This book is a worthy launch to an important research initiative connecting members of Congress with their constituents. Michael Neblo, Kevin Esterling and David Lazer report on a series of “deliberative town halls” held online with diverse samples of participants in dialogue with sitting members of Congress. I will note some of the merits of this important initiative as well as some of its limitations. I offer the latter in the spirit of suggestions for avenues that might expand and deepen the effort.
In their current form, congressional “town halls” are self-selected and unrepresentative. They are often captured by intense groups who show up not to deliberate but to mobilize their support or opposition. Citizens are not coming to listen to one another but to express their views, often as loudly as possible, to their member of Congress. Neblo, Esterling, and Lazer argue for a different model, for what they call “directly representative institutions” in which legislators engage with their constituents “directly as citizens rather than only as voters, campaign contributors, or members of interest groups” (p. 13). Most importantly, their approach engages random samples, not self-selected samples in substantive deliberation. While such events might look like the current “congressional town halls,” they are fundamentally different in structure and in the promise they hold for democracy.
There are few opportunities for members of Congress to get thoughtful and representative input from the public. Public opinion polls, we are told, “do not reliably track people’s considered views on policy when informed” (p. 16). Politicians should be interested in these deliberative town halls to restore trust and legitimacy. They might also reveal what V. O. Key called “latent opinion—that is, public opinion that will emerge after the official takes some action” (pp. 21–22). The idea is that deliberation may reveal where opinion may go with more information and experience. The results are more than the public’s current impression of sound bites and headlines. And they are more representative than the intense groups that show up at the conventional town halls.
The reform is proposed as part of a new “deliberative infrastructure” (p. 24), “in which people who disagree over policies exchange reasons about ‘what we should do’ (as opposed to ‘what do I want?’)” (p. 26). The authors convened 19 such events (p. 41) with samples of approximately 30 participants each. Each event had 35 minutes of dialogue, with the citizens submitting text-based questions and the representatives responding via voice. Then the representative left and the citizens continued to discuss for another 25 minutes. Senator Carl Levin (D-MI) also participated in a larger event with 175 deliberators recruited statewide from Michigan. That event was structured just like the House events, except that it did not have the 25-minute follow-up discussion among the citizens (because the group was too large, p. 115).
Much of Politics with the People is organized around five criteria for evaluating these events: 1) inclusion, which requires the representativeness of the samples; 2) the provision of balanced information; 3) the quality of the reasoning; 4) promotion of legitimacy and trust; and 5) whether the process can be scaled (p. 27). By this they mean: “The institution should be scalable so that a meaningful number of constituents can participate, and the process can perceptibly ramify through the larger deliberative system” (p. 48; italics in original). In what follows, I consider each in turn.
First, on representativeness, we are told that the participants in each of the 19 events are good cross sections of the eligible electorate and, in particular, that the events “tended to attract lower income participants and those with less steady employment, which strongly dispels the notion that deliberation only attracts those at the top of the distribution” (p. 65). However, no data on representativeness on the House town halls are presented in the book, so we are left just with general characterizations. Given that one of the criticisms of the conventional town halls is that they engage mostly those who feel most intensely, it would have been interesting to know how all the deliberative town hall participants, in aggregate, compare to all those who did not attend but responded when originally contacted.
Second, on balanced information, the book reprints good and balanced briefing materials on the issues. Further, the authors present data on information gains (based on knowledge questions). Many of the gains occurred in anticipation of the event, or even in discussions outside the event. This is consistent with other deliberative forums and shows, as they note, the merit of just being invited.
Third, the quality of reasoning is evaluated in terms of the Discourse Quality Index (DQI). The authors sensibly focus on the portion of the DQI that measures justified arguments—the offering of reasons and the connections between those reasons and what is advocated. The DQI results are impressive. However, they focus only on statements by the representatives, not on statements by the citizens.
Fourth, the authors report that there were increases in trust and political efficacy among the citizens. But these results are just summarized with no data presented (p. 101). Given the experience of other deliberative projects, these results are certainly plausible. However, it would have been nice to include more detail, even if the results were in an appendix.
Fifth, the authors rely on the one statewide session with Senator Levin from Michigan, which had 175 participants, to argue that the process is scalable. Participants talk to their friends, and thus there is a ripple effect. In addition, the online experience of being in a group of 175 could be compared, they argue, to the online experience of a session with 20,000 (p. 116) or even larger numbers. Note that the participants are not discussing the issues among themselves, they are proposing questions to the member of Congress. Critics might argue that the experience, even at 175, much less 20,000, is beginning to approach something closer to what Bernard Manin calls “audience democracy” than the kind of discussion typically identified with deliberative democracy. As noted earlier, the small group discussion following the 35-minute session with the representative had to be dropped in the session with Senator Levin, as the numbers were too large. Hence, the dilemmas of scaling.
Where might this go? The authors might consider variations in the model to extend and deepen the deliberations and to better scale them.
First, on duration of the deliberations: Thirty-five minutes is very short for deliberative events, many of which typically take place over a long weekend (Deliberative Polls) or multiple weekends (Citizens Assemblies or Citizens Juries).
Second, scale and design of the deliberations: The single shared-meeting format limits the numbers. Many deliberative events break the participants up into small groups (Deliberative Polls, 21st Century Town Meetings, etc.). The small groups formulate questions for competing experts and/or politicians. The number of moderated small groups could be multiplied to accommodate much larger numbers than in the single, shared-meeting design. The groups would gather for plenary sessions in which the questions agreed on in the small groups are asked. By having shared discussion in the groups about what questions are most important, it is possible that the deliberative quality of the questions would be increased.
Third, future studies might foster and study the deliberations of the citizens, not just the representatives. On the current model, only the speech acts of the representatives are coded. But what about the citizens? Is this a case of what the book calls “deliberative persuasion” (p. 85), or is it true deliberation by the citizens? Yes, the representatives offer reasons, but they are almost surely reasons to explicate and defend their own positions. What about the other side? The root of deliberation is “weighing,” and most deliberative advocates think that the process requires weighing competing arguments on issues posing trade-offs. But on this design, the expertise is mostly on the side of the representative. The authors note that there is systematic movement in favor of the positions favored by the representatives (p. 97). Is that an indicator of deliberation or one-sided persuasion? If we knew more about the citizen discussions (if the reasoning were coded), and if the sessions were longer, we could evaluate the balance of the process. A design with many small groups would generate both more data and more in-depth discussion.
Fourth, the mode of communication could be enhanced. Video-based small group discussions with moderators are being used for deliberation by citizens online and would approximate the conditions of face-to-face deliberation (whether conducted via Google Hangouts or Skype or other methods designed for the purpose). These would be especially suited to multiple small groups deliberating on the issues and then generating the questions.
While changing technology and variations in the design could increase the deliberative character of the interactions, the basic point is that this research initiative should expand and become a part of our collective political life. It is a pilot for an important idea—that members of Congress can deliberate about key policy choices with random samples of their constituents. Yes, it can be done. The authors are to be congratulated for starting something important.
