Three factors that may decide US poll outcome3 min read . Updated: 02 Nov 2008, 09:14 PM IST
Three factors that may decide US poll outcome
There are grounds for the scepticism. Consider the following facts: a series of polls conducted by polling and media organizations are giving Obama an average lead of 6.8 percentage points over McCain. The lead enjoyed by Obama has consistently risen over the campaign period as the US economy went into recession and markets plunged.
But, in the last lap, the gap has narrowed somewhat. Further, according to the opinion polls, 14% of voters are still undecided. Do the narrowing lead for Obama and a high proportion of undecided voters point towards a closer electoral race than what was suggested earlier?
Let me draw upon my Indian experience over the past two decades to make a few observations that could prove to be critical in determining the accuracy or otherwise of the opinion polls conducted in the run-up to the US presidential election.
Firstly, it is important to ascertain whether voters are telling the truth about their choice or are giving politically correct answers. Culturally, this is a problem with the Indian electorate and Americans are not known to be prone to it. Recall how Indian polls predicted a win for the Atal Bihari Vajpayee-led National Democratic Alliance (NDA) in the 2004 elections as voters told the media what they wanted to hear or what was considered politically correct. As Vajpayee is a widely respected leader and is perceived to have performed well, it may have been difficult for voters to say that they preferred somebody else to be the prime minister or would vote for some other party. Voters have begun to tell unsuspecting pollsters and the media what they anyway propagate, perhaps to please them.
Are Americans falling prey to a similar tendency in this election and telling pollsters they would vote for Obama, who is younger and considered to be more dynamic than McCain? It is possible that some supporters of McCain are not revealing their real preference to pollsters due to the “shame factor" associated with his age or with the non-performance of the Bush administration, which has failed on many fronts, including the economy.
Secondly, polls and election administrators are predicting a large turnout in the election. A high turnout would be more reflective of voter anger against the incumbent Bush administration and thus will be a negative vote against Republicans rather than a positive mandate for Obama.
American elections usually register moderate turnouts. The turnout was 55.3% in the 2004 election and around 50% in the 2000 and 1996 elections. It never exceeded 60% in the past four decades. If the turnout in this election was to be significantly more than 55%, it would be considered high.
An increase in the turnout would brighten Obama’s victory prospects. Conversely, if the election registers a turnout of around 50% and less, it could be big trouble for Obama as a number of voters may be voting for him on the basis of his age and merit only in opinion polls and not in the real election due to his racial background.
Thirdly, there is a lot of suspense about the sizeable chunk of 14% of the “likely voters" who are still undecided. Why such a large proportion of voters are still undecided despite being presented with a starkly different choice of presidential candidates is a question that is bothering poll pundits. It is possible that such voters have actually decided but have not revealed their preference to pollsters. Referred to as the “spiral of silence", this is a major reason opinion and exit polls get it wrong.
Polls show that most of these undecided voters are Republican leaning and if they all voted for McCain, the election could be very close. Even other undecided voters with no party affiliation may vote for McCain on election day owing to the American voters’ tendency to back the underdog in an election.
If American pollsters are right— and they have an excellent track record—Obama would be the first ever African-American president of the US. If the pollsters got it completely wrong, it could be because they failed to consider the three factors I have listed.
Also Read G.V.L. Narasimha Rao’s earlier columns
G.V.L. Narasimha Rao is a political analyst and managing director of a Delhi-based research consulting firm. Your comments are welcome at email@example.com