Response to The Trouble With Texting
By Benjamin L. Castleman 03/16/2016
In The Trouble With Texting, Jay Greene offers a thoughtful review of my book, The 160 Character Solution, and more broadly a critique of the use of behavioral nudge strategies in education and public policy. As with any new policy strategy gaining momentum—especially one being scaled quite rapidly at the state and federal level—it is important to ask critical questions about the efficacy and ethics of nudging.
I fully agree with Greene’s assessment of the importance of demonstrating that the positive results from early nudge experiments—such as our summer melt texting study which Greene describes—can be replicated across time, in other settings, and by other researchers. I believe Greene is selective in the related studies he chooses to describe which have not demonstrated positive impacts on college enrollment. In addition to a successful replication of the summer melt study that my colleague Lindsay Page and I conducted, numerous other researchers have demonstrated that behavioral nudge strategies, from integrating financial aid application support into the income tax preparation process to sending college students reminders to renew their financial aid, can lead to large improvements in college entry and persistence. This is not to dispute the importance of additional research to investigate the particular settings in and populations with which nudge strategies are most effective, or the mechanisms through which nudges affect students’ decisions, but rather to observe that, in fact, multiple researchers have demonstrated, through well-designed experiments, positive impacts from nudge strategies.
I also am in complete agreement with Greene about the importance of investigating whether nudge strategies that encourage students to enroll or stay in college lead to holistic improvements in their medium- and long-term outcomes. He is right to point out that nudges that encourage students on the margin of going to college to enroll have the potential to do harm if the student only stays for a term and leaves with either financial debt accrued or psychological costs from struggling in a challenging environment. With generous funding from the Laura and John Arnold Foundation, we are in the process of conducting a longitudinal analysis of several nudge experiments we have conducted over the last several years, to investigate whether these interventions have positively affected persistence and completion in addition to enrollment. But I would also note that this concern is not unique to nudges; we could ask the same of other college access interventions, like last-dollar scholarships or counselor- or peer-outreach campaigns. In reality the question of whether the longer-term benefits outweigh the potential costs of interventions that shift where students go to school applies as well to early childhood, elementary, and secondary school choice initiatives.
Greene astutely observes that many of the behavioral nudge strategies that have been applied in education would be unnecessary if we could instead simplify the complicated procedures that students have to go through. While I agree that procedural simplification would be optimal, researchers and practitioners have been advocating for FAFSA simplification for over a decade, and while the Department of Education has introduced some important steps to streamline the application, it remains quite complex. In my mind this begs an important question, and highlights the primary point of disagreement I have with Greene’s assessment of nudging: While we continue to advocate for legislative and bureaucratic solutions to burdensome procedures and applications, should we intervene to support students to make more active and informed decisions during these critical junctures, or should we let them figure it out for themselves?
Greene’s contention is that any kind of nudge subtly imposes our judgment on to students of what is an optimal educational pathway or outcome, and that these “educational prods are much more likely to be intrusive and much less likely to be effective.” I dispute this assessment for several reasons. First, we are not introducing nudges to a vacuum in which there are no “prods” affecting students’ decisions. The status quo of complex information, complicated processes, and limited access to professional assistance are nudges in their own right: a confusing financial aid application, an unanticipated application fee, the lack of opportunity to visit a college campus all have the potential to convey the message that “college isn’t for you” to a hard-working, talented, but socioeconomically disadvantaged student. In subtle and explicit ways, these aspects of the status quo can actively push students away from pursuing educational opportunity. By providing students with simplified information, reminders about important deadlines, and access to one-on-one assistance, our goal is to support students and families to make active and informed decisions about the pathways that they feel are best to pursue. The details matter greatly here; our nudges are not telling students they should go to college, or constraining the choices that are available to them. In the context of our summer melt work, these are students who have applied to college, been accepted, and made an active choice at the end of high school about where they want to enroll. Our nudges are designed to help students follow through on their own intentions by providing them with information and assistance to which they might not otherwise have access.
Second, Greene focuses on the potential downsides of nudging students to follow through on their postsecondary intentions and plans. We should also ask ourselves what the cost of Greene’s “leaving people alone” strategy would be, or stated differently, who is most likely to be negatively affected if we stopped nudging. It’s probably not the students from affluent, college-educated families. Their parents are intensely engaged in the college and financial aid process, and have only become more involved over time. What parents don’t know or can’t do on their own, they can compensate for through college-educated social networks or by hiring private college consultants. In qualitative work we did for an earlier book on summer melt, we found that college-intending students from low-income backgrounds are no less determined to go to college, no less tenacious in the face of adversity, and certainly no less talented than their more affluent peers. What they lack is anywhere near the same access to guidance and support. We find that parents often want to help but lack sufficient college literacy to do so. Many lower-income students attend schools or live in communities with limited or any college counseling. As an advisor in one of our summer melt studies once said, nudge strategies are just “filling in for what middle-class parents do with their children all the time.” In the absence of proactive outreach to and support for low-income and first-generation students, many bright, resilient students who have worked hard to get to college and ardently want to go may fail to do so as a result of the behavioral obstacles they encounter. I believe this potential cost—both for individuals and society—is far greater.
Finally, Greene posits that “Just getting students past the FAFSA challenge, if texting is actually effective in doing that, does not improve the skills students need to get past all of the rest.” Keep in mind that these students have already demonstrated a substantial skill set—both the academic skills necessary to gain admission to college and the non-cognitive skills necessary to persevere and succeed in what are often under-resourced school environments. A growing body of evidence suggests that providing students with simplified information, reminders, and access to assistance during critical junctures, like FAFSA completion and the summer after high school, can help students navigate challenging bottlenecks and transition into collegiate environments where they’ve already shown the potential for academic success. That being said, I believe Greene raises an important broader question about when we should nudge in education. Should we nudge students to go to school every day? To do their homework? To make new friends on a college campus? Each of these actions is positively associated with student success—but are these behaviors that nudges can positively influence and sustain over time? Is it the role of education leaders or policy makers to intervene at this level of student decision making? These will be key questions to wrestle with as nudging becomes more popular in education.
As behavioral nudge strategies continue to expand in education and across numerous policy domains, we should continue to critically examine and debate the questions of replicability, long-term impact, and appropriate use that Jay Greene raises in his critique.
Ben Castleman is an assistant professor of education and public policy at the University of Virginia and a Fellow at the Behavioral Science and Policy Association. He is the author of The 160 Character Solution: How Text Messaging and Other Behavioral Strategies Can Improve Education, available from Johns Hopkins University Press.