Political Polling Position Report
8301 Creekside Circle, #920, Bloomington, MN 55337
December 4, 2007
Reducing Negative Campaigning
Summary —The Civic Caucus is recommending a broadening of political polling to include depth of support for candidates, a step that might reduce some negative campaigning, immediately, for the 2008 elections, without waiting for legislation.
Negative, attack-your-opponent political campaigning has been intensifying with each election. As 2008 approaches, the trend is continuing. The attack approach is frequently defended by political strategists, who claim it is necessary, if for no other reason than to fend off the opposition, and, they say, it works. But the attack strategy is destroying confidence in government and is turning off the most constructive voters. We in the Civic Caucus have placed high priority on seeking ways to stimulate positive campaign commentary that would appeal to as many voters as possible on the political spectrum.
It is not necessary for the electorate to simply sit on the sidelines during the 2008 campaign and bemoan circumstances. A change can be implemented immediately, without new law, with potential to help candidates who are more interested in positively representing their own positions than in attacking their opponents. The change can apply to any election at any level of government without approval of lawmakers, candidates, political parties, special interests, or campaign organizations.
However, the change would require cooperation from the political polling organizations, although only one pollster willing to experiment is all that is necessary to get the idea off the ground.
The change lies in how political polls are conducted. Currently, public polls report the first choices of voters at any one time For example, a poll might reveal the following results: 35 percent, candidate A; 27 percent, candidate B; 21 percent, candidate C; 13 percent, candidate D, and 4 percent, candidate E.
Such results are helpful in identifying depth of support for candidates, that is, the number of their most committed followers. The assumption is that the plurality candidate is ahead, irrespective of whether the candidate has majority support or not. Consequently, a candidate can continue to target his or her main base of support and not worry too much about alienating voters who favor other candidates.
We suggest the polls make a slight change to illustrate breadth of support, as well. That is, the change will illustrate which candidates are attracting interest from the broadest field of potential voters. Candidates who seek broad support are more likely to use campaign tactics calculated to alienate fewer voters. It is likely that such candidates will de-emphasize negative campaigning. They might take positions on issues appealing to a broader constituency. They might refrain from direct attacks on opponents.
A simple polling change is required: pollsters would ask voters to identify their second and third choices, not just their first choices.
Pollsters could select any number of ways to report results. How the results are displayed may be less important than that the candidates are aware that second and third choices of voters are being reflected. Experts in political science, polling, and elections can be expected to advise which approaches communicate the best information to the public.
One possibility would be to report the percentages for second choice and for third choice, as is done now for first choices. Another possibility would be to assign points, say, three points for first place, two points for second, and one point for third; add up the points and show the rankings.
A third possibility would be to use a method of tabulating similar to that employed with instant runoff voting (IRV), which already is used in San Francisco and several other cities in the nation, and in some foreign countries, and which has been approved for use in city elections in Minneapolis. Such a method produces what might be called a consensus candidate.
Undoubtedly other possibilities exist.
Our specific recommendations :
* We recommend that polling organizations nationally, such as the CBS, ABC, and Gallup Polls, and in Minnesota, such as the Minnesota Public Radio/St. Paul Pioneer Press Poll and the Minnesota Poll of the Star Tribune, add ranking candidates to their regular polling.
The concept should be applied among candidates within a given political party and among candidates of all parties.
* We recommend that precinct caucuses in Minnesota next February use ranking candidates, in addition to their regular straw polls.
* We recommend that political parties use ranking candidates at their endorsing conventions.
* We recommend that individual organizations try ranking candidates among their own members.
Such steps will make it possible to learn more about, and experiment with, ranking candidates by preference without a legal commitment one way or another about its desirability or constitutionality for official elections.
In addition, voters will have more information about the candidates. Currently, voters only know how much support each candidate has, with limited information on breadth of support. Voters would like to know that while one candidate might be leading in the polls, using the current process, another candidate might have more depth of support, when second and third choices are considered.
We fully acknowledge that no one really knows whether adding a provision of ranking candidates to the political polling process will reduce negative campaigning. But why not try the idea out?
We are continuing to evaluate IRV and might make recommendations to the 2008 Legislature on whether the process should be implemented in more official elections.