Research Note 110, November 1998
The Corporation for Public Broadcasting provided funding for local public television stations to broadcast political debates prior to the 1998 election. CPB commissioned research to determine how the programs aided viewers in the election process.
Summary of findings
Viewers recruited to view the PBS station local debate night program found the program to be of value in terms of helping them to become more informed voters, learning more than they learned in other media and increasing their knowledge of the candidates and the issues.
- After viewing the program, 83 percent of viewers said that the program helped them become more informed voters.
- Eighty percent said they learned something other than what they had learned from print media or radio.
- Viewers' familiarity with the candidates increased from 81 percent to 92 percent.
- Viewers reporting good or excellent knowledge of the issues increased from 51 percent to 67 percent.
- Nineteen percent said that viewing the debate changed their voting preference.
Methodology
CPB commissioned Gordon S. Black to conduct research on the Local Debate Night project that aired in conjunction with the national PBS program on the 1998 election. Four local areas were chosen for the project: Portland, Oregon (KOPB); Illinois (WTTW, WTVP, WSIU, WQPT), Arkansas (AETN), and Arizona (KAET).1 In Illinois and Arizona, the debate was between candidates for governor. In Oregon and Arkansas, the debate was between Congressional candidates. A total of 397 viewers completed interviews. 2
About the sample
The sample included only registered voters who agreed to watch the local debate on their local PBS station. Not surprisingly, 91 percent of those polled were PBS viewers, who reported watching on average about 6 hours of PBS per week. The sample was 56 percent female. Household income distribution was as follows: 25 percent $30,000 and under, 29 percent $30,001 to $50,000, 29 percent $50,001 to $75,000, and 18 percent more than $75,000. The sample was skewed toward older viewers with a mean age of 55; 19 percent were ages 18-39, 30 percent were ages 40-54, 20 percent were 55-64 and 31 percent were ages 65 or over. Seventy-four percent of the sample attended or had graduated from college, including 17 percent who had a graduate degree.
Pre-viewing questions
Prior to viewing the debate, respondents were asked about their likelihood of voting and their familiarity with the candidates and the issues. Given the demographics of the sample, likelihood of voting was especially high, with 81 percent indicating a certainty to vote. Certainty to vote was correlated to higher income, higher education, females, and older respondents. There were no significant differences across markets.
Respondents were generally somewhat familiar with the candidates--81 percent reported that they were somewhat or very familiar. However, only 29 percent said they were very familiar. Respondents in Oregon were the least familiar with the candidates, which can be explained by the fact that neither candidate was an incumbent.
Fifty-one percent of respondents felt their knowledge of the issues was good or excellent. Only 11 percent indicated an excellent knowledge of the issues, with variations across markets ranging from three percent in Arizona to 17 percent in Oregon. Those in the lowest income category were more likely to report being "not at all aware of the issues." Public TV viewers were more likely to indicate knowledge of the issues than non-public TV viewers.
Viewing the debate
Sixty-nine percent of the sample viewed the entire debate; 31 percent viewed only a portion of the debate. Viewers in Oregon were more likely to watch all of it, as were public TV viewers. Likelihood of viewing the entire debate increased with age; 57 percent of those 18-39 watched it all compared to 76 percent of those 65+. White respondents were also more likely to watch the entire debate.
Post-viewing comparisons
The percentage of those saying they were certain to vote did not change after viewing the debate. This is not surprising, given the high percentage (81%) that said they were certain to vote prior to watching the program. The percentage of those saying they were very likely to vote also remained unchanged. Watching the debate, at least among this sample, did not appear to change the likelihood of voting.
However, there were changes in the familiarity with the candidates and the knowledge of the issues. The percentage of respondents who were now very or somewhat familiar with the candidates increased by 11 points. The biggest increase was in the somewhat familiar category. An even greater increase was reported in the knowledge of the issues with 16 percent more people reporting a good or excellent knowledge of the issues. Biggest gains in knowledge of the issues following viewing were reported by college graduates and people with the highest incomes.
Perceived value
Eighty-three percent of viewers said viewing the debates helped them to become a more informed voter. This was especially pronounced among younger viewers (92%), and non-white and Hispanic viewers. There was no significant difference among the educational levels of viewers who felt the debates helped them become more informed voters.
We also asked if they learned something by watching this debate other than what they have learned from print media or radio. While 80 percent said yes, the percentage was even higher among the youngest respondents and among Hispanic and Asian respondents.
Compared to the other geographic areas, the Arizona sample rated the program significantly higher on these perceived value questions; this may be due to the Arizona sample being younger and slightly more ethnically diverse than the other areas.
These data suggest that television is a preferred medium for information among younger and diverse populations. It also suggests that televised debates may have a stronger impact on these populations.
Changed voting preference
In addition to learning more about the issues, did viewing the debate actually change any voter's inclination to vote for a particular candidate? Nineteen percent of those viewing said the debate changed their voting preference. While the change cut across most demographics, it was most noticeable among ages 40-54, and those in Arizona. (Party or candidate preference was not asked.)
National PBS program
While these viewers were not asked to watch the national PBS program, PBS Election 98: The National Report, that aired adjacent to these debates, about 22 percent said they also watched the national program. Older viewers were more likely to have watched this program with 29 percent of those in the 65+ age group reporting they watched the Newshour special.
Internet chat
In Oregon, we asked about awareness of the Internet chat room related to the debate. Nearly half (47%) said they were aware of the Internet chat room. Surprisingly, the youngest respondents were not the ones most likely to be aware of the chat room. Those in the 40-54 age range and those with the highest income category were the most likely to have been aware of the Internet opportunity. However, no one in our survey participated in the chat room. A larger or more targeted sample would be needed to determine the effects of participation in the chat room.
Conclusions
Public television stations perform an important service to viewers by broadcastng political depabates prior to an election. Viewers find the debates a valuable aid for participation in the democracy process.
1. Selection of stations and markets was made primarily on the reliability of the debate to air on Oct. 4, as there was a question about Congressional candidates availability.
2. Beginning on the Wednesday prior to the Sunday debate, random sample phone calls were made to contact registered voters and ask them to view the Sunday night debate. Approximately 200 people in each area were recruited to view. Reminder calls were made on Sunday, and the follow-up calls to viewers were made Monday through Wednesday to administer the questionnaire. Gordon Black completed 397 interviews evenly distributed in the four areas, just short of our target of 400.
