Behavioral Economics and Parent Participation in an Evidence-Based Parenting Program at Scale

Theoretical frameworks like the theory of planned behavior (TPB) (Ajzen, 1991) and the health beliefs model (HBM; Rosenstock, 1974; Salari & Filus, 2017) describe how preexisting beliefs, attitudes, and values affect intentions and behaviors. The TPB posits that attitudes (i.e., personal opinions about the behavior), subjective norms (i.e., perceived social pressure to perform the behavior), and perceived control (i.e., perceived ease or difficulty of performing a behavior) determine intentions and, subsequently, behavior (Ajzen, 1991). The HBM posits that behavior depends on the value an individual places on a goal and the individual’s sense that a behavior will lead to achieving that goal (Janz & Becker, 1984). These frameworks complement conventional economic-motivated approaches of encouraging attendance by reducing transaction and related costs (e.g., providing transportation, child care, and food) or by providing financial incentives or rewards for attendance (i.e., covering opportunity cost, in the language of economics). Researchers have used the TPB and HBM frameworks to measure predictors of engagement and to recommend parent engagement strategies (e.g., newsletters to influence parents’ perceptions (Randolph et al., 2009) and norm-based strategies (Bracke & Corts, 2012). Winslow et al. (2016) harnessed insights from TPB to improve parent engagement by including motivational interviewing and goal setting to increase parents’ follow-through with intentions.

Although existing frameworks suggest useful ways to achieve desired engagement with programs, gaps remain. Some strategies are low-cost and scalable (i.e., giving families program information via a newsletter). Other strategies require more resources that may be feasible to offer on a large scale (i.e., child care cost subsidies or individualized motivational interviewing). Furthermore, existing frameworks offer little guidance on the role of timing of outreach strategies at key decision-making points.

BE starts from an understanding of human decision-making as a function of evaluating costs and benefits (as proposed in conventional economics) coupled with the way humans are influenced by their psychological biases and contexts. BE expands on existing theories by challenging the idea that parents consistently act in ways that reflect carefully constructed intentions or pre-existing values about a behavior and, further, do so in a context-free way. As such, BE turns a lens on how current and prior experiences with poverty, financial instability, structural inequality, and/or discrimination can shape or undermine intentions (Gennetian & Shafir, 2015; Mullainathan & Shafir, 2013). Poverty, as a circumstance of scarce and unstable economic resources coupled with repeated racially or ethnically marginalizing experiences, is especially draining on cognitive resources (Gennetian & Shafir, 2015). First, it is exhausting to engage in cost–benefit trade-offs and calculations to evaluate future choices and behaviors, and thus, the easier choice might prevail even if it is not the preferred choice or the choice known to have better long-term prospects. Second, the consequences of this cognitive (decision making) pressure under circumstances of scarce and unstable resources can be more damaging than for those with economic cushions who can recover, for example, from a financial error.

BE considers the role of cognitive resources, such as attention, as scarce and finite, similar to the conventional economic ideas of money and time as resources. Thus, heuristics or other patterns of information, such as those presented in society or by peers, are often relied upon to inform decisions implicitly or explicitly (Baddeley & Hitch, 1974; Valcke, 2002). The influences of such psychological resources and factors can sometimes conflict with or overcome the influences of targeted education or information-based strategies, particularly if they are complex and difficult to digest. The implications of finite cognitive resources are that allocation of such resources to one aspect of family life (e.g., immediate demands like paying rent) may drain cognitive resources for other aspects of family life (e.g., reading or playing with a child (Gennetian et al., 2016) or attending a parenting program). Thus, how parents make sense of information or situations might result in behaviors that deviate from values or intentions.

BE conceptualizes decision-making as consisting of fine-grained moments, such as picking up or deciding to read a brochure about a program, that have spiraling effects on larger decisions typically observed as behavior change. All of these micro-decisions—when combined—contribute to the goal of program attendance and, further, shifts in daily practice based on program learnings. BE focuses on these decision-making junctures with the goal of keeping parents on the ramp to attendance. Lastly, the BE framework offers insights on strategies that can be feasibly implemented within existing program infrastructure in low-cost and scalable ways. Typically called “nudges,” these BE-informed enhancements take the perspective of choice architecture, that is, whether programs are implemented in ways that facilitate or interfere with decision-making intentions (Thaler & Sunstein, 2008).

Recent randomized controlled trials have shown the effectiveness of BE strategies for increasing parent attendance and engagement in a variety of early childhood programs (Gennetian, in press). Researchers used BE strategies such as child-friendly planners, reminders, and personalized invitations to increase Head Start parents’ attendance in a school readiness program and to increase parents’ time spent with children on educational activities at home (Gennetian et al., 2019). Another set of researchers used the BE strategies of reminders, goal setting, and social rewards to increase the time parents spent reading to their children (Mayer et al., 2019). An experiment testing the behavioral concept of default options found that enrolling mothers in a text message-based informational program and providing the option of opting out led to significantly more mothers receiving information, compared with the default of requiring mothers to self-enroll in the program (Gennetian et al., 2020). A survey-based experiment showed that written self-affirmation exercises increased parents’ self-concept and their interest in parenting support programs, especially among parents with a high fear of judgment (Hill et al., 2020). Together, these studies provide increasing evidence of the potential of BE-informed strategies to support ongoing parent outreach efforts and optimize participation in programs.

ParentCorps is an enhancement of publicly funded pre-K programs to bolster parent and teacher capacity to support children’s development in the face of early childhood adversity, scaled up as part of New York City’s Pre-K for All starting in 2016 (Brotman et al., 2021). ParentCorps includes multiple components—a 14-week parenting program, which is the focus of this paper, as well as professional development for pre-K teachers and staff and a social-emotional learning program for pre-K classrooms. ParentCorps outreach begins at the start of the academic year with a variety of written materials including flyers, posters, and brochures. Pre-K teachers and family support staff host welcome events and strive to personally invite each pre-K parent. ParentCorps provides meals and small incentives (i.e., raffle tickets for gift cards), and when the program is held in the evenings, parallel programs are offered for children (i.e., social-emotional learning for pre-K children, arts activities for siblings). Parents are always welcomed regardless of constraints (e.g., arriving late from work), and the invitation extends to all adults in the child’s life (e.g., grandparent, aunt). Throughout the program, parents are encouraged to practice parenting strategies at home while receiving support from the facilitator and peers.

In two randomized controlled trials in pre-K programs attended primarily by Black and Latinx families, ParentCorps strengthened teacher-parent relationships, parent involvement in children’s learning, and use of evidence-based parenting practices (e.g., routines, parent–child play, positive reinforcement, discipline); prevented the development of child behavioral and emotional problems at school; improved academic performance and achievement test scores; and, among higher-risk children, prevented obesity and improved health behaviors (Brotman et al., 2011, 2016; Dawson-McClure et al., 2014). Careful analysis of attendance in the second trial, in which university-based staff and pre-K staff shared responsibility for parent outreach and program implementation, showed that 58% of families enrolled in the pre-K programs attended at least one session, with considerable heterogeneity across schools (range, 44% to 75% of parents attending at least one session). Attendance increased from 50% in the first year of implementation to 65% in the fourth year. Parent attendance was not predicted by ethnicity, neighborhood poverty, or baseline parenting or child behavior (Dawson-McClure et al., 2014), indicating that a wide spectrum of families engaged.

In moving from efficacy and effectiveness trials that were conducted in school-based pre-K settings to scaling up in schools and center-based pre-Ks throughout the city, ParentCorps aimed to support pre-K staff in implementing the program with comparable success in family engagement. This required revisiting materials and strategies to be user-friendly and streamlined so that pre-K staff could feasibly take on responsibility for parent outreach as part of their full-time jobs. Further, the revision of the outreach materials was an opportunity to codify the program’s approach to building respectful relationships with parents, not assuming that all parents should prioritize participation in the program, but based on the belief that all parents deserve to be offered the opportunity to participate in ways that affirm their worth and support their autonomy.

This study is a partnership between scholars with expertise in behavioral economics and ParentCorps developers and coaches who support program implementation by pre-K staff. The first aim was to document how BE concepts could feasibly be translated into concrete outreach materials and strategies for ParentCorps as it scaled up in a large urban school district. The second aim was to examine how testable components of the BE-infused outreach strategies influenced parents’ participation in the program, with the hypothesis that parents receiving BE-infused outreach would on average have higher levels of program attendance.

There was substantial descriptive variation in attendance across the core and BE-infused outreach sites (see Online Resource Table B). The percentage of families that ever attended varied from 12% (at site 3, a BE-infused outreach site) to 73% (at site 9, a core outreach site). Among families who ever attended, the average percentage of sessions attended varied from 23% (at site 9, a core outreach site) to 60% of sessions attended (at site 2, a BE site). Among all sites (see Online Resource Table C), attendance was highest at session 1, with 19% of all families attending, and decreased thereafter, with 11% of families attending session 5. In the core outreach group, 37% of families ever attended, whereas 30% of families in the BE-infused outreach group ever attended. However, those in the BE-infused outreach group attended a higher percentage of sessions than those in the core outreach group (38% vs. 33%).

As shown in Table 2, the site-level randomized implementation of BE-infused outreach materials did not show a statistically significant impact on attendance. Families at BE-infused outreach sites had a higher predicted probability of attending session 2 (b = 0.064, se = 0.04, p = 0.185), session 4 (b = 0.000, se = 0.02, p = 0.994), and session 5 (b = 0.005, se = 0.04, p = 0.904) and a lower predicted probability of attending session 1 (b =  − 0.031, se = 0.06, p = 0.605) and session 3 (b =  − 0.038, se = 0.04, p = 0.402); however, these estimates do not meet the threshold of statistical significance at p < 0.05. As we expected, larger sites (i.e., total pre-K enrollment) had lower rates of attending any session between sessions 1 and 5 and ever attending.

The sub-study showed positive impacts of time-varying text messages on families’ attendance across two attendance metrics (Table 3). In the BE-infused outreach sites, families randomized to receive messages at varying times were more likely to ever attend the program (b = 0.111, se = 0.06, p < 0.05), compared with families at the same site who received messages at a consistent time. This difference was significant for session 2 attendance as well (b = 0.102, se = 0.05, p < 0.05), which corresponds with the onset of text messaging.

This study describes the contribution of BE for optimizing participation in parenting programs through a unique collaboration between developers of an evidence-based parenting program, the team supporting implementation at scale, and researchers with expertise in BE. We used existing research on parent engagement and interviews and focus groups with ParentCorps parents, developers, and coaches to inform our identification of relevant BE concepts, such as drain of cognitive resources, stigma, fear of judgment, negative attribution, social norms, social identity, present bias, status quo bias, and endowment effect bias. These concepts in turn informed our development of BE-infused print and digital outreach materials tailored to the specific context of the ParentCorps parenting program. The BE-infused bundle included a “Real Talk” brochure insert, a teacher letter and wearable button, a postcard self-affirmation reflection, an active choice invitation to participate in the parenting program, and a suite of text-based messages. These BE-infused outreach materials were collaboratively designed with program developers and implementers aiming to minimize operational or varying labor and material costs. The BE-infused bundle was successfully implemented by pre-K staff, meaning that all BE outreach materials were distributed to program sites and subsequently distributed to families consistent with the standard outreach schedule, thus demonstrating feasibility in the context of scaling up efforts to serve families in historically disinvested neighborhoods. The ParentCorps program continues to use most of the BE-infused printed materials (i.e., the Real Talk insert, teacher letter, affirmation postcard, and active choice invitation) as part of their outreach. The prize drawing tickets were discontinued due to district-level challenges in funding prizes.

We utilized an iterative process to translate BE concepts into outreach materials that incorporated input from parents and practitioners consistent with the Rapid Synthesis and Translation Process (Thigpen et al., 2012). The design of the active choice invitation is illustrative of how our team collaboratively addressed challenges. We began from the idea of addressing inertia in decision-making by shifting the default from opt-in to opt-out, as is typical in BE research. This was initially operationalized as a ParentCorps membership card, given to all families at the pre-K site, along with a session date/time assignment as a strategy to signal enrollment without requiring any action from parents. Through review, we understood that this approach may not align well with principles of preserving autonomy, fostering engagement beyond initiation, and doing no harm. We also did not want default choices to be misinterpreted as parents falling short. Instead, we implemented an enhanced active choice invitation with an explicit option to choose “yes” or “no” that highlighted the benefits of attending sessions, avoiding the suggestion that non-attendance was an example of falling short as a parent.

In contrast, our embedded family-level randomization sub-study demonstrated that varying the time at which parents receive text messages every week—an insight around novelty of information as influencing behavior—improved attendance at session 2, which followed the onset of text messaging, and improved the overall likelihood of families ever attending, compared with families that received text messages at a consistent time.

We hypothesized that varying message delivery times capitalizes on two features of how families make decisions. First, receiving text messages at unpredictable times might increase the novelty of the communication and thus better draw attention to the messages, and, second, alternating the time of text message delivery increases the likelihood that any one message lands in a “viewing window” when parents may be attentive. Digital forms of outreach are increasingly popular methods for reaching children’s caregivers and are largely consistent with BE principles of light touch, low-cost, and streamlined approaches. We built on this by infusing BE thinking in message content, timing, and specifics (e.g., personalization). Less than 5% of the parents who received messages chose to opt out. However, text messages and related digital-based methods of communication come with trade-offs, particularly in the context of relationship-based programs implemented at scale. There are enormous labor costs of responding to messages if parents view them as modes of feedback or confirmation, with lack of response potentially reducing trust. As many more pre-K programs now communicate with families through text messaging and digital platforms, ParentCorps is working with programs to use BE-infused content in their messaging and varying the message timing based on the current findings.

This study relied on existing program data to measure outcomes, and thus, research costs were primarily to cover the cost of staff for conceptual work, design, and analysis, not data collection. Apart from the fixed costs of creating the reusable suite of BE text messages, a back-of-the-envelope calculation applied in a previous study finds that the per-family additional cost of BE-infused print and digital outreach was below $1.00, substantively lower than the approximate cost of $6.25 per family for motivational interviewing (Winslow et al., 2016).