POLITICAL CONVERSATION AS A PERSONAL SHORTCUT

Despite their drawbacks, partisan cues (and ideological cues) are appealing, because they give voters a way to make seemingly sophisticated choices without tremendous effort. What voters need is a more reliable cognitive shortcut to use during campaigns. If voters banded together into small groups, they might be able to pool their limited knowledge into relatively nuanced judgments. Like-minded citizens could compare their candidate preferences, and citizens with divergent views could consider one another's viewpoints. Even relatively passive and uninterested voters could use the summary judgments of these localized reference groups to make voting decisions. Such a system of information distribution, deliberation, and collective decision making could not only improve the quality of individual voting choices but also the likelihood of successful collective rejection of unrepresentative elected officials.

Many voters do, in fact, make "collective" judgments, whether as a couple, a family, or a group of friends.^[48] National surveys conducted in 1981 and 1990 found that roughly 70 percent of Americans "often" or at least "sometimes" discuss political matters when they get together with friends.^[49] One study of voters in the vicinity of Albany, New York, prior to the 1998 presidential election found that 88 percent had discussed the election in the past week. On average, people reported talking about the campaign with seven or eight people during that week.^[50]

The frequency with which spouses, families, and housemates vote together has led many political campaigns to target voters "by household." When a precinct walker or a telephone survey interviewer learns that one adult in a household is favorable toward a particular candidate, the campaign then attributes that preference to all voters in that residence. Campaigns are wise to aggregate voters within households: families do tend to vote together, because there is tremendous political influence within typical American families.^[51]

Do these social networks of political influence result in more informed candidate evaluations? Perhaps campaign conversations can make what Zaller calls an "unsophisticated" voter into a sophisticated one. Recall that in Zaller's model of public opinion formation, a sophisticated voter is distinguished by the ability to fllter out messages that are contrary to his or her preexisting ideology. Regular political conversations between sophisticated voters and less sophisticated ones permit the latter to substitute a social sieve for a cognitive fllter of massmediated messages.^[52]

Research suggests that this portrait is only half-true. It is certainly the case that much political conversation flows directly from the media: a recent focus-group study found that more than a quarter of all statements in an open-ended discussion had direct or indirect references to both factual news programming and entertainment media. A study of the 1992 Pittsburgh newspaper strike also demonstrates a relationship between mass media coverage of elections and the quantity of political conversation: the strike "limited voters' exposure to media reports concerning the local U.S. House races, and interpersonal discussion of those elections declined in response.^[53]

Though conversations do transmit media messages, the interpersonal relay messengers are not necessarily the most sophisticated voters. The most careful modern study of political conversation suggests that the people who most influence their fellow citizens are distinguished, not by their political knowledge and expertise, but rather by their mere interest in politics. Even if the most influential political conversants were politically sophisticated, the value of their influence is questionable, because they are noted only for their ability to fllter out ideologically dissimilar views. Political deliberation requires the consideration of alternative views, but research suggests that the most influential discussants are those who accurately understand that their conversation partners share views the same as their own.^[54]

Despite its limitations as a means of reaching sophisticated candidate evaluations and collective electoral action, campaign conversation has at least one redeeming quality. As a modest form of public deliberation, it motivates people to get more involved in elections and, speciflcally, to take the time to vote.^[55] At the very least, crude candidate evaluations combined with the act of voting come closer to attempted electoral rejection than the passive act of nonvoting.