How do American child welfare and obstetric healthcare converge? Matty Lichtenstein, a recent PhD from UC Berkeley’s Department of Sociology, studies how state and professional organizations shape social and health inequalities in maternal and child welfare. Her current book project focuses on evolving conceptions of risk in social work and medicine, illustrated by a study of the intertwined development of American child and perinatal protective policies. She is working on several collaborations related to this theme, including studies of maltreatment-related fatality rates, the racialization of medical reporting of substance-exposed infants, and risk assessment in child welfare.
In another stream of research, she has written on social policy change, with a focus on educational regulation and political advocacy, and she has conducted research on culture, religion, and politics. Dr. Lichtenstein’s work has been published in American Journal of Sociology, Qualitative Methods, and Sociological Methods and Research. She is currently a postdoctoral research associate at the Watson Institute for International and Public Affairs at Brown University.
In this podcast episode, Matrix content curator Julia Sizek speaks with Lichtenstein about her research on the transformation of American child welfare — and the impact of that transformation on contemporary maternal and infant health practices.
Excerpts from the interview are included below (edited for length and content).
How has the child welfare system changed over the span of time that you study?
I focused my research starting after the passage of the Social Security Act, because that is the major dividing line for American child welfare. Prior to 1935, when the Social Security Act was passed, we had a fragmented patchwork of mostly private child welfare agencies throughout the United States. The passage of the Social Security Act enabled an expansion of funding for state and local public child welfare. The main shift had to do with thinking about what welfare meant, and what it still means today.
In general, when we think about welfare, we are referring to government support for individuals or groups. The main distinction, especially in the 1930s, was between financial support — giving people money when they needed it, and couldn’t get it any other way — or providing services, such as funded medical services, educational services, or psychological counseling. Across social work, which was in a way the parent discipline of child welfare, there was a tension there. How do we help people — by giving them financial aid, or do we help them through social services?
The Social Security Act made that distinction quite clear for child welfare services, because the section that focused on child welfare services emphasized that this was about services in general, and financial aid was a separate part of the Social Security Act for families. One of the things that needed to be figured out was, what is child welfare, and how do you best serve children?
I’ve found in my research that there was an increased emphasis in the 1930s and 40s on the argument that child welfare should serve all the various needs children have. It was not just poverty-related needs. In fact, they veered away from poverty-related needs toward psychological needs, medical needs, health needs, etc. Child welfare advocates pushed for more funding and more resources for child welfare. What happened is that public child welfare grew exponentially in the 1950s and 1960s. The number of child welfare workers started rising dramatically. This led to a larger shift in child welfare and thinking about what child welfare meant in the 60s and 70s.
What was the focus of the child welfare system in the 1960s and 70s?
One of the major findings of my dissertation conflicts with the conventional narrative of child welfare history. The classic narrative is that the late 50s and 60s saw the discovery of child abuse as a social problem. Before then, scholars argue, nobody was talking about child abuse and neglect, and social workers and the public did not see it as a problem. And then by the 60s, it became a public and political issue, and you saw a number of laws being passed to mandate reporting of child abuse. This led to the creation of child welfare as we know it today, which is heavily focused on child abuse prevention and response.
The problem was that, as I dug through more archival resources, I found that that just wasn’t the case. The most damning piece of evidence I found was a publicly available report put out by the Children’s Bureau in 1959, which stated that 49% of public child welfare in home services related to abuse and neglect. This was in 1959, when current scholars were saying nobody talked about abuse and neglect.
I spent a few months in a sort of existential crisis: what is the meaning of my dissertation if everything is wrong? Eventually, I figured out that not everything is wrong, and that a lot of what was written about the history of child welfare was correct. There was much more of an emphasis on child abuse. But what it missed was this larger moment of transformation in child welfare.
What I show is that it’s not so much that child welfare agencies rediscovered child abuse, as much as they relinquished (sometimes willingly and sometimes unwillingly) jurisdiction over most other child welfare issues, including poverty, health issues, and education, and they retained jurisdiction only over child abuse and child neglect. I show that this happened largely due to larger trends in the American welfare state, specifically welfare state retraction and an increasing focus on efficiency and welfare governance in the late 60s and 1970s, which demanded that child welfare focus on issues that could be easily defined and services that you could put a price on.
The Children’s Bureau could no longer say they serve all of the needs of the population of children. Instead, there was an increasing shift toward, what is the problem you’re here to resolve? There were advocates that pushed for more focus, but it was all part of this larger shift in the American welfare state.
I also emphasize that the massive expansion of child welfare — that growth of staffing and funding — was also made possible by laws saying, you need to report child abuse. Where do you report it? To a child welfare agency. So now there were thousands of child welfare workers. It had unintended consequences. All the child welfare workers who were supposed to solve all of children’s problems were now there to solve one problem, which was the increasing the number of reports of child abuse and neglect.
How was the category of child abuse and neglect defined, and how did it transform over time?
Early research that tried to define what it meant to have abusive parents was primarily in medical journals. That was usually based on things like X-rays of children with broken bones and trying to figure out, was this an accident, or who caused this? There were also psychiatric evaluations of parents saying, what is wrong with parents who do this? It was a diagnostic model of approaching child abuse and neglect. The cases they were referring to were usually fairly severe cases of child abuse and neglect.
Originally, a lot of the laws addressed medical professionals, but they quickly expanded, in part because medical professionals pushed back and said, we can’t be the only ones mandated to report this. And so it quickly started to expand throughout the 1960s and 1970s to include professionals across the board who have any sort of interaction with children, including anyone in an educational setting, anyone in a medical setting, or people who work in funeral homes, for example. They became mandated reporters, which means they were supposed to be penalized if they did not report what were often very vaguely defined forms of abuse and neglect.
This varied greatly across states. Every state had different laws and different sets of mandated reporters, but child welfare agencies across the country started to receive a skyrocketing number of reports. This does not mean that everyone was reporting every suspicion, but there were enough reports pouring into child welfare that they had to figure out what to do with all these reports. In the 1970s, and increasingly in the 1980s, that forced a reckoning of the question of how to define child abuse — and how to figure out if what’s happening is child abuse and neglect.
Out of these millions of reports that started pouring in during this era, the majority were usually unsubstantiated. In the mid-1970s, usually around 60% of reports were unsubstantiated. The majority of reports that are substantiated were neglect reports that were highly correlated with poverty. There were eight times the rate of substantiated reports of physical neglect among low socioeconomic-status children versus non-low socioeconomic status children. So you had a broad category of neglect, which could include everything from passively allowing your child to starve to leaving your child home alone for a few hours when you go out to work. There was a huge range that varied by county and state.
The question then became, if you have this huge number of reports coming in, and the majority of them are not even abuse and neglect, or it’s not clear if it’s neglect or poverty, how do you create a system to prevent and treat a problem that we’re not even sure exists? And that’s really where you started to see this focus on risk. Child welfare and medical professionals affiliated with child welfare began to develop practical risk assessment tools to determine the risk that there’s an actual case of child abuse happening, or that it might happen in the future. These tools had all sorts of problems built into them.
What was wrong about the risk assessment tools that professionals were using?
In the 70s and 80s, the tools were often built on what was called a consensus approach to risk assessment. That was based on what social workers considered risk variables. This was deemed very problematic by the 1990s, but they were still widely used for the first 20 or so years. These tools tended to incorporate all kinds of variables having to do with the environment of the child. There may not have been any sign that the child was harmed directly, but you look at the environment and try to assess if there are risk variables there. That had to do with everything from the income status of the family to health issues of the parents to the marital status of the mother.
Childcare access could be a risk factor, as well as issues like the stability of the home. In the 1970s, there were risk assessment tools that had factors like, do the parents take this child to movies? Do they have a camera? Do they take the child fishing? Does the child have a mattress? You can see that it’s really hard to disentangle poverty from this.
There were also sometimes cultural factors. There was an early tool that was approved by the predecessor to the Department of Health and Human Services that asked whether the parent had wider family support in child care, and whether they were overly dependent on their family. That gets at something that is cultural, not just economic: studies have found that in families of color, there’s more interdependence and less of an emphasis on nuclear family units, so this could be problematic.
Drug or alcohol use was assessed as a risk factor. When you look at earlier surveys about child welfare services before this transformation toward a focus on child abuse, they would talk about health and family issues as issues of child welfare, but they weren’t risk factors for abuse. Child welfare might intervene if there was some sort of health issue with a parent, but that was seen as distinct, whereas when you look at the studies in the 1970s and 1980s, those same factors were not just a health issue, but a risk factor for abuse or neglect. So you saw a trend of structural inequalities and health issues turning into risk factors.
So instead of trying to say, how do we help this family as a whole, it became, how do we assess the assumption that the parent is harming the child? It’s an approach in which parent and child are seen as distinct units, and the question is, are they in some sort of conflict? What’s interesting is that this is a relatively rare problem, in which there’s an intentional effort by the parents to harm the child. It certainly happens, but it’s relatively rare.
How does what you’ve learned matter for people thinking about child welfare policy today?
First, child welfare is under-equipped for multi-dimensional problems. In some states, they might have access to more resources, and in other states, the only thing that can really do is child removal or interventions that are often quite disruptive to the family. Having child welfare in charge conflicts with the multidisciplinary approach that’s favored by most professionals.
Second, child welfare is associated with an enormous amount of trauma, especially for families that are low-income and for families of color in the United States. Fifty percent of African-American children in the United States today have experienced a child welfare investigation — one out of two. That’s just crazy. Huge numbers of children are experiencing these kinds of investigations. Perhaps some are very minimal, but some of them are not going to be so minimal.
What we have is potentially traumatic family surveillance and separation that’s intrinsically linked to child welfare, because no matter how helpful or well-meaning a child welfare worker might be, ultimately child welfare has the authority to take your child away, possibly forever. Even if they do that rarely, it can still be something that is laden with fear and anxiety for families.
Adding to that, lower standards of evidence are applied in child welfare proceedings, so that makes it particularly problematic to have child welfare involved in cases of substance-exposed infants, especially because (at least based on the limited data we have, for example, for California), a significant percentage of these infants are taken away from their mothers. Taking a newborn away from their mother is not necessarily an evidence-based approach to dealing with substance use issues. But the paradigm of child welfare is not necessarily to approach the best interests of the family as a whole. The paradigm of child welfare is to reduce and mitigate risk of future child abuse and neglect.
There have been significant shifts in child welfare over time. My research largely ends in about 2000. In the first couple of decades of the 21st century, there has been a concerted effort by child welfare agencies on every level to try to counter some of the intense racialization and income inequality that is reproduced by the child welfare system. We’ve seen a dramatic decline in child removals. For example, in New York City in 1995, there were 50,000 children in foster care. In 2018, there were 8,000 children in foster care. That is a dramatic decline. However, even though there were 8,000 children, there have been an enormous number of children investigated, and in New York City in 2019, 45,000 cases were in preventative services. So you still have a lot of child welfare involvement. What that means for families is not really clear yet.
The second major shift is that there’s been an intensification of the focus on risk assessment. We have seen the development of quite sophisticated risk assessment tools, not just the consensus tools, but actuarial tools and algorithmic tools that use computational methods to assess risk. And there have been a lot of critiques of some of these tools. The main issue is, do these tools funnel multiple problems, many of them poverty-related, into child welfare? And even if racial disproportionality in some states has declined, we still have a lot of racial disproportionality in child welfare, and income inequality continues. We don’t have enough data on that to fully assess it. And so we’ve continued to have significant issues with child welfare today, even as it has changed in this new century.
What are the approaches that different states take to the issue of infants who have been exposed to substance use during pregnancy?
In the 1980s, you have an increasing number of reports coming into child welfare of substance use during pregnancy, and a lot of this was highly racialized, in terms of how it was conceptualized. During the 1980s, this problem received a lot of media coverage. And what that means is that state legislators felt they had to do something; they had to respond in some way. And their options were basically to say, well, we can mandate medical intervention in such cases, we can criminalize these women for harming their children and mandate essentially law enforcement interventions, or we can mandate civil interventions through child welfare. The current scholarship on this period — and really on this issue — tends to focus a lot on criminalization, on how pregnant women are thrown into jail and how women are jailed or prosecuted for these kinds of uses. And then there’s also a lot of conflation of child welfare interventions and medical interventions, all part of this larger criminalization and policing of pregnant women. And there’s a lot to be said for that framework. But I think it’s actually really important to distinguish between those things, because criminalization is actually relatively rare compared to the thousands of women who are reported in each state to child welfare every year. By far the predominant response is child welfare reporting.
So how do we essentially manage and mitigate this risk of substance-exposed infants? Child welfare has this risk prevention framing, and also, it’s supposed to be dedicated to protecting children. So they are the perfect response. And what’s interesting about this is that child welfare increasingly across states becomes the primary authority for intervening in such cases, even as simultaneously, the professional consensus increasingly converges on the idea that we need a multidisciplinary response to the issue of substance-exposed infants. If you’ve read reports that are put out on this issue of substance-exposed infants, including from the federal government, the consensus is that we need doctors and social workers and financial aid, and perhaps even law enforcement. Everyone needs to work together to deal with this issue of substance-exposed infants. But in practice, the state laws overwhelmingly favor child welfare interventions, and child welfare is mandated to mitigate risk of child abuse and neglect. They’re not there to provide a multidisciplinary approach. They can and sometimes they do; it varies greatly by state. But that’s not their primary mandate. And there are very concrete consequences to having a child welfare response to this issue.