By Lisa A. Gennetian, New York University
The role that the behavioral sciences can play in nurturing interdisciplinary frameworks was recently showcased in the May 2018 article in this journal by List et al., describing the ways in which theories from behavioral economics, public health, and early childhood development collectively, more so than independently, can improve our understanding of strategies to reduce socio-economic gaps in children’s educational outcomes.
In an earlier blog, Kalil shows how behavioral sciences provide a novel complementary framework for modeling parenting, which is otherwise treated as an idealized set of characteristics and styles according to purer interpretations of child development theory. These are fine demonstrations of the promise of behavioral economics as an interdisciplinary framework, bridging disciplinary divides: I argue here that behavioral economics is also an ingredient for an interdisciplinary framework of household production that bridges models of parenting (espoused by child developmentalists) and recognizes the essential interaction of home inputs into the education production function (that otherwise offers little guidance about how parents, per se, allocate resources).
Children spend the majority of their time with parents or comparable adult caregivers in their home, with recent estimates showing on average that young children spend less than 5 hours/weekday in any nonparental care arrangement. As children age, parents continue to make decisions about where and with whom their children spend their time. Indeed, parents are the most influential actors affecting children’s development until young adulthood; and, at a minimum, gatekeepers of how resources are allocated on their children’s behalf. Conventional models of household production predict that these resources are allocated as trade-offs in paid work, home production, and leisure. How time spent reading, going on educational trips, or money put away into savings accounts is determined depends on comparative advantage across paid, home and leisure activities, or is some optimal outcome of a negotiation.
Under assumptions of full information, if parents behave optimally as the household production framework purports then this leads to the conclusion that sub-optimal parenting (from the child developmentalist perspective) or suboptimal educational production (from the educator perspective) is due to not enough time or not enough money. And, herein lies the dilemma: neither time nor money seem to have the influence on parent behavior as much as household production theory predicts. Parents miss opportunities to sign their children up for programs, misestimate their children’s development, and misjudge parenting information, resources or services. Perhaps these latter factors can help partially explain why the public investments motivated by Heckman’s quantified returns to early childhood interventions are often unused or underutilized, and often at high financial social cost.
To get beyond time or money solutions behavioral economics offers some ideas: what if parents, like all people, get distracted with other crises and do not have attention available at the right time to fill out a form, or show up to a meeting? What if parents, like all people, hold certain identities that shape their decisions or harbor stereotypes about who needs or participates in programs? Money may certainly help overcome some of these barriers, but so might other less costly options. Indeed, both high and low income parents report similar feelings of meaning and stress during the time they spend caring for their children. The behavioral economic lens supplements the household social contract as one that might be subject to a variety of other factors influenced by caregiver identity, social norms and expectations, and even leaves room for optimization based on available attention and mental bandwidth of prospective caregivers.
In my research lab (beELL at New York University) we are seeking to better understand these influences on parent engagement and parent behavior
In one of our studies we integrate an existing city wide early language texting campaign (aka “big” policy solution) with an existing newborn home visiting program by defaulting new mothers, and any other interested caregiver, to receive the early literacy and education content during the first home visit. Mothers can opt out at any time. When we tested this through a randomized design, we learned that this relatively low cost strategy of shifting the default from an opt-in (via pamphlets and city wide media campaigns) to an opt-out increased enrollment in the text-based early literacy program by over 90 percentage points. Switching to an opt-out minimized inertia and procrastination as potential bottlenecks interfering with intentions of enrolling, and did so without impeding on free choice. Conventional household production theory would predict no difference between opt in and opt out; but rather that such decisions are an active choice based on an evaluation of costs-benefits with opt-in or opt-out resulting in the same outcome.
In another study we are learning that parents, like all individuals, experience stigma, and hold preconceived judgments that interfere with self-concept in ways that are situation specific. Using an online platform to recruit over 1,000 parents of children less than age 13, we randomly assigned half to receive a pride-based affirmation writing exercise (write about a proud moment) vs. a neutral writing exercise (write about your day yesterday). Participating parents spent roughly the same amount of time writing for either prompt. Parents who received the pride-based affirmation were more likely to express interest in parenting materials and attending a parenting program, and felt better about themselves, than parents who had the neutral writing prompt. Stigma has been proposed as an ex post explanation for observed lower uptake of public assistance and often falls into the unobserved parameter bucket in formal household production modeling. This example points to the ways in which stigma and identity can confound evaluations of benefit or cost, and identity priming and affirmations can shift and shape interest and follow-through and, as a result, potentially affect daily interactions with children.
Throughout our studies we find that grandparents are as likely as parents to drop and pick up children from programs and serve as gatekeepers of teacher communication and out of classroom program curricula. Fathers are lingering around door corners when home visitors come by and, are passive influencers of information and education. Finally, siblings are trusted social friends and relied upon by adults to engage with their younger brothers and sisters in educationally productive ways. These actors in children’s lives are not explicit in the household production function nor commonly the featured adult in models of children’s development. Only targeting mothers, whether deliberately or by default, misses opportunities to capture others who are present and available for children.
A household production lens to child development makes good sense. Parallel theories of parenting also make good sense. And, we need models to help guide optimal production of education. Behavioral economics offers a way to thread these frameworks together in ways that complement, rather than compete, and that acknowledge parents, and other caregivers, as the optimizing and active agents influencing children’s lives.
With this I note a challenge inherent in any interdisciplinary endeavor, related to identifying mechanisms supporting children’s development. Ex ante behavioral economics, like other frameworks and theories, can inform strategies that ex post may stimulate changes in parent behavior or children’s development – but not clearly inform the mechanism of change. Goal-setting tools, as an example, are consistent with theories of planned behavior (to motivate individuals), as well as a behavioral economics to facilitate follow-through (among already motivated parents). Any resulting goal-setting tool evaluated in the field may suggest (but is not a direct test of) a mechanism, and indeed could support ex ante insights from either theory. The extent to which this might matter depends on the objective: policy focuses on outcomes and might be agnostic to the theoretical mechanism – if it is not a mechanism being targeted. One goal for scholarship, however, is to delineate mechanisms in the way that can feed back onto theory. In her blog Kalil offers one example of a study to pinpoint the influence of present bias. Being alert to the contributing role of mechanisms born out of heuristics and biases within behavioral economics will not only enhance theory but also support effective policies for children’s positive development.