Ariel Kalil is a Professor of Public Policy at the Harris School, University of Chicago. She is a developmental psychologist by training, and her work links developmental psychology with economics, e.g., the effect of parental job loss on child development. I had a conversation with Ariel about her work and thought it would be of great interest to our readers.
Kwang: Congrats on the NY Times feature last week. Not many social scientists make it to the front page. How do you feel about that?
Ariel: It was very exciting! It was fun having friends e-mail to tell me they’d read the article on their train ride that morning. What I was especially happy about was that the reporter got the story right. He was a very curious and thorough guy, and we spent a lot of time on the phone and exchanging e-mails over the past few weeks. He read all of my original papers carefully, and came up with some very good questions for me. And the families that he interviewed had stories to offer that really illuminated some of the quantitative findings from my work.
Kwang: Tell me a little more about your work. What are main themes you’ve researched and what motivates you to pursue these questions?
Ariel: I’m interested in how socio-economic conditions are associated with families’ well-being and children’s development. So, I’m interested in parents’ mental health and their interactions with one another and their children, and I’m interested in children’s behavior and academic performance. In many instances, there is a link between, say, family income or families’ employment experiences and these outcomes. I care, as all economists do, about whether these links are causal. But, in thinking more like a developmental psychologist, I am also interested in “getting inside the black box” to understand why these links exist, and what kinds of individual differences shape how strong these links are for different types of children and families. I’ve always been interested in applying social science to real world problems. The idea that my work might someday shape public policy that could help improve the lives of families and children is very motivating!
Kwang: In the paper featured last week by the NY Times, you show that young people are badly affected when their parents lose their jobs, and that this is true in single parent and dual parent families [paper]. Could you tell us more about these effects and what you think drives the differences between single mothers and dual-parent families? Between male and female parents becoming unemployed?
Ariel: I think there are likely different factors at play for single parent vs. dual-parent households when jobs are lost. First, these families look a lot different from one another in terms of a whole set of demographic characteristics. So, in some sense, it’s a bit difficult to compare the two kinds of families. One of the biggest and most obvious differences is that when a single mother loses a job, the family has typically lost its only breadwinner. These families are likely to already be strained economically, and to have few (if any) people in their set of friends and relatives who can help them out. In many cases, a job loss sets off a cascade of adverse events that can be hard to stop, such as getting evicted or having to move in with others to save housing expenses, and this might disrupt child care arrangements or where kids go to school, and so on. There is just a lot more instability in these families related to the families’ economic circumstances.
In dual-parent families, I think the situation is a little different, and, at least in the short term, I think the impact on well-being and child outcomes has less to do with the economic impact of the job loss than the psychological one; for instance, in the way that parents relate to one another and to their children. For example, most dual-parent families have two earners, and so the family hasn’t lost all of its income at once. And many of these families also have some resources they can draw on, either savings or help from other family members. The immediate economic threat may not be quite as great. Also, in the families from whom I’ve collected data, I’ve found that parents will typically try to cut back on other things before they cut back on spending for their children, so the kids are often spared disruptions in their daily lives. In these families I think the adverse impacts that we see have a lot more to do with stress and anxiety, which we know can be very damaging to family relationships and ultimately to children’s development. And I think a big factor in the current recession is how long it’s taking people to find new jobs. The number of “long-term unemployed” is at an all-time high, and parents are very worried. We may eventually see more of these families exhausting their savings, losing their homes and encountering the same kinds of hardships that single-parent families have been more likely to face.
The different impact when fathers and mothers lose jobs is a really interesting one. In our work, we have consistently found that the negative impact of fathers’ job losses is greater. And this is not simply because fathers’ earnings losses are greater than mothers’ (in fact, in the US, in 40% of dual-earner households women are the primary earners). This is an interesting puzzle that I’d like to try to figure out; unfortunately the data are not readily available on this particular issue!
First, I think that “stereotypical” gender roles are still alive and well in many families and that the idea of being the “breadwinner” is still very important to many men and that is may be a bigger psychological blow to them when they lose their job. Second, working women occupy a variety of roles – we see in time use data that women still do the lion’s share of caring for children and tasks around the house (cooking, cleaning, etc), even when they are employed full-time. It turns out that working mothers cut back on their sleep and leisure time to do all of these things. So it may be that during periods of unemployment these women spend their time at home more effectively than a similarly unemployed man – because they were already occupying those roles anyway. Also, in the families from whom I’ve collected data, there seems to be more strife over figuring out what fathers’ “roles” are going to be during a period of unemployment. Many fathers viewed spending 40 hours per week in an outplacement office or a networking group searching for a new job as a full-time job, whereas many of their working wives thought they could usefully be spending more of that time helping out around the house or with the children. And that created a lot of conflict, which I think is rooted at least in part in “societal” or individual views about how the responsibilities of running a family should be divided between mothers and fathers.
Kwang: The data in your paper is very comprehensively analyzed. The data appears to be from the 1990s. Do you think anything has changed since then, or are the fundamentals still the same? Also the main variables in the paper are whether the student (a) had to repeat a grade or (b) was expelled, both of which are relatively drastic. Why do you think the impact is so great (and not just the child doing a little less well during exams, for example), and are there any other impacts that you might have encountered on the children apart from in their schoolwork?
Ariel: In terms of the data, I should admit that a lot of what we can learn about this question is really constrained by data limitations. There just aren’t that many big data sets that have high-quality information on parents’ job experiences and more than a smattering of child outcomes. And if they have any measures at all of parents’ psychological well-being or parent-child interactions that’s really incredibly rare! Right now I’m working with some new US data that has more “intermediate” outcomes, such as grades and test scores, and child behavior problems, that I expect would be antecedents to the outcomes we examined in the paper you read and that was cited in the NY Times. I’m also very interested in how the phenomenon of parents’ job loss affects kids’ expectations for the future, because I think those expectations can affect choices they make and options they pursue. But again, these measures are really hard to find in a data set that is collected by an organization such as the US Bureau of Labor Statistics!!
You are right that the data are from a few years back. One thing that I think is different now, which I mentioned before, is the rate of long-term unemployment. And the rates of job loss and unemployment in general are much higher in the current recession than in the mid 1990’s. You might think this could go in different directions. On one hand, losing a job during a period when lots of other people are also losing jobs may make a person feel less “singled out,” because the experience is so common to one’s peer group. On the other hand, I think the issue of long-term unemployment is an important one. I think it’s quite possible that families would be more worried in the current era because they know that others in a similar situation are having a lot of trouble finding work.
Kwang: What do you hope government policy makers will learn from your study, and what actions do you hope they will take? What about managers at private companies: how should they respond to your findings (if at all)?
Ariel: Most of the government solutions we have right now relate to economic solutions for the unemployed; i.e., issues concerning the unemployment insurance program. Similarly, managers at private companies think about issues like severance pay, or how many weeks of an “outplacement service” they will provide to former employees. I think our work suggests that workers and their families could also benefit from some counseling services – to try to figure out such issues as how to discuss the job loss with the children, how to negotiate new roles and expectations within the household, how to find ways to emphasize positive mental health in these circumstances. My guess is that in times of high stress and anxiety many families have a hard time solving these problems entirely on their own.
Kwang: And what do you hope policymakers will not do based on your findings, i.e., are there things you worry about that they might misinterpret from your research?
Ariel: Well, as we discussed above, my research shows relatively fewer ill effects on children’s development when mothers lose jobs compared to when fathers lose jobs. I certainly don’t want this to be interpreted as thinking that we should neglect mothers in this phenomenon of widespread job losses. If our findings imply simply that women are somehow better able to cope with this type of event than are men, it doesn’t follow that they should have to!!
Kwang: In another recent paper, you show that being married makes a similar contribution to a person’s happiness as spending $20k per year on leisure activities. One of my colleagues jokes that he is now tempted to buy a $20k bike, while another says this evidence that governments are taxing singles unfairly for the benefit of married couples. What are your reactions?
Ariel: On to quite a different topic indeed! My colleague Tom DeLeire and I had a lot of fun discovering a few things about consumption and happiness in our recent work. We’re pretty certain ours is one of the first papers on this particular question using actual data on how people spend their money, though it addresses the age-old question “can money buy happiness?” I guess one of the first things to tell your colleague is that he will have to keep buying that $20k bike year in and year out, because our analysis, as you indicate, looks at the relationship between spending and happiness within a given year. Also, the happiness gains to marriage may accrue over time in a non-linear fashion, so your friend might have to keep ratcheting up his spending!
On the question of whether singles are being untaxed unfairly for the benefit of married couples, I could also point out that our results showed there is a “happiness boost” that comes from educational attainment, from being in good health, from not being lonely, and from having frequent social interactions (above and beyond the boost that comes from being married). Looked at this way, a lot of people are being taxed unfairly for the benefit of those whose circumstances make them happy.
Kwang: This particular paper claims to have some good measures of ‘happiness’; could you tell us a bit more about these?
Ariel: We used data from the US Health and Retirement Survey for this paper. One of the unique merits of these data is that it has a whole wealth of psychological variables we could use as controls in our analysis. For example, happiness is strongly related to personality type, which might also be related to people’s expenditure decisions. The HRS recently measured, for the first time, individuals’ personality, and many dimensions of their psychological well-being. So this allowed us to control for variables that typically go unobserved in other studies. And, as I said, it is the only data set that we know of that allows us to link happiness with people’s actual expenditures (controlling for their income and wealth). Most studies on income and happiness have focused on GDP across countries, or household income, but have not actually looked at differences in how individual people spend the money that they have.
Kwang: Thank you for this time with me. Can you provide some contact information so that people who are interested can follow up to learn more about your work?
Ariel: I’d be delighted if anyone wanted to read any of my work further. My webpage and a link to my CV is at http://harrisschool.uchicago.edu/Faculty/web-pages/ariel-kalil.asp. If you send me an e-mail I’d be happy to send you any of my papers. Thanks Kwang!!