Avner Seror, Journal of Political Economy, forthcoming
Understanding how children develop is a fundamental challenge for academic research in many disciplines, from medicine to economics. Since the pioneering works of James J. Heckman and co-authors in the early 2000s, a vibrant economic literature on child development has emerged. Through estimations of the technology of skill formation, economists working in this field demonstrated that gaps in child development reflect large differences in parents’ investment in their children. They also showed that parents profoundly affect children’s personality and other skills not captured by cognitive abilities. Undoubtedly, these findings have enabled social scientists to better understand how children develop, and policy makers to design public policies that foster human development.
Focused on the technology of skill formation, however, the economic approach to child development says little about the interactions through which parents shape children’s multiple skills. This appeared to me as an important gap in the economic literature. I started to investigate this issue during my PhD at the Paris School of Economics, around 2016. As often in my research, I began with readings well beyond the (usual) scope of economics: Anna Freud, Donald Winnicott, and John Bowlby. Although their views of child development differ on several important points, they all describe the incredibly powerful link that parents and children develop. A link that emerges from care, and that literally becomes the backbone of a child’s emerging identity. In the words of François Mauriac, which I used to open the paper: “We are, all of us, molded and remolded by those who have loved us, and though that love may pass, we remain none the less their work” (The Desert of Love, 1949).
In the paper, I introduce a theoretical model of how parent-child interactions impact child development. My approach is rooted in developmental psychology. It is based on the idea, first introduced by John Bowlby, that through care, sensitivity, and positive feedback, a parent enables her child to build mental representations of himself as loved and competent. These mental representations shape the child’s noncognitive skills by increasing his motivation to learn and explore his surroundings. Hence, the parent’s interactions with the child create a fundamental complementarity in the formation of cognitive and noncognitive skills. As the child learns from the parent, he forms cognitive skills while acquiring noncognitive skills that increase his motivation to further learn.
This model explains the evidence from a vast literature, at the same time generating new insights. First, the model provides a single unifying explanation for well-known properties of the production function of skills. It is well established that (a) skills are selfproductive, (b) noncognitive skills foster cognitive skills, and (c) there is a dynamic complementarity between early and later investments. All these properties can be explained by the mechanism of the model.
Beyond the established properties of the production function of skills, I show that a child who increases his noncognitive skills by interacting with his parent is better equipped to learn from his surroundings. This prediction echoes the celebrated first experiments that enabled generations of psychologists to classify different types of parent-child relationships, depending on children’s observed behavior. Indeed, a child’s exploration of his surroundings while his parent is absent reflects a secure attachment to the parent, which is built through the parent’s care, sensitivity, and positive feedback. Conversely, limited exploration reflects a parent-child relationship where the child is insecurely attached to the parent.
I also use this model to study the effect of early exposure to the media devices that are ubiquitous in most economies. There is a growing consensus that early media exposure is negatively associated with a variety of child outcomes. This negative association is not straightforward, since media devices (e.g. smartphones and tablets) stimulate children and can often foster substantial learning through game apps or educational content. Additionally, children growing up in more disadvantaged families appear to be particularly exposed to media devices. The model provides an intuitive explanation for these phenomena. First, I find that parents expose their children to media devices as a substitute for their own caring effort. The more a child is exposed to media devices, the less effort he puts into learning from the parent. Thus, he is less able to see himself as competent and loved, he acquires less noncognitive skills, and he becomes less motivated to further learn from his surroundings. Second, according to the model, a parent’s caring effort is less likely to be replaced by early exposure to media devices when the parent has “better” characteristics, which implies that such exposure can amplify child development inequalities.
I see at least two potentially interesting avenues for future research. The first is related to causal estimates of the effect of early exposure to media devices on various measures of child development. To my knowledge, little attempt has been made to explore this issue, although there is a consensus in the medical literature that early media exposure is negatively associated with child development outcomes.
As long as we lack rigorous evidence, we will remain unable to provide clear guidelines for policy makers or to estimate the cost of early media exposure on a variety of societal outcomes. Second, this theory is grounded in the attachment theory, which enabled generations of psychologists and psychiatrists to study the complex linkage between children’s mental health and parent-child relationships. It could also serve as a starting point for quantitative analysis of the socio-economic determinants of parent-child relationships and early childhood mental health.
→ This article was issued in AMSE Newletter, Summer 2022.
© Kelly Sikkema / Unsplash