided, and it will not solve the issue, said Thompson. Federal funds for such prevention efforts is crucial, he said.

“It’s really hard for the Health Department to get to even a fraction of the houses and really work with the families and they lost support to do that,” Thompson said.

Members of Detroit’s Health Department spoke to TalkPoverty on background but referred questions to the city’s communications department before going on the record. The city’s communications department gave TalkPoverty basic information about its lead program after more than a week of requests, but gave vague answers about whether losing out on the CDC funds hurt the city’s lead prevention efforts in any way. At times, Nixon told TalkPoverty to “file a FOIA” (Freedom of Information Act request) for such information.

It is unclear why the city downplayed the importance of missing out on the federal grant. However, after being denied the CDC grant, the city’s former Health Department Executive Director, Joneigh Khaldun, in a July 10, 2017 appeal of the federal agency’s decision, characterized the federal funds as a “severe need.”

“Addressing lead exposure remains a critical need given the history of Detroit as a large industrial community and the subsequent ubiquity and permeation of lead in our neighborhoods,” Khaldun said.

As American cities like Detroit scrap for federal funding to address very important issues facing their communities and their residents, an accurate count in 2020 is crucial.

Unfortunately, it’s very real, and the root cause of many  of the problems in child welfare cases.

“Some people do phrase it as a conspiracy theory,” acknowledged Richard Wexler, executive director of the National Coalition for Child Protection Reform. “When they say the government makes money on foster care, that’s not true … on foster care they still lose money, but they lose less money [than on reunification]. And private agencies do make money on foster care in many cases.”

In the United States, child welfare agencies are tasked with ensuring the health and safety of the nation’s children. Each agency receives a complex web of funding from federal, state, and local sources, leaving it accountable to a hodgepodge of authorities. Although these agencies are often referred to as “child protective services” and considered by many as a cohesive national program, state and local agencies are only linked by a loose set of federal guidelines that provide broad definitions for child maltreatment, along with the Adoption and Safe Families Act (ASFA).

First enacted in 1997 under the Clinton administration, ASFA has undergone several rewrites, but its overarching purpose has remained steady: to ensure “timely permanency planning for children.” Part of the emphasis on “permanency” includes financial reimbursements for foster care programs, as well as adoption bounties, which are lump sums in the thousands paid to states for each child they successfully adopt out after a certain threshold.

Get Talk Poverty In Your Inbox

This starts with the Federal Foster Care Program (Title IV-E of the Social Security Act), which functions as an open-ended entitlement grant. There is no upper limit to the amount of funding that can be provided for eligible foster children each year. States receive reimbursements ranging from 50 cents to approximately 76 cents for each dollar spent on daily child care and supervision, administrative costs, training, recruitment, and data collection.

But when it comes to programs that support family reunification, the budget slims. Title IV-B of the Social Security Act, which governs federal reunification funding, includes a capped entitlement component and a discretionary component. So, unlike foster care funding, these dollars come with a set limit.

And that limited money isn’t all for reunification services. Title IV-B also includes provisions that allow for some of this funding to go toward foster care programs. A portion is also required to go toward adoption promotion.

The result of this imbalanced funding structure? The federal government spends almost 10 times more on foster care and adoption than on programs geared toward reunification.

One of the less-known sources of federal funding for child welfare programs is the Temporary Assistance for Needy Families (TANF) program. TANF is supposed to be a cash-assistance program servicing low-income families with children, In reality, TANF funds can be used to support many services designed to help “needy” children, including child protection agencies. The result is that many states use TANF funds to finance foster care, child welfare investigations, and adoption or guardianship payments.

Because child welfare program data are self reported, it can be difficult to track exactly how each dollar is spent, but Wexler was able to identify eight states using TANF to pay for adoption subsidies, 23 states funding CPS investigations, 27 states funding foster care, and three states diverting TANF money to fund residential treatment facilities for child welfare involved children.

Considering that three-quarters of substantiated child maltreatment cases are related to neglect, which is often the result of poverty, it seems exceedingly unjust that funds supposedly intended to offset the worst effects of poverty are instead being used to finance the separation of mostly poor families.

Under ASFA, states are — with few exceptions — required to file for the termination of parental rights when a child has been in foster care for 15 of the past 22 months. In an attempt to curtail the infamous foster care hopscotch, which leaves children whose parents have lost their rights bouncing from foster home to foster home, the government created adoption payment incentives.

Adoption bounties range from $4,000 to $12,000 per child. As Wexler explained, “the harder the system deems the child to place, the higher the bounty.”

But in order to begin collecting that money, a state must exceed the last year’s number of adopted children, thus incentivizing states to permanently re-home an ever-increasing number of children each year. As can be expected, the number of adoptions increased in the five years after the implementation of ASFA, while reunifications declined. The Bush administration’s Adoption and Promotion Act of 2003 further codified this adoption bounty system by allocating $43 million yearly to states that succeed in increasing the number of adoptions from foster care.

Many states contract with private agencies that oversee out-of-home placements and service referrals for child welfare involved children. Said Wexler, “that agency will probably be paid for each day that child remains in foster care … So the private agency has an incentive to convince itself that the child really, really can’t go home and has to stay with them for a long, long time.”

What does this look like on the ground? Painfully delayed referrals to support services such as parenting classes and addiction treatment, judges hesitant to find fault with the way agencies and providers handle cases, and private agencies eager to deem parents unfit for reunification.

There have been some recent moves at the federal level aimed toward shifting some of these financial imbalances. The Family First Act, signed into law in 2018, now allows federal reimbursements for mental health services, evidence-based substance use treatment, and in-home parenting support. Its purpose is to create similar incentives for helping families stay together.

Unfortunately, the act does not support many of the common needs that lead to family separation, such as housing or child care support. And because the programs it does support must meet stringent requirements in order to be eligible for reimbursement, foster care and adoption subsidies continue to exceed reunification programs by the hundreds of millions.

But after decades of collecting this data, the government is putting it to use. This information is feeding algorithms that decide everything from whether or not you get health insurance to how much time you spend in jail. Increasingly, it is helping determine whether or not parents get to keep their kids.

Get TalkPoverty In Your Inbox

When someone phones in a report of suspected child abuse — usually to a state or county child abuse hotline — a call screener has to determine whether the accusation merits an actual investigation. Sometimes they have background information, such as prior child welfare reports, to assist in their decision-making process, but often they have to make snap determinations with very little guidance besides the details of the immediate report. There are more than 7 million maltreatment reports each year, and caseworkers get overwhelmed and burn out quickly — especially when a serious case gets overlooked. New algorithms popping up around the country review data points available for each case and suggest whether or not an investigation should be opened, in an attempt to offset some of the individual responsibility placed on case workers.

The trouble is, algorithms aren’t designed to find new information that humans miss — they’re designed to use the data that humans have previously input as efficiently as possible.

“If you give it biased data, it will be biased,” explained Cathy O’Neil, mathematician and author of the book Weapons of Math Destruction, while speaking with me for a story I wrote for Undark last year. “The very short version is that when you’re using the past as a kind of reference for how it works well, you’re implicitly assuming the past is doing a good job of rewarding good things and punishing bad things. You’re training the system to say if it worked in the past, it should work in the future.”

Historically, low-income families have had their children removed from their homes at higher rates than wealthier families. As a result, these new algorithms work to codify poverty as a criteria for child maltreatment. Some of the variables that these tools consider are public records that only exist for low-income parents, such as parents’ poverty status, whether they receive welfare benefits like SNAP and TANF, employment status, and whether they receive Medicaid. Other factors, like previous criminal justice involvement and whether or not there have been allegations of substance misuse in the past, are also dramatically more likely for families living in poverty.

This bias exists even in systems that have been highly praised, like the Allegheny Family Screening Tool currently being implemented in Pittsburgh, where prior arrests and parents’ mental health histories are considered factors in whether a child should be removed. It’s similar in other, less-transparent systems, like one in Florida where tech giant SAS contracted with the Florida Department of Children and Families to research which factors were most likely to contribute to the death of a child by maltreatment. According to press releases by SAS (some of which have been unpublished since they began garnering media attention) the company used public records such as Medicaid status, criminal justice history, and substance-use treatment history.

The results led jurisdictions in Florida to zoom in on factors that apply to huge swaths of families, including mine. In April of last year, an allegation of drug use and child abandonment led Broward County, Florida child welfare investigators to investigate my family. When my drug tests were negative, the investigation pivoted to my recent financial setbacks, which had been caused by my husband’s acute health crisis. My children were ultimately removed from my care, and we have been separated for nine months for reasons that are primarily financial. My case is far from unique. Three-quarters of child protective cases in the United States are related to neglect, not abuse, and that neglect usually means lack of food, clothing, shelter, heating, or supervision: problems which are almost always the result of poverty.

Ira Schwartz, a private analytics consultant, thinks he may have found a way to help re-balance this system. He conducted a research study in Broward County — the same county in which my case is based — that discovered the current approach to child welfare substantiation is highly flawed. According to his research, 90 percent of system referrals were essentially useless, and 40 percent of court-involved cases (which typically involve child removal) were overzealous and harmful, rather than beneficial, to the families. He created his own system that, like the Allegheny tool, predicted the likelihood that a family would become re-involved with the system. But he admits quite openly that predictive algorithms like his target the poor.

“We found in our study that lower socioeconomic status was one of the significant variables that was a predictor [for reinvolvement with the system],” said Schwartz. “The issue with higher-income families is … they just don’t really come into the system because they have other options. With higher-income families, when there’s child abuse or neglect or even spouse abuse and it’s reported, they can afford to go to private agencies, get private mental health services; they can see a psychiatrist or social worker or psychologist … it’s a discrimination factor.”

Schwartz believes that these types of admittedly discriminatory computer programs can still be put to good use when combined with prescriptive analytics, which would determine the services that high-risk families need in order to remain out of the system in the future. Schwartz says this would include services like rental assistance, food assistance, day care funding, and housekeeping services. This would help welfare agencies understand which families need what services, and streamline the process of providing them. (All jurisdictions are legally required to make “reasonable efforts” to help families resolve the issues that brought them under investigation, but how agencies go about meeting that standard varies widely by location.)

The issue with these algorithms is certainly not malice on the part of their creators. Even the more secretive, proprietary algorithms being created by companies like SAS claim to want to create a safer system that results in less child maltreatment. But it’s unclear if that is possible with the data that’s available. Without comparable data from wealthier populations, which are better protected by privacy laws, the new systems cannot produce accurate results — and even if more data were added, it would mean more families are being separated and surveilled.

Despite public statements to the contrary, there is mounting evidence that the administration is continuing to separate asylum-seeking families like Brayan and Julio’s. President Donald Trump holds fast to the belief that family separation effectively deters families from Mexico and Central America from seeking refuge in the United States, despite evidence to the contrary, and immigration attorneys are reporting that the administration is taking advantage of a loophole in the federal court’s injunction against separations. According to Neha Desai at the National Center for Youth Law, border patrol officers are using the pretext that children’s safety is at risk to separate families: “If the authorities have even the most specious evidence that a parent was a gang member… anything they can come up with to say that the separation is for the health and welfare of the child, then they’ll separate them.”

The Trump administration’s decision to systematically separate children from their parents, is, in its specifics, unprecedented. But family separation was enabled in the first place, and it continues today, because our immigration system, like other public systems, has been built to separate families — particularly families of color.

Get TalkPoverty In Your Inbox

The immigration system is one of three systems that routinely separate families in the United States. The criminal justice and child welfare systems are the other two. In the immigration and criminal justice systems, separation is most commonly an unconsidered, if not quite unintended, consequence of policy, as parents are incarcerated and sometimes deported without their children. In child welfare, separation is the deliberate result of policy, as children are removed from their parents’ custody over concerns for their immediate safety. In each system, however, children are harmed by family separation. And in each system, children of color are more likely to be separated from their parents.

The very first federal restriction on immigration resulted in family separations. In 1875, Congress barred Chinese involuntary laborers and suspected prostitutes from entering the United States. In practice, the law made it almost impossible for Chinese women to immigrate, including those who wished to join their husbands, as government officials “demonstrated a consistent unwillingness, or inability, to recognize women who were not prostitutes among all but the wealthy applicants for immigration.” In the years that followed, an increasing number of laws excluded more Asians from the United States. Separations continued as part of this: At Angel Island, the notorious immigration station in San Francisco Bay, many Asian-American families were separated for weeks at a time so that they could not coordinate their answers before they faced interrogation.

By the mid-20th century, Latinx immigrants had become the subject of nativist ire, and many Latinx families were separated as a result. During the Great Depression, local and state governments colluded with social welfare agencies to encourage and sometimes coerce Mexicans—and in many cases Mexican Americans — to “repatriate” to Mexico. Two decades later, concern about rising undocumented immigration in the Southwest led to “Operation Wetback,” a federal deportation drive that was once again focused almost exclusively on Mexicans. The legacy of this targeting of Latinx communities by immigration enforcement is visible today. Though immigrants from Latin America make up an estimated 77 percent of the unauthorized population in the United States, they have constituted well over 90 percent of immigrants removed by U.S. Immigration and Customs Enforcement (ICE) in recent years. 27,080 immigrants with U.S. born children were deported in 2017.

Like immigration enforcement, our system of mass incarceration mechanically separates families. Incarceration creates financial and emotional hardship for families by default, but there are additional ripple effects that can last long after release. According to an analysis of 3 million child welfare cases, parents who have a child placed in foster care because they are incarcerated are more likely to have their parental rights terminated than those who physically or sexually assaulted their kids. Again, this falls disproportionately on children of color: Approximately 11.4 percent of African-American children have a parent in prison, compared to 3.5 percent of Hispanic children and 1.8 percent of white children. This disparate impact has been true for the history of the criminal justice system in the United States, and it has grown with the rise of mass incarceration since the 1970s.

Families of color are also disproportionately separated by the child welfare system, which from the beginning saw its role as removing children from their families for their own protection. Originally, the child welfare system focused on removing poor children from their families, whether or not there were signs of abuse. As William Pryor Letchworth, a famous advocate of children’s causes, declared in 1874, “If you want to break up pauperism, you must transplant [the child].” Charities in New York, Boston, and other East Coast cities sent thousands of poor children on “orphan trains” to towns in the Midwest, where they were assigned foster families.

As the child welfare system developed in the late 19th and early 20th centuries, children of color were for the most part excluded from services, but other public institutions separated them from their families at high rates. A Children’s Bureau report observed that from 1750 to 1960, “the black child’s chance of ‘receiving care’ [a polite euphemism for being incarcerated] from a correctional facility was still much greater than that of receiving any other type of care.” Meanwhile, the United States undertook a concerted campaign to remove American-Indian children from their families in order to facilitate their “assimilation.” Starting in 1879 and continuing well through the 20th century, children as young as five years old were packed off to boarding schools, where they were prohibited from speaking their native languages and, often, from visiting home.

When the formal child welfare system began to integrate following World War II, it continued to identify symptoms of poverty as grounds for removing children, and separated American-Indian and African-American families at startlingly high rates. Starting in 1959, the Indian Adoption Project, part of the Bureau of Indian Affairs’ (BIA) larger effort to undermine tribal sovereignty and erase American Indian cultures, purposefully placed American Indians in white homes. Surveys in 1969 and 1974 documented that between 25 and 35 percent of all American-Indian children were placed in foster or adoptive homes or institutions. During this period, child welfare scholars also began to document the high rates of removal of African-American children, a legacy that lives on despite attempts to address racial inequities. A 2014 study found that 4.9 percent of white children will experience foster care placement before their 18th birthday, compared to 15.4 percent of Native American children and 11 percent of black children.

This history reveals that Julio and Barayan are not alone, even under less openly racist administrations. Thousands of families are separated every year by public systems, and families of color are much more likely to suffer this fate. In order to ensure that families like Julio and Barayan can remain together, we need to transform these systems. In the criminal justice and immigration systems, this means severely limiting incarceration and deportation, particularly of parents. In the child welfare system, this means increasing the services and supports available to families so that they can thrive together, as well as significantly raising the threshold to remove children from their homes. Children need their families in order to develop and flourish. As a nation, we cannot continue to tear children from their parent’s arms.