The Enlightened Citizen
Is Made, Not Born
Can deliberative polling cure democracy?
One of my contributions to this magazine’s State of the State page—a reflection on the possibilities inherent in deliberative polling as an alternative, or rather enhancement of, democracy. Probably more interesting for me to write than it was for most people to read. Which is unfortunate.
Bernard Shaw—the playwright, not the CNN talking head—noted that in a democracy every person was expected to be his own Solon and his own Plato.
"Why not," he asked, "every man his own Shakespeare and his own Einstein?"
A good question that no lawmaker dares to answer. To assume that every citizen is capable of government, simply by virtue of being a citizen, is as absurd as to assume that" every legislator, simply by virtue of holding office, is capable of wisdom. The broader American voting public does not grasp the concept, much less the tedious details, of such essential government programs as Social Security, foreign aid, trade or property taxation. It is useful to recall that its members enjoy the right to vote not because they are capable of governing but because they are—occasionally—capable of revolution.
Traditional voter education in the form of pamphlets, issues forums and League of Women Voters show-and- tells err in assuming that voters lack only facts. That they do, and the most basic facts; the number of Illinoisans who can recall the names of both their U.S. senators, it seems fair to say, is much smaller than the number of people who vote to select them. But voters' ignorance owes not merely to missing or errant facts per se but to the inability to make sense of the facts they have, thanks to their shaky grasp of history, economics, even basic arithmetic. It is not the failure of our schools to teach civics that is undoing democracy, but their failure to teach everything else.
Can it be done better? During the 1970s, a number of worried democrats experimented with methods by which voters might learn the essential facts about selected public issues and ponder their meaning in context. Details varied, but the basic plot goes like this: Invited volunteers gather for a period of hours or days to discuss a particular issue face-to-face. Their conversations are based on information presented to them in issue-pamphlets prepared by the organizer, usually an academic or a do-gooder foundation; occasionally this printed information is augmented by testimony from experts imported for the purpose. After the palaver, a group opinion is rendered, the results of which are announced to reporters who say, "That's interesting" and go back to speculating whether the collapse of the ringgit will affect the primary campaign for secretary of state.
These experiments in democratic group therapy have taken several forms over the years, from "people's budget" exercises to the Kettering Foundation's National Issues Forums. The best known is probably the "deliberative poll" as conceived in 1988 by Dr. James Fishkin of the University of Texas at Austin.
The point of these exercises is not merely to gauge opinion but to deepen it. In the process, participants almost always discover what the brighter politician, bureaucrat, and, yes, even reporter knows, which is that most issues do not look the same up close as they do from the couch in the TV room, if only in being more complicated.
Not surprising, the more the average earnest citizen learns about an issue, the more likely she is to conclude that her previous opinions were a bit undercooked. In the summer of 1993, Larry Hansen, now with the Joyce Foundation in Chicago, then at George Washington University's Democracy Agenda Project, organized a series of chinwags on the topic of campaign reform. Called "Heartland Voices," the project gathered citizens-at-large for deliberations at 19 sites across the Midwest, including five in Illinois. Hansen recalls that many people who favored term limits going into the exercise came out of it opposing them, once they had a chance to think about it.
It has been an item of faith since the Progressive era that if all voters took such care, politics and government would not be quite the dog's breakfast that it has become. Alas, all voters never will. Volunteer participants are hardly representative, almost always being what Hansen calls "engaged, informed citizens." This self-selection bias may enhance the intelligence of the results, but it calls into question their political legitimacy.
One way to fix this is to select participants who constitute a statistical random sample of citizens. The Jefferson Center for New Democratic Processes uses random telephone dialing to select its "policy juries," panels of 24 citizens who are seated to hear evidence and pass verdicts on issues from national health care to housing for the poor. Fishkin's deliberative polls pick a representative sample of several hundred of the body politic using the University of Chicago's National Opinion Research Center.
Such random samples pull in minorities, women, the poor and the alienated, few of whom usually show up for such soirees. Of course, widening any franchise in this way usually achieves legitimacy at the expense of good sense. Think of baseball All-Star balloting, or the citizens' initiative, in which voters choose their laws and not just their lawmakers. Unfortunately the initiative is a tool to create a more perfect democracy, not a more perfect government. In the American West, millions have proved what an Illinoisan never would have believed otherwise, which is that there is a class of people—voters—even dumber " than legislators.
What is interesting—indeed, nearly miraculous—about a well-conducted deliberative poll is that there occurs little drop-off in the quality of decision-making when the participant pool is expanded from elites to the hoi polloi. Almost any citizen who troubles to learn about complicated issues and to refine his views in conversations in small groups can come up with sophisticated solutions to the policy puzzles of the day. When done by a random sample of voters, the opinions thus arrived at do not quite reflect the views of the larger polity the sample represents. As Fishkin puts it, they reflect the views the larger polity would hold if the latter's members troubled to do their homework.
The enlightened citizen is made, in short, and not born. This should reassure skeptics on the left who fret that the deliberative poll is a sort of re-education camp, set up to brain-wash regular guys into parroting the views of our educated elites. But rather than co-opt the views of the working class and poor, a properly run poll would finally give voice to them. The problem of democracy in the '90s is not that the poor and working class do not vote in numbers, but that the world has grown so complex that few among them know enough about where their interests lie to cast their votes to any effect.
As Fishkin has explained, a deliberative poll can do more than merely describe or predict public opinion. Ideally, such polls should have a "recommending force"—in other words, prescribe as well as describe. Unfortunately, there can never be enough minds changed this way to directly make a difference in even local elections. (No one talks seriously about involving all voters in what is a time-consuming and expensive process.) Nor is there evidence that voters at large consider the results of their neighbors' deliberations as seriously as they consider, say, the views of talk show hosts.
Might deliberative polls lead to deliberative decision-making? Hawaii has experimented with such techniques to gauge public opinion on school issues, and they are used in Davis, California, to test public views on urban planning. In Texas, the state's Public Utility Commission now accepts deliberative polls conducted on behalf of utility companies as satisfying its requirement that customers be "consulted" about how those companies ought to meet customers' future electricity needs—in effect substituting for public testimony.
The Texas experiment suggests how deliberative polling may be institutionalized as a means to convey the public will to decision-makers. Joyce's Hansen says, "I've thought about what might be done in a place like Chicago, where you might have citizen assemblies across the city—maybe 25, probably organized on a multiward basis—that gather several times a year to take up public policy matters that confront the city, such as reviewing the city budget."
Good thinking, but using deliberative polling merely as an alternative to public testimony and/or public comment wastes much of the technique's promise. Why not use it as a surrogate for lawmakers at various levels of government? The problem with representative government in our age is not that it is not representative. The problem is that what it represents is the voters' ignorance, their hasty judgments, their credulity. Making the deliberative poll the centerpiece of the democratic process would be cheaper and fairer, and promises more sensible results at every level of civic endeavor.
What? You ask. A bunch of unqualified people the voters don't even know, empowered to make decisions on the voters' behalf? It's no worse than the system we have now. Some complain that randomly selected poll participants will not be typical of voters. Indeed they would not, but neither is the voting public, especially in primaries and single-issue referenda.
There are other quibbles. That briefing materials will be biased. (Unlike, say, political ads?) That publicity about the process will influence the outcomes. (Unlike a campaign, maybe?) That delegates will be swayed by other members of the project. (People compromising their positions to oblige the interests of other members of the community? How, well... democratic.)
Democracy does not require that citizens govern, only that citizens freely consent to be governed. Problems don't arise when decisions for all are made by too few, only when they are made badly by the too few. ■