Listen to Part 1 of the interview
Listen to Part 2 of the interview
Listen to Part 3 of the interview
Pablo Argote, Jennie E. Brand, Stefanie DeLuca, Nazar Khalid, Kristin L. Perkins, Alexa Prettyman, Emily Rauscher, and Florencia Torche are contributors to RSF: The Russell Sage Foundation Journal of the Social Sciences issue “Disparate Effects of Disruptive Events on Children,” edited by Florencia Torche (Princeton University), Jason Fletcher (University of Wisconsin-Madison), and Jennie E. Brand (University of California, Los Angeles). The issue offers a systematic examination of the variation in the consequences of disruption in early life. The interview has been edited for length and clarity.
Pablo Argote is a postdoctoral scholar at the University of Southern California. He is a co-author of the article “Politics Matter: How Political Experience Mitigates Learning Loss Caused by Natural Disasters.”
Jennie E. Brand is a professor of sociology and statistics and data science at the University of California, Los Angeles. She is a co-editor of this issue of RSF and a co-author of the article “Disparate Effects of Disruptive Events on Children.” She is also the author of the RSF book Overcoming the Odds: The Benefits of Completing College for Unlikely Graduates.
Stefanie DeLuca is the James Coleman Professor of Social Policy and Sociology at Johns Hopkins University. She is a co-author of the article “Exploring the Trade-Off Between Surviving and Thriving: Heterogeneous Responses to Adversity and Disruptive Events Among Disadvantaged Black Youth.” She is a co-author of the RSF book Coming of Age in the Other America, a contributor to RSF journal issue “Administrative Burdens and Inequality in Policy Implementation” and RSF volume Choosing Homes, Choosing Schools, and is the recipient of multiple RSF research grants.
Nazar Khalid is a doctoral candidate in demography and sociology at the University of Pennsylvania. He is co-author of the article “Floods and Children’s Education in Rural India.”
Kristin L. Perkins is an assistant professor of sociology at Georgetown University. She is the author of the article “Heterogeneous Household Change Among Children.” She is also a contributor to the RSF journal issue “Severe Deprivation in America.”
Alexa Prettyman is an assistant professor of economics at Towson University. She is a co-author of the article “The Effects of the Great Depression on Children’s Intergenerational Mobility.”
Emily Rauscher is a professor of sociology at Brown University. She is a co-author of the article “Unequal Effects of Wildfire Exposure on Infant Health by Maternal Education, 1995-2020.” She is also a contribute to RSF journal issue “Administrative Burdens and Inequality in Policy Implementation” and an RSF research grant recipient.
Florencia Torche is the Edwards S. Sanford Professor of Sociology and International Affairs at Princeton University. She is a co-editor of this issue of RSF and a co-author of the article “Disparate Effects of Disruptive Events on Children.” She is also an RSF research grant recipient.
Q. What are some examples of disruptive events? Why is it important to study the impact of disruptive events on children? And why is it important to study how the impacts vary among different groups?
Torche: So, what are disruptive events? We identify unexpected events that have negative consequences for people’s lives in several domains – in the economic domain, in the household and family domain, in terms of health, and in terms of environmental exposures. So, let me give you an example in each one of those domains. In terms of economic events, we're thinking about things such as an economic recession or economic downturn at the macro level. And at the micro level – or at the level of the families or individuals – we are thinking about something like losing a job. In terms of disruptive events that may happen at the household or family level, we're thinking about exposures such as parental divorce or the incarceration of a parent or a family member. In terms of health-related events, we're thinking about things such as illness of a family member or the death, say, of a parent. And in terms of environmental exposures, we're thinking about the unfortunate increasingly common events related to climate change, such as floods or hurricanes or other macro level disasters that shape entire communities. Jennie, do you want to build on that?
Brand: That’s a very good characterization. We think of these macro- and micro-level events of interacting. So, during some macro-economic disruption, like an economic recession, there are more micro-level events, like job losses, and similarly, across these other domains. Those different events shape the consequences for different groups.
So, why is it important to study the impact of disruptive events on children? Generally, disruptive events can have very broad and far-reaching consequences for individuals and families. Long-term trajectories can be altered by disruptions, especially for children. Children are particularly vulnerable to disruptive events because shocks that happen in early life can alter developmental trajectories and then result in long-term consequences for educational attainment, socioeconomic attainment, and overall health and wellbeing. So, things that happen in early life can really condition how children’s lives unfold over the long term. We think it’s particularly important to consider how these types of disruptions impact the wellbeing of children. Did you have anything you wanted to add, Florencia?
Torche: That was great, you said it all. We emphasize early life because we now know, based on research from many different fields—including the biomedical field and developmental psychology—that early life exposures to disruptive events may have long-lasting consequences and even intergenerational consequences. I think it is worthwhile to mention that consequences are not only for the people who are exposed when they're young, but also for their children. The event may contribute to the intergenerational persistence of advantage or disadvantage.
But the question of why to study variation is central. The focus of this volume is not only to try to understand the impact of these sources of disruption on children and their families, but especially to understand the extent to which there's a variation in this impact and the explanations for that variation. The reason for that is there is substantial evidence suggesting that the same event can have very different consequences for different people or different groups. The event may be quite harmful for some groups, it may be neutral for other groups, and it may even provide a positive turning point for other groups. So, this idea of examining the overall effect of disruption on children is very important, but it is incomplete if there is indeed substantial variation in the population.
We really want to understand which groups are more vulnerable and more affected not only for scientific reasons, but also if we want to inform decision-making. That is the reason why this is our focus. And there is good theoretical rationale to expect variation across the population. Jennie, would you like to outline our different theoretical expectations?
Brand: We know that the risk of experiencing different disruptive events is often stratified by socioeconomic factors. So, things like family instability or economic instability, and disruptions that occur within those domains, are more likely to occur among those with fewer socioeconomic resources. And some theoretical expectations are that those with fewer economic resources also have more trouble responding to those types of disruptions. So, disruptions may compound the disadvantages that some groups face. That's one theory about how disruption may vary across the population. But then there's another theory that suggests that people who are unlikely to experience a disruptive event may be more shocked when those events occur. So, children who may otherwise have more privileged circumstances in childhood may experience shocks that disrupt their trajectory in unexpected ways.
Throughout this volume, we examined these two theoretical propositions and how different disruptions play out for children in different ways – whether we see this compounding of socioeconomic disadvantage or people more negatively affected because it was a greater psychological shock and disruption to an expectation of a more advantaged trajectory.
Q. What were the effects of the Great Depression on the upward mobility of children of the Great Depression? How did these effects vary by group?
Prettyman: As many people are likely aware, the Great Depression was the deepest and most protracted downturn in U.S. history. Such an economic crisis could either increase or decrease upward mobility. Before I explain how, let me first explain what is meant by upward mobility. Upward mobility measures whether an individual’s outcomes surpass their parents’ outcomes. In other words, do children do better than their parents?
Now how could the Great Depression have a theoretically ambiguous impact on upward mobility? Well, on the one hand, the Great Depression may have hit poorer families harder making it more difficult for them to invest in their children. So, we would expect to observe a decline in upward mobility. On the other hand, the Great Depression may have taught grit and resilience to the families hit harder, so we would expect to observe an increase in upward mobility. The Great Depression may have also “leveled the playing field” for children in different social classes if wealthier families had more to lose. This would break the link between parent and child outcomes and increase relative mobility, at the expense of declines in upward mobility among the rich or most advantaged.
My co-authors, Martha Bailey, Peter Lin, and Shariq Mohammed, and I use historical data that links children from the Great Depression to their parents and their locations to understand how the severity of the Great Depression impacted upward mobility in terms of occupation and educational attainment. We find that the Great Depression increased upward mobility for teenage boys but reduced upward mobility for teenage girls.
These results are specific to education, meaning, teenage boys living in places more negatively impacted by the Great Depression were more likely to get more education than their fathers, relative to their peers living in places less negatively impacted by the Great Depression. Perhaps sons from more negatively impacted households learned how to work hard and hustle, whereas sons from less negatively impacted households saw forgoing education to participate in the labor market as an agreeable trade-off. However, these educational differences did not necessarily translate to occupational gains.
For teenage girls, the opposite is true. Perhaps because teenage girls had more to lose from Depression-era school closings. This finding is especially evident among daughters with more siblings, presumably because daughters in larger families had to drop out of school to help care for their siblings.
In summary, we find differences in upward mobility among teenage boys and girls, suggesting that disruptive events for individuals on the cusp of completing their education or entering the labor market have consequential impacts on their lives.
Q. What factors make a child more or less likely to experience changes in their household composition? How does the likelihood of experiencing family instability impact Black children’s educational attainment? Do the effects vary by the type of family member moving in and out of the house?
Perkins: Other studies show that changes within a child’s family can prompt changes in household composition, in particular parent divorce often means a parent moves out, job loss, health shocks, and death in a child’s network may mean household members leave or join. Living in a shared or doubled-up household predicts experiencing changes since shared arrangements are often unstable. In prior work, Matt Desmond and I find that residential mobility, or moving, is often accompanied by changes in household composition.
Looking at the propensity score models I used to predict household change for this article, we see that having an employed head of household and living in an owned home reduce the likelihood of experiencing household change involving a parent, all else equal. If the child’s parent is not the head of household, if they live with a female head of household, or have more people in the household the child is more likely to experience change involving a nonparent. Having married parents, having an older sibling, a parent head of household, a more educated head of household, living in an owned home, and having a higher household income all reduce the likelihood of experiencing both parent and nonparent change. Essentially, the more resources the child’s household has, in terms of education, income, or assets (like owning a home), the less likely they are to experience changes in household composition.
When examining how the likelihood of Black children experiencing family instability impacts educational attainment, we should first understand that most Black children in my sample experienced household changes during childhood. Only 13% experienced no changes compared to 30% who experienced changes involving nonparents and 45% who experienced changes involving both parents and nonparents. So, experiencing household change is very common.
But there were some differences in the effects of household change on educational attainment. In general, across types of household change and educational outcomes, I find that the negative effects of changes appear to be strongest among Black children who were relatively unlikely to experience change. That is, the negative effects of household change are strongest among the group of children whom we would not expect to experience change based on their family and household characteristics.
My conclusion that changes in household composition negatively affect educational attainment among some Black children runs counter to research concluding family instability is less consequential for Black compared to White children and I think it provides support for the call other researchers have made to examine within-group variation in family processes.
I frequently get asked if the effects vary by the type of family member moving in and out of the house. In some ways it inspired my whole research agenda on changes in household composition. There is a huge amount of scholarship focused on the effects of parental divorce and new romantic relationships, but much less focus on nonparents who leave and join children’s households. There is not, however, a straightforward way to estimate effects by type of family member since children rarely experience only one type of change during childhood. For most children I cannot isolate one type of change because they experience multiple changes during childhood. If I look at children in my sample who had a parent leave at some point during childhood, more than half of them also had a parent join, and about a third had a parent leave more than once. If I look at children who had an extended family member or nonrelative join their household, 90 percent also experienced an extended family member or nonrelative leave, and 65 percent experienced this type of change multiple times. I provide more statistics like this in the article, but to summarize, these changes do not typically happen in isolation, and trying to model them independently would require I ignore the complexity that characterizes most children’s households and lived experiences.
In an article I published in Demography, I show that the association between household changes involving parents and teen childbearing is statistically indistinguishable from the association between changes involving nonparents and teen childbearing, suggesting that in that context, predicting childbearing instead of education, household composition shifts involving nonparents can be as disruptive as those involving parents.
Now, I’m not saying that a grandparent or uncle can necessarily replace a parent in a child’s life. Instead, what I’m finding is that changes involving different types of household members are similarly disruptive to children’s trajectories, and I encourage future research to take a holistic view of households and household instability.
Q. Can you speak briefly about the different ways that disadvantaged Black youth respond to adversity?
DeLuca: Part of the reason we had such interesting insights in this paper was because we had qualitative data that allowed for us to see the range of responses to adversity.
As one might imagine, in response to some adversity – like housing shocks, the experience of the premature death of important people, their own experience of victimization – some young adults withdrew socially from friends and from school. Some turned to coping mechanisms such as substance use. But we saw another set of responses were more common, which included some isolation, but it was more of a selective social isolation. They’re in a risky environment and there are peers who may or may not be what they want to be about. Maybe some peers were running a corner. Maybe some were in college. And these young adults focused on spending time with those who aligned with their goals. A total withdrawal is the extreme case and may be what you might expect for some of these young people who have experienced so much disadvantage and trauma. But what we saw instead was really a more adaptive selective social engagement with peers and an investment in relationships with peers, parents, and some mentors, coaches, and teachers that were really aligned with where they wanted to go in life. And we saw an avoidance of risky peers or peers they thought might derail them. In other words, for most, not a full-scale retreat and withdrawal into substance abuse or running away, but something more adaptive. Connecting and investing more in relationships that were more likely to help promote their long-term success and divesting from relationships that might get in the way. In addition, we saw that some that some young men and women were investing in enrichment activities at school – some clubs, some sports, some extracurriculars – or they were investing just in school itself, or they were focusing on work as another protective and promotive strategy. So, in other words there’s a range of responses to adverse experiences, which I think is interesting and stands to reason, but you need this rich, qualitative work, to help that really pop out.
The qualitative data also importantly underscores the tradeoffs that have to be made when deciding how to protect oneself or deciding the right thing to do for one’s family. I think another key point of this paper is that the tradeoffs that some of the young men and women that the team met in Baltimore had to make were really stark. If you have family members who are compromised by health issues or some parental estrangement, then making sure your siblings are okay, or your mom is okay takes precedence over leaving home to go to college. So, there again, is the range of responses. You have a complex response – in order to take care of people important to them, they have to forgo their own aspirations for college or certain jobs. in order to survive—or help others survive—in the near term, you may have to make decisions that compromise your ability to thrive in the long run. And that’s a brutal trade off that many of these youth are faced with when having to contend with some of these violent environments, resource and financially constrained families, unstable home lives, and under resourced schools.
This paper is mixed methods and it’s interdisciplinary. We’ve got an economist, Nicholas Papageorge on the paper with a former grad student, Joseph Boselovic and me, who are sociologists. And we really dig in and see what this study reveals to us about behavioral theory with regards to these tradeoffs. For example, it may seem suboptimal to not go to the best school you can go to that might be far away. But you need to understand a bit more about what’s underneath those decisions. One young woman in the study, Rhiannon, whose brother was murdered, was worried about something happening back home if she went away and decided to stay closer. These are the sorts of tradeoffs that are hard to see without the perspectives of the young people themselves, that underlie traditional statistics, which tend to focus on dummy variables like low income or high income, White or non-White, experienced in adversity or not. But we want to understand the complex adaptations, tradeoffs, and strategies that come with having to contend with some of these circumstances.
Q. It has been shown that the stress induced by natural disasters is detrimental to learning outcomes. How does the experience of local politicians affect the severity of natural disasters’ impact?
Argote: In our paper, Manuel Alcaino and I look at the impact of the 2010 earthquake in Chile, which was actually the fifth strongest earthquake ever recorded. It had short- and long-term effects on learning outcomes in math and Spanish, but especially Spanish.
But I am a political scientist, and I want to give you a political angle. We all know that natural disasters are an opportunity for politicians to show up and try to do something and mitigate the situation. We claim in our paper that these events are more of an opportunity for local politicians in the sense that they know better what’s happening in the field. There is a large distance between that and the president. In the U.S. it would be like the distance between the federal government and what is happening, for example, in North Carolina or Florida right now.
So, in the case of Chile, instead of focusing on the central government, we focused on local government. And we looked for a proxy or indicator of quality or experience for local politicians and how those qualities impact learning outcomes. Perhaps a more experienced politician or a veteran politician will know how to more effectively delegate resources in reconstruction efforts, for example. The earthquake damaged schools and many schools closed. Some students were out of school for weeks or even months. So, reconstruction was crucial after this event.
What we show is that more experienced politicians – and we measure experience by whether they were re-elected or if they were second term mayors – were more able to mobilize educational resources. Mobilize resources can mean either being more effective in the use of their own local resources or being able to more effectively transfer resources from the central to local government. More experienced mayors were able to actually completely mitigate the effects of the disaster. Let’s put it in numbers. If on average, the detrimental effect on test scores was, let’s say, 0.2 or 0.3 standard deviations, among the localities with an experienced mayor, who was able to mobilize the resources, the effect was zero. It was actually as if the earthquake hadn’t existed.
So, we find that the effect is zero and then we say, “Okay, why?” And then we start looking at all these measures of research mobilization, particularly pertaining to education. And this connects to literature in political science that suggest that re-elected mayors could be higher quality. That is one option, in the sense that at least winning a few elections say something about your capacity to connect to voters, at least. And that is not mutually exclusive from the fact that they may also be more able to mobilize the kind of resources that are crucial in this moment.
There is some black box here that we don’t know. Someone can ask, “Does mobilized resources mean you’ve skipped the line?” Perhaps. Perhaps, they were able to skip the line. Say I have a direct contact with the president or direct contact with the governor. You know, things like that. We will need more qualitative evidence to know that. But that is, of course, a possibility.
Q. What groups are more likely to experience flooding? How does flood exposure impact learning outcomes?
Khalid: Our research in rural India reveals a clear pattern of unequal flood exposure that follows traditional social and economic hierarchies. To understand this, it's important to note that Indian society has historically been stratified by caste - a hereditary system of social categorization - as well as by religion and economic status.
We found that socially marginalized groups face disproportionate flood exposure. This includes what are officially termed 'Other Backward Castes' and 'Scheduled Castes' - historically disadvantaged groups that have faced centuries of social and economic discrimination, somewhat analogous to racial inequalities in the United States. Muslims, who constitute India's largest religious minority, also experience higher flood exposure. What's particularly striking is that these patterns persist even today, reflecting how historical disadvantages translate into geographical and environmental vulnerabilities.
The economic dimension is equally important. Children from the poorest households and those whose parents have less education are significantly more likely to live in flood-prone areas. This creates a compound effect where social marginalization and economic disadvantage overlap. We also found that agricultural households face greater flood exposure, which is particularly concerning as these families often depend directly on weather-sensitive farming for their subsistence.
Interestingly, while Indigenous communities in India, officially known as Scheduled Tribes, generally face various forms of marginalization, they show lower flood exposure in our study. This is largely because they typically reside in forested or hilly regions rather than flood-prone areas.
Using a unique dataset that assessed children's actual learning outcomes – not just school enrollment – our analysis indicates that flood exposure appears to significantly disrupt children's education in multiple ways.
Our findings suggest that children in flood-exposed villages tend to show poorer performance in both mathematics and reading assessments. However, what's particularly noteworthy is that these associations don't appear to be uniform across different groups of children. Two key patterns emerge from our analysis.
First, our data points to a notable gender disparity in how floods may affect learning. The evidence shows that girls experience disproportionate negative impacts on their educational achievements when exposed to floods compared to boys. This potential gender gap is especially concerning given India's broader efforts to achieve educational equality.
Second, our analysis indicates that family economic resources likely play a protective role. Children from higher-income households appear to show more resilience in their learning outcomes during flood events. They seem better able to maintain their educational progress, possibly because their families can afford alternative educational resources or better mitigate flood-related disruptions. In contrast, our findings suggest that children from poorer households may face greater challenges in maintaining their learning progress when floods occur.
One of our most intriguing findings relates to what our data did not find any evidence for. While socially marginalized groups – such as lower-caste children and religious minorities – are more likely to live in flood-prone areas, we found limited evidence suggesting they face disproportionately larger educational penalties when floods occur compared to other rural children. This pattern might indicate that the primary educational inequality stems from their greater exposure to floods rather than from a greater vulnerability to flood effects. However, another possible explanation is that learning levels among these groups are already so low that there may be limited room for further decline when floods occur – what social scientists sometimes call a 'floor effect.' This possibility underscores the complex relationship between existing educational disadvantages and the impacts of environmental disruptions.
These findings potentially have important implications for policy. They suggest that addressing educational inequalities in flood-prone regions may require both reducing disparities in flood exposure and providing targeted support to vulnerable groups, particularly girls and economically disadvantaged children, when floods do occur. Moreover, the complex patterns we observe – including the possibility of 'floor effects' among marginalized groups – suggest that policymakers need to carefully investigate these underlying mechanisms. Effective interventions likely require understanding not only the immediate impacts of flood-related disruptions but also how these may interact with pre-existing educational disadvantages. This underscores the need for more detailed research into these complex relationships to inform more effective policy responses.
Q. What are the effects of wildfire exposure on infant health? How do maternal education levels impact these effects? What might explain this variation?
Rauscher: The first question I was asked to address: What are the effects of wildfire exposure on infant health? Wildfires harm infant health. Infants who are exposed to a wildfire in their second or third trimester during pregnancy are about 0.2 percentage points more likely to be born low birth weight than those born to mothers in the same county but conceived earlier or later relative to the wildfire. Exposure to wildfires in the third trimester increases the likelihood of preterm birth by 0.2 percentage points and increases the likelihood of fetal death by 0.04 percentage points. These effects on fetal death suggests that most studies, that look only at live births, underestimate the detrimental effects of wildfire exposure on infant health. The effects of wildfire exposure on infant health that we found are small. However, even small effects on the likelihood of low birth weight have substantial long-term consequences for a child's life chances. This can result in higher risk of disability, mortality and poor health later in life, as well as lower educational and labor market outcomes. With about 3.6 million births per year in the United States and about 4% exposed to a wildfire in their second or third trimester, even this slight increase in the rate of low birth weight would amount to about 288 additional low birth weight infants. And each underweight hospital birth costs about $24,000 more than a normal weight birth, which would result in an estimated annual cost of nearly $7 million.
How do maternal education levels impact these effects, and what might explain this variation? Wildfire effects differ by maternal education. The effects of wildfire exposure are worse for mothers with low education, especially earlier in pregnancy, in the second trimester. For mothers with higher education, the effects of wildfires are worse later in pregnancy, in the third trimester. In terms of what might explain this variation, the pattern seems to reflect selective survival with lower fetal survival among mothers with less education, resulting in infants who are less sensitive to additional environmental insults, if they make it to the third trimester. To say this another way, wildfire exposure earlier in pregnancy reduces the chances of fetal survival more among mothers with fewer resources. By the third trimester, fetuses carried by less educated mothers are more highly selected. They've already made it through more and are less sensitive to additional environmental insults. In contrast, fetuses carried by more educated mothers are more likely to survive wildfire exposure earlier in pregnancy, if they're exposed, suggesting that then they're less selected and may be more sensitive to later environmental insults.
A key thing to note is that the differences by maternal education in wildfire effects are not due to differences in exposure or different maternal behaviors, such as changes in prenatal care or tobacco use. Looking at what explains the effects of wildfire exposure, it seems that the effects on infant health are due more to maternal stress than to air quality. Wildfires are stressful, they're disruptive, and they have substantial economic effects, increasing stress for mothers and for others living there. And stress has negative health implications for mothers with both low and high education levels. However, mothers with less education have higher baseline risk factors, such as exposure to chronic stress, that can increase the likelihood of negative health outcomes from additional stress or poor air quality during a wildfire. So, wildfires increase stress for everyone, but the effects are worse for mothers with fewer resources due to higher initial risk factors like chronic stress. This leads to worse health outcomes for low-educated mothers, especially earlier in pregnancy.