Online campaigners, citizens, and portals in the 2002 elections
Part Two: The Online Citizenry
I Still Haven’t Found What I’m Looking For
This section of the report is based on a tracking survey of 2745 U.S. adults conducted by Princeton Survey Research Associates (PSRA) for the Pew Internet & American Life Project and the Institute between October 20 and November 24, 2002. The section also draws on a continuous body of research conducted by PSRA for the Project, and analyzed by Project staff.
Of the month-long survey population, 1707 (62%) were Internet users. Of the Internet subpopulation, 741 (43%) said they got political news and information online. The margin of error for the full sample and the Internet subpopulation is plus or minus 2 percentage points. The margin of error for the sub-subpopulation of political news and information viewers –henceforth referred to as “the online citizenry”—is plus or minus 4 percentage points. Bracketed notes at the end of sentences in this section, such as [POL 02], designate survey questions; the basic or “topline” responses to every question in the tracking survey may be viewed along with the questions at www.pewinternet.org and www.ipdi.org.
A Growing Constituency for Online Politics
The online citizenry constitutes the segment of Internet users who pay attention to politics. Its ranks swelled considerably in the 2002 election cycle, as Table 4 illustrates.
A 39% increase is remarkable. What can account for it? Campaigners were probably not the primary force drawing 13 million more people into the online citizenry between the summer of 2000 and November 2002. As we learned in Part One from the IPDI questionnaire, campaigners had a hard time using the Internet just to reach supporters, let alone uncommitted political observers and non-political users. The survey data confirm the campaigners’ sense of their own spotty effectiveness in connecting with the electorate. A mere 22% of Internet users (roughly half of the online citizenry) paid attention to election 2002 information.14 [POL03] Most of these election information consumers did so less than once a week. [POL 04] Only 12% of Net users (a little more than a quarter of the online citizenry) visited partisan Web sites such as those run by the political parties, a candidate, or a campaign. [POL09c]
Nor can a technological innovation explain the 2000-2002 expansion of the online citizenry’s population. While the popularization of Google, the second-generation search engine, may have eased Internet users’ ways toward getting news and political information, Google did not provide them with the impetus to head in that direction. No public service announcements or paid advertisements lured the viewers of Google’s home page into civic and political Web sites. Nor can the online citizenry’s growth be attributed to a surge in general Internet usage. As the table shows, there was no big spurt in people coming online; indeed, overall Net population growth slowed after 2000. Finally, the data suggest that the growth of the online citizenry did not depend on the financial health and social cachet of the high-tech industry, in as much as those were plummeting at the same time the online appetite for politics was rising.
A more likely explanation attributes the expansion of the online citizenry to three factors: the maturation of users, the spread of broadband, and big news stories. In the fall of 2000, 36% of online Americans had been on the Net for three or more years; by November 2002, that figure had nearly doubled, to 70%. Experienced Internet users do a wider range of activities while they are online.15 The growth of broadband is a compatible development. Broadband enables people to look for political information faster, in richer media forms, and without tying up a telephone line. Project research has determined that, even when controlling for online experience and a number of demographic factors, having a home broadband connection has an independent and strongly positive impact on the number of activities conducted online, the frequency of logging on, and the amount of time spent online daily.16
Big news stories, in which crucial details could break at any moment of the day over a period of weeks, also powered the expansion of the online citizenry. As people headed to news and public information sites in hopes of learning who won the Florida electoral votes and who was responsible for the 9/11 attacks, they discovered in the process the convenience, depth, and breadth of online political communication. The popularity of these factors may be seen in Table 5:
In sum, the enhanced capacity and facility of users to check in on breaking news and “drill down” for details is probably responsible for the sharp growth in the population of the online citizenry. There remains considerable room for further expansion: as of November 2002, 70 million online game players, e-mailers, music down-loaders, office workers, students, and other American adults had yet to look for political news and information. The U.S. confrontation with Iraq has probably lured some of that 70 million into the online citizenry since November 2002, and war would draw in more of them.
The important lesson here for political activists, and electoral office-seekers in particular, is this: Campaigners who can make news, shed light on the news, and empower people to act in response to the news, have a greater incentive with each passing day to try and connect with the Internet population. Online citizens are looking for just those qualities, and perhaps more, as they search the Web and comb through their email. In 2002, most of them did not find such information at campaign sites.
Online citizens are, on the whole, a dissatisfied bunch. Comparative survey data compiled for the Project reveals that people who have used the Internet to get political news and information were less likely to find what they were looking for than those who have sought health information and information from government agency Web sites. Political information seekers had fewer bookmarks of favorite sites than health and government information seekers. They relied more on portals (the subject of Part Three of this report), and browsed more Web sites. They said they learned something new less often. And they said they found information which helped them make a decision less often. In short, the average experience of the online citizen is not as rewarding as that of the constituencies for health and government information.
Some of this dissatisfaction derives from qualities inherent in political information. Politics is often about controversial matters for which no clear answers exist, and the two-party winner-take-all electoral system tends to polarize the fuzzy answers with which we all must make do. A Web searcher looking for a tax form does not have to choose among Democratic, Republican, and news media versions of the form, as does a Web searcher looking to learn about tax reform. A Web searcher in need of relief for a strange rash may be bewildered by the variety of potential diagnoses and remedies offered online, but at least the rash-bearer will not be subjected to one source of advice attacking another, as will a Web searcher attempting to bone up on the nation’s health care system.
A majority of survey respondents said that, as they searched, they tended to encounter viewpoints different from their own. [POL25b] That is a good thing from a civic standpoint, because public debate becomes brittle when people just have their own views reinforced. Nevertheless, being exposed to contrary facts and opinions can be a discomfiting experience for an individual, generating what psychologists term cognitive dissonance. Another source of anxiety for the online citizenry may stem from the fact that while most Net searchers tend to be task-oriented, political Web sites frequently turn the tables on their audience. In politics, unlike health care and government, (and business and education, for that matter), it is the sources, not the users, who come loaded with tasks: “Call so and so and tell him thus and such.” “Vote for me.”
And then there is the credibility of political information, the matter of how much the online citizenry trusts that what it sees is accurate and not misleading. As Table 7 illustrates, Internet users place much less stock in the veracity and integrity of what partisan and issue group sites provide than in what non-partisan and news media sites provide.17 That is a bleaker situation than it may appear, because the information supplied by the latter is largely derived from the former. For example, DNet, the online database of campaign information produced by the non-partisan League of Women Voters, consists mainly of a matrix of candidate issue positions submitted by the campaigns themselves; DNet solicits the information and arranges it for ready comparison. Campaign reporting in the news media consists mainly of excerpts (sound bites) from candidate, issue group, and party press releases and news conferences, often embedded in cynical speculations about the strategy and money behind them.18 What is “trusted” in non-partisan and news media versions of political information is thus more a matter of format than content.
Wherever the online citizenry alights in cyberspace, it is likely to see political figures issuing self-interested promises to fix society’s problems, accompanied by attacks and allusions by opponents and journalists which cast doubt on the likelihood that the promises will be kept and the problems solved. These discouraging depictions may account for the low percentages of Internet users who visit non-partisan (14%) and partisan (12%) Web sites. [POL09c-d] A final indicator of citizen dissatisfaction: political news and information searchers act more out of a sense of duty to be well-informed (71%) than because they enjoy following campaign news (24%). [POL02]
This is not a pretty picture. An online supplier of political information, like a struggling stand-up comedian, plays in a tough room: the audience is sparse, suspicious, impatient, and inured to the contentiousness and raggedness of the show. All the same, there is a chance for one or more of the performers to succeed. The survey points up several ways they can polish their act. Campaigners, civic groups, and the news media can provide some of the features the online citizenry says it wants to see at little cost and low risk.
What Online Citizens Like to See and Do
The Internet offers a variety of activities to the online citizenry. The survey asked whether respondents ever engaged in eleven of them, and the results appear on the following page.
Online citizens prized the research function of the Internet as highly as campaign managers. Searching for more information about candidate issue positions (cited by 64%) was the most popular response by far. The second and third most popular answers also fall under the category of research: online citizens hunted for candidate ratings and endorsements by organizations and groups (38%), and for candidate voting records (34%). Getting information about when and where to vote ranked seventh (22%), but this type of research has grown steadily since 1998.
The survey discloses some things about how systematic citizens are as they look for political news and information online. Prompted to recall the last time they searched, 33% of the online citizenry said they made a beeline to a specific site for a specific reason, while 56% did not have a specific site in mind. [POL 19] To a question about the number of sites searched, two to three sites received the highest response, 45%, while 18% said one site, 17% four to five, and 12% six or more. [POL22] Asked how they decided what Web site to visit first after using a search engine, a sizable majority (63%) said it read the explanation of each Web site and chose the one that best fit what was being sought. In contrast, only 10% clicked on a site whose name or sponsor was recognized, and 24% started at the top of the list. [POL21]19
After research, participating in online polls proved to be the most popular activity (32%). Selecting a favorite from an array of statements about issues, candidates, and other political news developments was thus the top mode of online political engagement, well ahead of sending email (17%), taking part in discussions and chat groups (7%), and contributing money (5%). It is unclear how seriously people regard these polls, whether they think they are expressing an opinion that the Web site’s operators and other political elites will heed, or having fun seeing how their choices square with those of previous participants. If it is the former, they are being fooled. But, again, there is a lesson here for online campaigners: the online citizenry enjoys interactive polls.
Getting or sending email jokes about the campaigns and elections ranked fifth among the activities asked about, at 31%. This is a big drop from 2000, when 54% of respondents participating in a Democracy Online Project survey answered in the affirmative to a similar question. Of course, the 2002 election took place in a more somber time. Yet humor remains an effective rhetorical tactic for online campaigning.20 It assuages the fuzziness, contentiousness, and anxiety of politics. It suits the activity of breaking from work, which many Net users are engaged in when they turn to look at political news and information.
These findings suggest several practices for campaigners to adopt in order to increase their appeal to the online citizenry:
Campaign Web sites should present information in database formats, which are more conducive to research, and less evocative of news media reports and advertisements, than the prevailing news release and brochure forms.
Campaigns should pay attention to search engine listings and the descriptive language in them.
It’s hard to be genuinely funny, but when a campaign develops or even discovers humorous material which suits its goals, it should pass it along to supporters via email, and suggest they pass it further along, to friends, colleagues, and family.
Campaigns should offer blogs, public diaries which salt observations on the passing parade of life with links and photos. Blogs are increasingly popular on the Internet; Blogger, a free do-it-yourself blog software service recently purchased by Google, counts one million regular users. Blogs speak to the universal human interest in how someone else is fielding the challenges of daily living, and in particular, what someone else has found worthy of a look in the vast catacombs of cyberspace. Blogs, in short, are friendly guides to online research and humor. A campaign with a successful blog will be a campaign with a loyal following.
The Online Citizenry and the 2002 Elections
As the online citizenry inhales and exhales political information during an election campaign, are minds being changed or just reinforced? This is perhaps the biggest question about the Internet for campaigners, who want to do each to different groups at different times with different messages. Reinforcement is easier, less challenging, and less useful to campaigns, grass roots mobilization being valuable mainly at the beginning (for seed money) and end (for voter turnout). Persuasion is much harder, so much so that an efficient campaign will try to minimize it, and concentrate on wringing every last bit of existing support out of an electorate, instead of exhausting its resources on quests to create new support. Nevertheless, candidates and causes who enter a campaign with less than 50% of the likely electorate’s support (and political mediators seeking to enlarge their market share) must engage in persuasion to some degree.
Only 13% of those who used the Internet during the 2002 election said that the last time they went online for political/campaign/election news and information, they were looking for information to help them decide how to vote. In contrast, 24% sought information to reinforce an existing preference (that is, information about a candidate or issue already supported or opposed), and 43% selected the option “to generally learn more about what’s going on in the campaign.” [POL 18]
We posed the persuasion question in two other ways. Internet users who voted in 2002, a different sub-subpopulation than the online citizenry, were asked “How important has the Internet been in terms of providing you with information to help you decide how to vote in the November election.” [POL 12] The results for 2002 appear below along with results from Democracy Online Project polls conducted in November 1998 and 2000. For the majority of respondents, the Internet was not that important.
We also asked the online citizenry whether “any of the information you have received online about the 2002 mid-term elections made you decide to vote for or against a particular candidate.” Phrased this way, with an emphasis on particular candidates instead of general experience, a larger minority (25%) said it was influenced. However, the percentage dropped markedly from 2000 (43%) and 1998 (34%).
This data needs further investigation. Still, it fits the pattern suggested by the popularity of the Internet as a research medium and, for that matter, by the preponderance of political science studies of voter behavior: on the whole, the opinions that the online citizenry possesses when it logs on tend to be reinforced, not altered, by the information it encounters, and the activities it engages in.
- The characteristics and behavior of the online electorate will be analyzed in a subsequent report. ↩
- “Getting Serious Online,” 2002: http://www.pewinternet.org/Reports/2002/Getting-Serious-Online-As-Americans-Gain-Experience-They-Pursue-More-Serious-Activities.aspx ↩
- “The Broadband Difference,” 2002: http://www.pewinternet.org/Reports/2002/The-Broadband-Difference-How-online-behavior-changes-with-highspeed-Internet-connections.aspx ↩
- A 1999 Democracy Online Project survey found similar low levels of trust. It is archived at www.ipdi.org. ↩
- On the negativity bias of campaign reporting, see Thomas E. Patterson, The Vanishing Voter: Public Involvement in an Age of Uncertainty (New York: Knopf 2002), Chapter Three. ↩
- Only 135 people responded to this question, so the results are more suggestive than statistically significant. ↩
- For an example from the 2002 campaign, see the account of the DNC Web video cartoon “Social Insecurity” in Michael Cornfield, “Let’s Roll,” Campaigns & Elections, December 2002/January 2003, p. 76. ↩