Defining 'Moral Values'
After relying on misleading exit poll data to announce the wrong winner in Campaign 2000 and after failing to produce any data at all in the 2002 mid-term elections, the television news divisions promised a new and improved exit poll system this year.
They promised not to jump to early conclusions from the exit polls and to use them instead for their proper purpose: to provide insight into voter motivation, decision-making and ideology.
Yet again on Election Day the exit polls failed. The networks are taking a false rap for rushed early bungling. A second, larger error is now only coming to light.
First, no blame attaches to the network news divisions for the fact that partial data was disseminated and misinterpreted online on Election Day leading to the misleading mood – before the polls closed – that things would turn out badly for Republican George Bush. On television, that mood was mostly conveyed not by journalists but by cable news guests, political pundits and operatives.
Examination of TV news itself – for example the broadcast nightly newscasts on election night before any results were declared – shows that anchors, in-house analysts and reporters were scrupulous in observing their vow not to use their in-house data, which apparently favored John Kerry, to taint their characterization of the likely outcome.
Sure enough, all through election night broadcast and cable networks conveyed the accurate impression that this was a close race, with almost no changes since 2000, where the results could only be called quickly in non-battleground states, and where the races were close enough in the swing states to remain uncallable for hours. We saw no hasty 2000-style misuse of the exit polls on air – only unauthorized use online, for which no blame attaches to the networks.
Blame does attach, however for the sloppy wording of the exit poll questionnaire itself. In trying to discover which issues were crucial, the poll cited the major themes of the campaign – the Economy, Iraq, Terrorism – but then mixed apples and oranges by adding a category – "Moral Values" – which is not an issue at all.
So our understanding of key issues was left clear as mud in the election aftermath when "Moral Values" turned up as the single most frequently cited issue, a category chosen disproportionately by supporters of the president. Pew Center pollster Andrew Kohut protested its inclusion on PBS' News Hour on the day after Election Day, explaining that, especially for devout Christians, all of life is seen through the prism of moral values, so faced with that category on any list it would feel un-Christian not to choose it.
What does the category mean? Is it shorthand for a cluster of genuine cultural issues such as abortion, gay rights, church-state relations or stem-cell research? Perhaps not. Perhaps, instead, it refers to the personal attributes of the two candidates and the importance of the President's born-again faith to a huge proportion of his electoral base.
Evidence for this latter reading comes from the Citizens Debate Scorecard conducted by MediaChannel.org and Media for Democracy. The Scorecard was an interactive online monitoring panel of the conduct of the presidential debates. It evaluated the appropriateness of the moderator's choice of questioning both on the issues and on the personal attributes of the candidates.
When the panel monitored the third debate on domestic policy in Tempe, Bush supporters and Kerry supporters diverged very little on their assessment of the mix of social issues chosen for discussion. However, when monitoring the relevance of the candidates' personal attributes – such as consistency, honesty, judgment – there was a striking divergence in attitudes towards morality, values and religious faith. A large proportion of Kerry supporters on the panel found too much time spent on values; while Bush supporters found the opposite – either the right amount of time or too little.
The media monitoring of the debates indicates that moral values are seen as a personal attribute of the candidate not as a complex of public policy issues akin to the economy or Iraq or terrorism.
When the networks designed their Election Day exit poll questionnaire to decipher voters' decision- making, they had an opportunity to deconstruct the morals question into its distinct components.
The networks' big exit poll failure was not in using their data too early but in not asking the proper questions in the first place.