Informing Public Policy with Social and Behavioral Science
Many of the challenges facing our society today—from military preparedness to climate change—have a social or behavioral dimension, as do the policies considered by government officials to address them. A better understanding of the factors that influence how people act and interact can help policymakers design more effective procedures.
The vast majority of policymakers are not trained as scientists. As a result, they have varying degrees of understanding about how the social and behavioral sciences can help them do their jobs. Likewise, the vast majority of researchers have little to no policymaking experience. As a result, researchers often approach policymakers in ways that policymakers find unhelpful.
As social scientist Robert Cialdini observed at a gathering of researchers and policymakers on Capitol Hill a little over a year ago, if the social sciences were a corporation, they would be renowned for research and development. But they lack a crucial element: a shipping department. Social and behavioral scientists do not have a distribution system to deliver what they know to decisionmakers, packaged in a form that they can use. As a result, a wealth of potentially useful information that could yield practical benefits for the public never realizes that potential.
I propose that the social and behavioral sciences move quickly to develop an efficient and effective “shipping department”—a mechanism for delivering the most useful findings and methods into the hands of public policymakers. Having been both an active researcher and a member of Congress, I have seen how the absence of a concerted effort to effectively communicate social science has limited its impact in critical policy contexts. The social and behavioral sciences can increase their social value by working to translate and transfer their insights to real-world policymakers in ways that stay true to the content of the scientific research, while responding to and reflecting the policymakers’ actual informational needs. In other words, to help policymakers better understand science, it is critical that social scientists better understand the kinds of information that policymakers do and do not need.
To that end, I suggest five actions that can help the social and behavioral sciences make a more positive impact on policy and our society. Although any one of these actions taken independently could increase the relevance and public value of the social and behavioral sciences, each endeavor will have greater value if all are pursued together. Let us consider each of the proposals in turn.
Use a collaborative, consensus process to identify robust scientific methods and findings that are of potential interest to policymakers. The social and behavioral sciences should join together to create a high-level, cross-disciplinary project involving leading experts in communication and learning along with actual policymakers. The goal would be to produce practical, empirically driven, and readily applicable presentations that are accessible and usable for policymakers at different levels of government.
This would, emphatically, not be a typical academic work. Rather, it would be practical and translational in purpose, and it would be guided by an awareness of real-world policy needs and how our methods and findings can help impact that policy. Just as important, it would derive communicative content and presentational strategies from the substantial knowledge base on these topics that the social and behavioral scientists have taken the lead in producing.
From the research side, the consensus process would be guided by the following questions: What do we believe are the most significant principles of how social and behavioral scientists approach questions relevant to public policy? Which of our theoretical insights and robust empirical discoveries are most relevant from a policymaker’s point of view? At the same time, we would ask policymakers to articulate the insights that they most want from the social sciences. In other words, what are the situations in which they would most value the knowledge that social and behavioral sciences produce? Are there situations in which social and behavioral science methods and findings can help policymakers avoid ineffective or counterproductive policies and programs while crafting more effective ones?
An example of the type of outcome that this process could produce pertains to “regression to the mean”—i.e., the long-known tendency in social science research for scores at the extreme high or low ends of a distribution to “regress” toward the mean on subsequent measurement owing to chance factors alone. Researchers are aware of the potential illusory effects of regression to the mean. Policymakers, however, often want to help those at the most extreme ends of the curve, such as students in the very lowest performing schools. As a result, laws may be enacted and programs created whose apparent effectiveness is an illusion. For example, a school where students test at an extremely low level one year is likely to perform at a less-dire level the following year, regardless of any policy intervention, because of the tendency to move toward the mean. This could happen for a variety of reasons unrelated to the new program—for example, if several particularly difficult students leave the school or some high-performing students join. Those unrelated changes, however, could produce higher average scores for the school, leading the policymaker to see an apparent improvement and attribute that improvement to the intervention. Researchers know that there are methodological design and statistical analytic techniques to guard against this error, but again, policymakers are not only unaware of the problem of regression itself, they are also likely to have very little knowledge of research design or statistical techniques.
Helping decisionmakers who care greatly about problems and the people affected by them to craft policies that are not prone to this and other comparable errors could help direct more resources to policies that have real effects and avoid policies with spurious or even harmful impacts.
As a second and related example, it might be very useful to provide policymakers with practical tools to understand how the findings of randomized, controlled trials can or cannot be appropriately transferred from one setting or application to another. In general, a collaborative effort among researchers and policymakers to first identify, and then more effectively communicate, methods and ideas for greater policy effectiveness and efficiency is one route to increasing the public value of social and behavioral science knowledge that already exists.
Develop a comprehensive and outcome-oriented entity to create more effective communication strategies. This entity would not just produce content, but also commit to evaluating and making public the relative effectiveness of different science communication strategies. When we discover social and behavioral science knowledge that has the potential to benefit the public, the realization of that potential will depend on the effectiveness with which the information is conveyed. We should develop a means of producing and disseminating such information that makes use of many modes of communication. These modes can include printed products, electronic publication and distribution, and, possibly, a new online journal directed to an audience of both policymakers and social and behavioral scientists. We should also consider developing practical handbooks or, perhaps better yet, massive open online courses (MOOCs) to communicate with the widest possible audience. Another possibility would be developing apps that employ decision trees, algorithms, or augmented analysis and decisionmaking to help policymakers use what the panel develops. In all such cases, we should commit not just to developing content, but also to evaluating the extent to which our target audiences find it valuable. If this entity is linked to the researcher-policymaker collaboration described above, that group could advise it on how best to evaluate the impact of its activities.
This entity’s main target audiences would be policymakers, those who support policymakers, and those who seek to aid policy processes. If sufficiently effective and accessible, these resources could also be used as part of graduate training in the social and behavioral sciences—providing templates for researchers and organizations that want to deliver valuable advice to policy communities. This resource, if sufficiently effective, should also be incorporated into the orientation and other services provided to members and staff on the Hill and in other legislative and policy bodies. The key is to develop content that this population believes is necessary for them to achieve their ambitions. Another, much broader aspirational goal would be to make this information available to the general public as they seek to understand social problems and policies.
Create an independent, non-governmental resource to which policymakers can turn to have more personal and ongoing interactions. Policymakers can use this resource to obtain credible and objective information about existing or proposed programs and legislation. In contrast to the model used in the United Kingdom, where the government has a central office of behavior, we should consider an alternative approach: establishing an independent, external resource—like the Congressional Research Service or the Congressional Budget Office—to provide an expert, nonpartisan sounding board to which policymakers could turn for feedback about the likely social and behavioral consequences of current and proposed policies.
This is a subtle but important distinction from how things work currently. The proposed resource would not replace existing entities such as the National Research Council or other organizations that offer analysis of social challenges and policy options. Rather, we can augment and amplify the effects of such analyses by creating a way for policymakers to routinely gain assistance in thinking things through as a regular part of their own policy development processes.
As this entity responds to requests from legislative or administration policymakers, specific policy examples might be offered and specific findings of research would be presented, but the purpose of doing so would be to demonstrate how certain actions and consequences are related. This resource would not advocate for a specific policy for a specific problem.
Suppose, for example, that a policymaker is concerned about the consequences and costs of elevated high school drop-out rates. This resource would provide a venue for policymakers to learn about how social scientists have examined the issue, what attributes of the problem are most and least likely to be affected by various policy alternatives, and what mistakes or successes researchers have documented. If the resource could provide this type of information to policymakers in an accessible and actionable way, it could help them make more effective decisions.
Our purpose in this endeavor would be to transform how individual policymakers and their staffs understand and use directly relevant scientific methods, findings, and concepts in their thinking and actions. To make this project work, however, it is essential that we focus not just on how to educate policymakers about science but also to help social and behavioral science researchers better understand the situations that policymakers regularly face. This resource will be of value to policymakers only if researchers understand enough about policymakers’ needs to provide the kinds of information that policymakers can use.
Establish a series of presentations that are readily accessible to policymakers and staff on Capitol Hill and elsewhere in government. As the consensus panel does its work and identifies the key challenges and relevant insights and findings, we also need to take our findings and methods to where the audience is and show them what we know, how we know, and why it matters. The previously proposed journal, MOOCs, apps, and other mechanisms would all contribute to this, but we should also begin to have events on the Hill at convenient times, with food and other incentives to attract interested staff.
When I chaired the Energy and Environment Subcommittee of the House Science and Technology Committee, we initiated a series of what I called “Gee Whiz” presentations on the Hill. These were intended simply to highlight for staff and members of Congress the most interesting and exciting findings from Department of Energy scientists. The events were a huge success and drew increasingly large and very interested audiences.
There is no reason the social and behavioral sciences could not do the same. For these to be successful, what is presented at such events must be offered by compelling speakers—not merely people with impressive academic or research credentials—but strong, engaging communicators. These events must also address topics beyond esoteric subjects or psychological parlor tricks, and the presentations must not be laden with the usual “on the one hand, on the other hand, more research is needed, etc.” unless it is relevant, interesting, and meaningfully illuminates the topic. Topics and speakers must be tightly and strategically chosen and the information must be practical, have substantial magnitude of effect, and must speak to people on both sides of the aisle. And again, it must incorporate what we know about cognition, emotion, and behavior change.
For example, in the fall of 2013, the National Research Council organized an event in the U.S. Capitol that featured public benefits of social science. However, the event was not billed as such. Instead, we framed the proceeding as “How Social Science Saves the Government Money.” The event was framed this way to reflect the needs of the target audience—in this case, staff who could gain a type of knowledge that members of Congress could then use to benefit their constituents. The event featured leading scholars from several disciplines and former members of Congress from both major political parties. The presenters delivered sharp and cogent examples of how the social sciences transformed the provision of health services, enhanced the effectiveness of military strategies, and increased the efficiency of environmental programs. Instead of engaging an audience of congressional staffers in abstract conversations about science, the presenters highlighted how science could help them do their jobs more effectively. If done right, these presentations should become the kinds of events that members and key staff look forward to and make a point of attending because they value the intellectual stimulation and the practical policy implications.
Develop and implement a parallel media communications plan, based on social science research, to enhance public awareness of social science methods, findings, and impacts. In other words, social and behavioral scientists need to use what we know to communicate how we know and why it matters. If a behavioral and social science method or finding does or could change the world for the better, but no one who makes policy knows that, why would policymakers support the science that produced it to begin with?
In response to a general lack of awareness among policymakers of the many potential and actual contributions of the behavioral and social sciences and a devaluing of social science research, an independent funding source should develop a communications campaign directed toward increasing awareness, understanding, and support for the social and behavioral sciences. This campaign would incorporate principles and findings of the behavioral and social sciences to maximize effectiveness. The initial focus would be on policymakers inside the Washington Beltway, but consideration would be given to a broader public market so that average citizens will be better informed about the social and behavioral sciences and their value.
As one example of how such a campaign might be developed using behavioral science principles, the “Trans-theoretical” or “Stages of Change” model suggests that there may be merit to an initial messaging strategy designed to move people who may be at the pre-contemplation level to contemplation of the methods and benefits of the social and behavioral sciences. Based on this model, several striking examples of proven applications could be highlighted, with the initial focus not on the specific findings, but on the methodologies and disciplines that produced them.
For instance, a media campaign might use the following: “What is the best treatment for PTSD and how effective is it? How do we know?”
Another message, perhaps at one of the Metro stations leading to National Airport, might include an image of the 1982 jet crash in the Potomac with a caption: “This tragedy has not been repeated, and your air travel is much safer today because of fundamental research. What changed?”
As a third example, “With no change to the tax code and no new government expenditures or mandates, millions of Americans are saving billions of dollars more in their retirement accounts. Who figured out how to do that?”
Ideally, these or other messages should be tested empirically and compared with other messages and media. If they are shown to be effective, they would be deployed strategically through media and locations identified by research and with expertise and evidence from communications firms.
As part of a comprehensive strategy, these messages could be used to drive interest to further information, or they might be used as part of a series in which the first messages move from pre-contemplation to contemplation, with subsequent messages moving through other stages of change toward the desired end of greater awareness and support for social science research. The goal would be to craft messages that reach out to different audiences in different ways so that each can, in its own way, recognize that the social and behavioral sciences can help contribute to better outcomes, financial savings, and more effective and efficient policy.
Our disciplines have established a body of methods, findings, and knowledge that is directly applicable to a host of public policy areas. The task before us now is to turn that “applicable” into “applied” in ways that benefit our society and demonstrate the value of our disciplines to policymakers. Whereas this article has suggested several ways to go about that endeavor, there are undoubtedly many other possibilities. What matters most is that we consider a number of options and then put in place a strategic plan to implement the initiatives that seem most promising.