A merican politics is suffering from a near-fatal attraction to direct democracy. Symptoms of this attraction include the proliferation of referenda, particularly in the western states, and the credibility given to Ross Perot's proposal to introduce "electronic town halls" in which television viewers would call in votes on current policy issues.
We have also brought elements of direct democracy into presidential selection by creating a nominating system dominated by the direct primary. Since 1968, when Hubert Humphrey won the Democratic nomination without entering a single primary, the number of states holding primaries has grown from seventeen to thirty-nine; primaries are now the televised battleground where the nomination is effectively decided.
In addition, a near-daily supply of opinion polls, reporting the approval levels of candidates and the popularity of various positions, has given us a system that is far more plebiscitary in its use of direct democracy than textbook analyses of American institutions would suggest. Political scientists from the late V.O. Key, Jr., to Giovanni Sartori have compared the interaction of television and opinion polling to an "echo chamber" in which polls bounce back impressions presented in the mass media. Despite their volatile character, these polls set the terms of political competition and dominate the agenda for public debate.
What's more, the media reports poll results as if they were solid constructions able to support platforms and candidacies. The political landscape is altered beyond recognition when presidential approval ratings drop from 91 percent to 30 percent, as they did between spring 1991 and summer 1992 for President Bush. Or when an enigmatic billionaire is able to climb the polls from nowhere to become, if only briefly, the leading candidate for the presidency without contesting a single election. When General Schwartzkopf was substituted for Ross Perot as a presidential challenger in one poll, he did almost as well, revealing the flimsiness of the public information base on which the Perot challenge rested.
Three central factors television, polling, and the impulse to bring the people directly into the process have given us a thin democracy of stylized impression management. Yet, as I will suggest, these same three factors could be turned to a constructive purpose to give greater substance to our democratic processes.
Couch Potato Democracy
This campaign season has initiated a number of new variations on these themes. Some changes have improved the system in minor ways, while others have only increased the superficiality of our increasingly plebiscitary televised democracy.
The hallmark of the season has been the proliferation of opportunities for citizens to respond to what they see on television. Citizens have been given opportunities to join in the dialogue, to call in for information, to call in questions to live broadcasts, and to participate in town meetings. Supplementing the shrinking soundbite democracy of conventional news coverage, many hours of "talk show democracy" have been broadcast on national television. The advantage has been an increase in the breadth and spontaneity of the televised political dialogue. The disadvantage has been the addition of new pseudo-voices for "we the people" to the campaign process.
Conventional news coverage continues to filter opportunities for the candidates to talk directly with the public. In well-known, parallel studies, Kiku Adato and Daniel C. Hallin showed that the average candidate "soundbite," the period in which a presidential candidate could speak uninterrupted on the evening news, shrank from about forty-two seconds in 1968 to about nine seconds in 1988. Recent studies by the Center for Media and Public Affairs show that during the 1992 primary season, this shrinkage continued. The average candidate soundbite has now diminished to 7.3 seconds.
The move back to a more extended discourse would depend on the interaction of network norms of coverage and candidate calculations. CBS announced that it would attempt to counteract this trend with a guaranteed minimum length for soundbites of thirty seconds, but the policy has produced controversy because in practice, it ends up omitting candidate statements. Meanwhile, the candidates have learned how to speak in nine-second bites to get on the news. So for now the effective political discourse reaching the mass public is mostly the shrinking soundbite, a medium that reduces political debate to messages worthy of bumper stickers or fortune cookies.
A major factor counteracting this trend, at least during the primary season, was the loosely coordinated effort to broadcast debates. From December 15, 1991 to March 15, 1992, eleven debates were televised nationally. While these debates offered a substantive contribution for those who listened, they did not transform the effective political discourse reaching the mass public.
By and large, the public learned about the candidates from the soundbites the evening news produced rather than from watching the debates themselves. Ratings of the major network broadcasts ranged from a low of 2.1 for the climactic CNN/League of Women Voters debate before the New Hampshire primary to a high of 5.5 for the ABC debate March 5 in its "Nightline" slot. (Each rating point is a percentage of the 92.1 million television households and represents 921,000 households.) These ratings put the debates squarely in the bottom ninth of network programming in their respective weeks. Unlike the final presidential debates of the general elections, primary debates, when candidate selection is a live issue, have not attracted large audiences. (In contrast, the 1988 presidential debates in the general election were estimated to have drawn more than 160 million viewers.)
Notably, when these debates were turned into soundbites and newspaper stories, they were reported mostly in terms of whatever conflict, controversy, or confrontation they generated. The first debate, on NBC, was most notable for the flap over Jerry Brown's advertising his 800 number on the air and for Harkin holding up a dollar bill to symbolize the value of the middle-class tax cut. The CNN debate in New Hampshire was reported mostly in terms of the fire Paul Tsongas drew for his support for nuclear power. The Denver debate is remembered for Tsongas's response to Clinton that while he might not be "perfect," at least he is "honest." The Dallas debate was notable for Clinton's rejoinder to Brown that he should "chill out," a phrase that Hillary later took credit for in the press.
The debates were only intermittently enlightening (with the possible exception of the MacNeil/Lehrer debate on PBS), even for the few citizens who watched them. But the greater damage may have been that to the extent the debates did reach the public, they came as sensational, recycled soundbites. The debates, then, while a noble effort, did not transform the effective political discourse reaching the public.
Another noble effort worth mentioning is the opportunity the Discovery Channel provided to all the major candidates to communicate for twenty minutes each directly to the public, without the filters of pundits or editors. Unfortunately, the broadcast achieved a rating of only about 1.5, reaching about 1.2 million of the nation's television households. Because the format produced neither drama nor conflict, it was not widely reported and produced very few soundbites.
Talk, Talk, Talk
A major departure this campaign season has been the use of talk shows. Perot announced his possible candidacy on "Larry King Live" on CNN. Both Clinton and Perot fielded questions from viewer call-ins on the "Today Show" and "CBS This Morning." Perot held a two-hour "Nightline" town meeting on ABC while Clinton held a ninety-minute town meeting on MTV. In one of the most ambitious talk show forays into politics, Clinton and Gore appeared for two hours on "CBS This Morning" with questions from a studio audience, live satellite connections to remote locations around the country, and questions collected from viewer letters. This format was successful in combining viewer input from around the country with follow-up questions from the talk show hosts so as to yield a more sustained dialogue.
President Bush tried his hand at the town meeting format by talking with hand-picked visitors to the White House on "CBS This Morning." Bush argued that town meetings were nothing new. He had, after all, campaigned in 1980 and 1988 with broadcasts of voter forums entitled "Ask George Bush." However, those forums were scripted, a practice that got Bush in trouble more recently when he complained to a live mike last December that he had been asked the questions in the wrong order in an ostensibly spontaneous question-and-answer session in a teleconference to a California teacher's convention.
The main benefit of talk show discussions with ordinary citizens has been the injection of spontaneity into the political dialogue. Talk shows permit us to examine candidates for longer than a shrinking soundbite and they do so under conditions that may produce spontaneous exchange. Instead of press questions focusing on the horse race and political strategy, ordinary citizens have tended to raise more substantive questions about the economy, health care, and other issues that touch their lives directly.
Still, while citizen questions on the talk shows have been notable for their substance (if not for their follow-ups), host questions retain a whiff of the sensationalism that is part of the regular agenda of such shows. Talk shows tend to treat politicians as just another group of celebrities. Hence it was Phil Donahue who relentlessly pursued Clinton about Gennifer Flowers and draft issues before the New York primary, only to be upbraided by a member of his studio audience who wanted substantive questions. And when Stone Phillips interviewed President Bush on "Dateline NBC," the president threatened to cut the interview short rather than face a question about alleged adultery, a question Phillips was told he should be "ashamed" to ask in the inner sanctum of the Oval Office. Conversely, talk show democracy blurs the distinction between politician and media figure by allowing such figures as Pat Buchanan, Jesse Jackson, John Sununu, and David Gergen to alternate between roles.
O f course, the talk show or town meeting ideals hold out the promise of even more radical departures from conventional political coverage. In addition to lengthier, more spontaneous dialogue, television viewers may also see their reactions tabulated, in some process that appears analogous to voting. This was, of course, the basic idea behind Perot's proposal for the "electronic town hall." As Perot described it, major issues, such as the budget deficit or health insurance, would be explained on the air "in depth, not in soundbites." Viewers would then call an 800 number "by Congressional district." This feedback, Perot promised, would be tallied and used to get the White House and Congress "dancing together like Fred Astaire and Ginger Rogers."
Something very close to Perot's proposal briefly saw the light of day last January, after the President's State of the Union address. In a pilot for a possible series called "America on the Line," CBS tabulated about 300,000 viewer responses to questions about the President's speech and the state of the union. However, the viewers who decided to phone in their responses to the CBS program presented a distorted picture of public opinion, at least when contrasted with poll results from a representative sample asked the same questions (and reported by CBS). For example, 53 percent of "America on the Line" respondents said they were "worse off" now than a year ago, while only 32 percent of the representative sample said so. Only 18 percent of "America on the Line" respondents reported they were in the "same" economic situation as a year ago, while 44 percent of the representative sample reported being "the same."
This kind of electronic town hall has two fundamental defects it is neither representative nor deliberative. It is not representative because the sample is self-selected. Instead of being chosen through the methods of modern survey research, through a random statistical process, viewers at home select themselves by their decisions to call in.
Viewers calling in to an 800 number constitute what Norman Bradburn, Director of the National Opinion Research Center at the University of Chicago, has called a SLOP a self-selected listener opinion poll. A SLOP played a role in distorting media coverage of the Carter/Reagan Presidential debate in 1980, when ABC used viewer call-ins (in this case with charges for calling a 900 number) to declare Reagan an instant two-to-one winner, as compared with random samples that viewed the debate as a close contest. Like the Literary Digest fiasco of 1936, which predicted a landslide for Alf Landon over Franklin Roosevelt, self-selected samples draw disproportionately from those who feel strongly enough to call. Large numbers do not by themselves offer any indication that the self-selecting viewers represent public opinion. CBS has reported that over 24 million calls were attempted to "America On the Line," but far more accurate results could have been achieved from a carefully constructed random sample of several hundred.
Neither is the electronic town hall deliberative. Citizens are expected to phone in their reactions off-the-cuff, have little opportunity for debate or for consideration of alternative views, and often they have little background information. This points to one of the biggest quandaries of direct democracy in a large nation-state: the belief that there is little reason to bother learning about candidates and issues because it is so easy to calculate that one vote is unimportant and will have little effect on the outcome.
One of the more inventive departures this election season has been directed at this problem of underinformed voters. The Center for National Independence in Politics has launched "Project Vote Smart" providing an 800 number, advertised on CNN, which citizens can call to get nonpartisan information about candidate positions. The same device, advertising a number for citizen information on television, has been employed by the notorious Floyd Brown (of Willie Horton ad fame). In a new twist on negative campaigning, Brown offers callers the chance to hear tapes of conversations between Gennifer Flowers and Bill Clinton.
In either case, the problem of individual incentives for information remains. Primaries, referenda, and opinion polls have brought power directly to the people, but under conditions where the people have little motivation to think about the power they are supposed to exercise.
I s there some way of getting over this problem of effectively motivating ordinary citizens to acquire political information and deliberate about it? Some recent experiments, both in this country and in Britain, suggest a new way of combining television and survey research. In five different British elections, Granada television took a random sample of 500 citizens from a benchmark constituency in northern England. After discussing the issues for a couple of weeks, these 500 citizens were transported to London, where they appeared in a televised question-and-answer session with the three party leaders, to be broadcast a few days before the British general election.
Unlike American "town meetings," which have employed either haphazard collections of people in a studio audience or viewer call-ins, the Granada 500 group was a statistically representative sample that was also prepared on the issues. However, the Granada 500 took no votes, made no decisions. It simply offered a new kind of forum for questioning the candidates, forcing them to confront issues of direct relevance to ordinary citizens.
The Jefferson Institute in Minnesota has been experimenting with "citizens' juries" that question candidates and deliberate about their positions. In cooperation with the League of Women Voters, it has held such juries in Minnesota (in a Minneapolis mayoral election and on various policy issues), and it plans to hold a similar forum, in cooperation with local television stations in Pittsburgh and Philadelphia, in the Pennsylvania Senate race this fall. In contrast to the Granada 500, citizens' juries vote on the candidates. However, it is worth noting that with juries of eighteen people, they cannot be statistically representative of the entire population, as a full-scale random sample would be.
Both the citizens' jury and Granada 500 use randomly chosen citizens who are giv en the opportunity to deliberate about public policy and who are, in addition, offered the chance to question candidates on television. These elements are both included in my own proposal for a "deliberative opinion poll" at the start of the primary season on national television. Instead of a random sample of a benchmark constituency, as with the Granada 500, or a group the size of a jury, as with the Jefferson Institute, I have proposed that we take a full-scale national random sample of 600 people and transport them to a single site where they can question the presidential candidates in person on national television. Even if the viewing audience were limited to the comparatively small audience of primary debates, the results of such a deliberative poll at the start of the primary season would receive enough coverage to play a major role in launching candidacies and issues so as to reform the "invisible primary" the initial period of the nomination struggle that has come to largely determine the nominee in an increasingly front-loaded presidential selection system. Hence the timing of such a televised deliberative poll could give it an influence far beyond the 600 delegates and far beyond its viewing audience.
A deliberative poll harnesses to a constructive purpose the same factors that have, thus far, only trivialized our mass democracy television, polling, and the impulse to bring the people directly into the process. A deliberative poll brings the people into the process but in a statistically controlled way; it uses the techniques of polling but under conditions where the responses represent more than echoes of shrinking soundbites; it uses television to amplify deliberation rather than to disseminate canned material or advertising.
An ordinary poll models what the public thinks, however little the public knows or pays attention. A deliberative poll, by contrast, models what the public would think if it had more opportunity to think about the questions, more information about the issues, and opportunity to question candidates extensively. Like the advent of talk show democracy, a deliberative poll would add depth and spontaneity to the dialogue. However, it would do so with a carefully constructed sample that would have every incentive to pay attention and take the process seriously. Members of the sample, unlike ordinary citizens viewing the campaign at home, would be motivated to pay attention because they would appear on national television with the candidates. The problem of incentives for information and deliberation is solved for members of the sample, and the results are then amplified by the television broadcast.
I offer this proposal not as a panacea but as a televised demonstration of an alternative democratic model. Recall that the direct primary is, itself, a relatively recent innovation fostered primarily by the Progressives early in this century and, in a second wave of reform, by the McGovern-Fraser Commission reacting to the fiasco of the 1968 Democratic nomination struggle. Experimentation with alternative models is clearly called for, even if there is no single proposal that can, by itself, guarantee a credible system.
Last year I joined with WETA, the Washington PBS station, to attempt to mount a televised deliberative opinion poll, the "National Issues Convention" to be broadcast for three days over PBS at the start of the primary season in January of 1992. While the event was cancelled for 1992 because it ran into funding difficulties, WETA has since joined with all ten of the nation's presidential libraries to sponsor the event in 1996 at the LBJ Library at the University of Texas at Austin. With luck and proper funding, the next primary season will begin with a deliberative opinion poll on national television.
To be sure, a televised deliberative poll or kindred forms of policy jury are neither a new form of direct democracy nor a substitute for the more usual forms of representative democracy. A functioning polity requires high voter turnout, responsible parties, a well-informed electorate, and competent elected officials whose job is to deliberate day in and day out.
However, our elected legislators are truly representative only to the extent that the voters pay attention. Otherwise, they exist in an echo chamber of their own. In that sense, a deliberative poll can serve as a kind of role model. It can demonstrate the capacity of ordinary citizens to appreciate the complexity of pressing issues; it can give elected representatives a more authentic form of feedback; and it can shame pollsters into resisting the temptation to oversimplify. Perhaps it can even help restore public interest in public issues. We shouldn't expect new forms of public deliberation to replace the ordinary mechanisms of democracy. Rather, we should appreciate their potential to infuse representative democracy with new life.
