Legal Affairs

Current Issue


printer friendly
email this article
letter to the editor

space space space

January|February 2004
The Big Kozinski By Emily Bazelon
Righting the Ship of Democracy By Bruce Ackerman and James Fishkin
Smooth Sailing By Richard Posner
The Wrong Tack By Arthur Lupia

The Wrong Tack

Who's to say that people make better decisions in groups than they do on their own?

By Arthur Lupia

Bruce Ackerman and James Fishkin are to be commended for the energy that they bring to their cause. They take a notion of deliberation originating in the often arcane field of 20th-century political philosophy and propose to make it the basis for a large-scale public policy. This is no small feat, and these are not waters in which timid intellects wade. I would be happy if their proposal could improve civic competence. But there are reasons for doubt.

The problems with Deliberation Day have little to do with the authors' specific proposal, though its grand scale raises the stakes for deliberation's advocates and makes it essential that they address the problems with the concept. The Achilles' heel of Deliberation Day, like so many philosophical proposals to make deliberation a remedy for perceived civic incompetence, is the fact that it is based on idealized versions of human nature rather than on basic scientific facts about how people think about politics and what, if anything, they learn from interactions with others.

DELIBERATION DAY WILL SURELY ACCOMPLISH SOMETHING. It will expose some people to new ideas and perspectives. But mere exposure does not guarantee an improved electorate. What matters is whether participants learn from each other and, if so, whether what they learn enhances their civic competence.

Ackerman and Fishkin argue that deliberation is likely to elevate the "force of the better argument." Decades of research on human cognition and decision making, however, indicate that deliberation increases civic competence only if some difficult conditions are satisfied.

By civic competence, I mean a citizen's ability to accomplish well-defined tasks in her role as voter, juror, or legislator. If deliberation is to increase civic competence, it must cause people to think about politics in very specific ways; not just any change will do. Suppose that we can define a competent vote as the one that a person would cast if she knew where a specific set of candidates stood with respect to a well-defined list of major policy debates. For deliberation to increase her competence, she must not be voting competently initially. Deliberating must cause her to do so. If knowing a candidate's political party leads her to draw the correct conclusion about which candidate takes the positions she prefers, then deliberation cannot increase the competence of her vote.

How, then, can deliberation affect civic competence? Many deliberation advocates describe deliberation as if it is a place where ideas travel from one mind to another in original form—as if listeners interpret ideas whole, exactly as speakers convey them. In human communication, however, all but the simplest statements and stimuli are parsed. People pay attention to only a tiny fraction of the information available to them, and they can later recall only a tiny fraction of the things to which they paid attention.

The physical attributes of working (short-term) memory force us to ignore almost everything around us. To get our attention, a statement must fend off competitors such as our preoccupation with certain prior or future events, the simultaneous actions or statements of others, and even the color of the wallpaper. For the statement to deliver a specific idea, the target audience must also pay attention to it in a very specific way.

For example, if someone says, "Colin Powell contends that the Iraqis have not disarmed," audience members must parse the statement in a certain way and combine it with previously held views about the relationship between Powell and the Iraqis to make sense of it. If the target audience focuses exclusively on one aspect of the statement—say they hear Colin Powell and think instead about a foreign policy view of his unrelated to Iraq and the Iraqis—then exposure to the statement will not increase the audience's competence.

THE KEY TO INCREASING COMPETENCE is not putting people in a place where they and others get to state their opinions; it is putting them in situations where they are motivated to pay attention to information that will help them make competent choices. If deliberation advocates focused on the latter as well as the former, they would find that science provides mixed messages about the relationship between deliberation and competence. Many studies reveal that some group interactions actually decrease competence; one example is the organizational malady known as "groupthink." Other deliberative interaction, as the Harvard political science professor Jane Mansbridge points out, "accentuates rather than redressing the disadvantage of those with the least power in society."

Together, such findings reveal that it's wrong to presume that the more knowledgeable people in a deliberative environment are necessarily the more influential—which, in turn, undermines key claims about deliberation's benefits. Particularly imperiled is the claim that proposed deliberative mechanisms necessarily or even frequently elevate the "force of the better argument."

Other scientific research reveals deeper problems. Memory is complicated, and participants in a deliberative democracy session aren't going to remember everything the authors want them to. Even if a piece of information is attended to, it can only increase competence if it is processed in a way that leaves a unique cognitive legacy in long-term memory, or LTM. If it is not processed in this way, it is, from a cognitive perspective, gone forever.

LTM depends on chemical reactions within and across specialized cells in the brain, with a particular reliance on each cell structure's "activation potential." Activation potentials correspond to probabilities of recalling things once they have been noticed; what we usually call learning involves changing these activation potentials. The physical embodiment of learning that smoking is highly correlated with lung cancer, for example, is a change in activation potentials that makes you more likely to associate pain and death with smoking.

Two facts are important here for understanding the impacts of deliberation. First, if a speaker's attempt to increase another person's competence does not lead to a change in that person's activation potentials, then the attempt does not increase competence. Second, not every change in activation potentials is sufficient to increase competence; the change must be significant enough to help someone accomplish a task that couldn't be done before.

This suggests that participants may not walk away from open deliberation remembering what Ackerman and Fishkin might want them to remember. To see why, think about the most important events in your life: marriage, the birth of a child, completion of personal goals, and great disappointments. Chances are that most of these events took place over a series of hours or days. How much do you remember about them? Even if you focus hard, you can probably generate only a few seconds of distinct memories, tiny fragments of these critical events. Recall from LTM is not like bringing up an old document on your computer that comes back exactly the way you saved it. There is significant decay. Deliberation organizers who ignore how citizens think about politics will be surprised to learn that they have very little control over what participants will remember. In LTM, "the better argument" (if one exists) can easily be crowded out by something else, such as an outrageous statement or gossip conveyed between sessions.

WHAT WILL DELIBERATION DAY ACCOMPLISH? Ackerman and Fishkin offer evidence from Fishkin's previous deliberative opinion polls, which sampled groups of several hundred citizens before and after they deliberated for two days. They treat large changes as evidence that participants were "better informed." This is problematic. The changes need not signal increased competence. Instead, they may be a sign of decreased competence; someone taking part might have changed an answer because he misunderstood information presented or because he became confused by a complicated argument. A closer inspection of the authors' data will likely reveal participants' opinions moving in opposite directions and many not moving at all.

What must we conclude about the citizens whose views either did not change or moved in the opposite direction of those whom the authors label "better informed"? Perhaps they learned important things that drove them even further from people who disagree with them. Ackerman and Fishkin do not report the extent to which this happens; nor do they comment on how such occurrences would affect the value of their proposal.

The authors' treatment of heuristic decision making—basing decisions on cues such as a candidate's partisanship or an automobile's brand name—is similarly flawed. They note that "poorly informed voters already use various simplifying devices to approximate the same conclusions they would reach if they were well informed." But since such findings imply that voters can be competent without deliberating, which undermines the case for spending billions on Deliberation Day, the authors attempt to dismiss the research. They claim that the results rest primarily "on statistical models developed from survey data."

This claim is false. Almost all of the evidence on this subject is empirical evidence from experiments. The hundreds, if not thousands, of experiments on heuristics usage (described in books such as Gerd Gigerenzer and Peter M. Todd's Simple Heuristics That Make Us Smart) go into far greater detail than any deliberative poll about how people think and learn from others. Ironically, the only evidence that Ackerman and Fishkin present to the contrary, a study by the political scientist Larry Bartels, rests exclusively on statistical models developed from survey data, the same technique that they reject above. If we adopt their standard of evidence on this issue, then they have no evidence.

This is not to say that heuristic decision making is a panacea. Like deliberation, sometimes it works—and sometimes it doesn't. The key to increasing civic competence effectively and efficiently is understanding the conditions under which each method yields competent choices. While Deliberation Day's authors presume that deliberation improves citizens' competence, they never establish empirically that its effect on civic competence exceeds that of the heuristic decision making in which we all regularly engage.

DO WE NEED DELIBERATION DAY? After all, informative news programs, well-crafted political commercials, conversations with friends, and newspaper editorials can also change opinions. Underlying Ackerman and Fishkin's argument is the presumption that the kind of knowledge people gain by deliberating with strangers is superior to other kinds of knowledge. If it isn't, much of the rationale for Deliberation Day disappears.

Given the importance of this question, it is surprising that the authors never compare deliberative polls to other means of increasing civic competence, such as encouraging target audiences who need more information to read a book or an article or visit a well-designed website. These and other strategies may outperform Deliberation Day or have an effect so similar that they call into question the day's bang for the bucks and return on the man-hours it requires.

Similarly, Ackerman and Fishkin cite a desire to make people more like "ideal citizens." They are annoyed that so many citizens learn about politics from the mass media, especially television. Like other deliberation advocates, they make no distinction between people who learn from sources like C-SPAN and The Economist and those who get their political news from Jay Leno.

Ackerman and Fishkin repeatedly portray citizens as "ignorant" and "selfish" because they do not pay attention to politics as the authors do. Like other deliberation advocates, they dismiss or underemphasize the importance of what citizens do instead of deliberating about politics. While some people engage in activities that may have limited social value (in their book, the authors mention shopping mall trips and ski weekends), millions of others are raising families, helping neighbors, counseling friends and co-workers, and engaging in a wide range of socially beneficial activities. It is presumptuous to conclude that society will benefit by taking people away from such activities and inducing them to deliberate about politics with strangers.

PHILOSOPHERS, CIVIC LEADERS, AND SCIENTISTS ALIKE recognize the importance of a competent citizenry in a properly functioning democracy. Most also recognize that civic competence can be improved. I agree. To that end, I ask that people working in this field stop for a moment and contemplate their own competence for the task at hand. It is ironic that a movement founded on the premise that citizens don't know enough to vote competently thinks it can change how citizens deliberate without really understanding how they think. Science shows that deliberation can lead to thoughts that increase citizens' competence—and to thoughts that make citizens less competent. In many cases, deliberation may not change thoughts at all.

Ignoring these facts may have dangerous consequences. Deliberation advocates who insist that citizens learn a specific set of facts or engage in a particular set of practices may not really understand how these actions affect the target audience. In many cases, advocates presume that the practices they prefer are better for others. In other words, elitism fills the void left by advocates' inattention to basic facts about how people think. The result is that advocates who are overconfident in their ability to change human beliefs and behavior end up imposing on others values and programs that favor elite worldviews yet make the target audience worse off and that fail to improve civic competence. Outcomes like these are tragic, wasting resources that could have been used for activities that would more accurately diagnose and remedy problems caused by a lack of civic competence.

From those who ask that America change its democratic practices or that citizens change their political ways, we should demand a demonstration of their own competence in increasing citizens' democratic skills. We should ask how the conditions under which their approaches can increase civic competence. The deliberative movement's current foundations are insufficient and dangerous when proposed as the foundation for large-scale public projects. At a minimum, anyone who supports or agrees to participate in such mechanisms should put faith only in those people who can successfully address the challenges posed above.

Arthur Lupia is a professor of political science and senior research scientist at the Institute for Social Research at the University of Michigan. His books include The Democratic Dilemma: Can Citizens Learn What They Need to Know?

printer friendly email this article letter to the editor reprint premissions
space space space

Contact Us