Exit polls: Popcorn entertainer
Going by its track record, the super prime time show this evening will be a bad trailer of the actual show on Friday
Just a few hours from now, the CEO and her driver or the friendly neighbourhood grocery store owner and his delivery boy would be united – albeit for some time. They will all be glued to the idiot box to watch the Monday blockbuster on the new time slot brilliantly packaged as super prime time by a television channel. To be sure, recipients of bad news from the exit polls will beat their chests and abuse the channels; beneficiaries of good news will strut and crow like obnoxious school kids.
But what’s the fuss all about? Come to think of it – exit polls are not even a good trailer of the actual show to be unveiled on Friday as they have proved to be hopelessly wrong for two times consecutively – in 2004 and 2009, not to talk about similar disasters before several state election results. In 2004, all the exit polls predicted a return to power for the National Democratic Alliance giving the ruling alliance a lead of 40-90 seats and more; in the end, the BJP-led alliance was routed by the UPA. In 2009, all the exit polls predicted a thin lead for the ruling UPA; some predictions sighted a single-digit margin between the two alliances. In the end, the Congress-led UPA ended up almost 100 seats ahead of the NDA
The exit polls were quickly mocked for their inaccuracy. How could they be so wrong? people wondered. Yet the artificial excitement continues exit polls after exit polls.
And India isn’t alone. In 2004, for example, exit polls gave Senator John Kerry such a massive lead over President George Bush that Democrats began to prematurely celebrate based on what turned out to be unreliable figures. The same thing happened again in 2008, albeit to a somewhat lesser extent.
Many would say terming exit polls unreliable is too charitable a view. Others have termed exit polls a fake and a con job even in the US. The Collier brothers, authors of the book, VoteScam: The Stealing of America, talked about this long back. And here’s why. In 1970, Channel 7 in Miami projected with “100 per cent accuracy” the final vote totals on Election Day. When the Colliers asked the networks where they got their exit poll data, the broadcasters told the Colliers that a private contractor used the data from a single voting machine to project the winners. On his part, the contractor said he got the data from a University of Miami professor, who in turn denied it. In the end, the news broadcasters appeared to have pulled the polling numbers out of thin air.
The expose prompted many to demand that the practice of indicating winners using manipulated exit poll data must be terminated now. The news networks have also not helped their cause by not giving hard evidence of what kind of exit polls they carry out.
There are many theories doing the rounds on why exit polls have failed in India. First, the greater willingness of voters of the perceived winning parties to be polled. Second, the diversity of India extends to every village, so it is almost impossible to ensure that any sample size is completely representative and we are forced to suffer the consequence of the belief that wrong data is better than incomplete data. That means when ‘learned’ analysts, highly rated statisticians and media-savvy social scientists pore over exit poll figures to give you insights on how caste and community influenced the “result”, they are largely an exercise in fiction. The pollster’s job is also getting more and more difficult with the Indian polity becoming increasingly fragmented with many local parties emerging as formidable entities at the national level. The characteristics of the electorate are just too many to be incorporated in any statistical model.
Besides, there are practical problems, too. A researcher for an exit poll firm who spends the morning outside a polling booth, for instance, may meet voters predominantly from one community or voting group, whereas an opposing block may vote late in the afternoon — after work — in large numbers.
There isn’t any doubt that exit polls give us a treasure trove of information about the voting behaviour of subgroups of the electorate but they are after all imperfect instruments. Meanwhile, switch on your TV this evening – a popcorn entertainer is assured at least.
But what’s the fuss all about? Come to think of it – exit polls are not even a good trailer of the actual show to be unveiled on Friday as they have proved to be hopelessly wrong for two times consecutively – in 2004 and 2009, not to talk about similar disasters before several state election results. In 2004, all the exit polls predicted a return to power for the National Democratic Alliance giving the ruling alliance a lead of 40-90 seats and more; in the end, the BJP-led alliance was routed by the UPA. In 2009, all the exit polls predicted a thin lead for the ruling UPA; some predictions sighted a single-digit margin between the two alliances. In the end, the Congress-led UPA ended up almost 100 seats ahead of the NDA
The exit polls were quickly mocked for their inaccuracy. How could they be so wrong? people wondered. Yet the artificial excitement continues exit polls after exit polls.
And India isn’t alone. In 2004, for example, exit polls gave Senator John Kerry such a massive lead over President George Bush that Democrats began to prematurely celebrate based on what turned out to be unreliable figures. The same thing happened again in 2008, albeit to a somewhat lesser extent.
Many would say terming exit polls unreliable is too charitable a view. Others have termed exit polls a fake and a con job even in the US. The Collier brothers, authors of the book, VoteScam: The Stealing of America, talked about this long back. And here’s why. In 1970, Channel 7 in Miami projected with “100 per cent accuracy” the final vote totals on Election Day. When the Colliers asked the networks where they got their exit poll data, the broadcasters told the Colliers that a private contractor used the data from a single voting machine to project the winners. On his part, the contractor said he got the data from a University of Miami professor, who in turn denied it. In the end, the news broadcasters appeared to have pulled the polling numbers out of thin air.
The expose prompted many to demand that the practice of indicating winners using manipulated exit poll data must be terminated now. The news networks have also not helped their cause by not giving hard evidence of what kind of exit polls they carry out.
There are many theories doing the rounds on why exit polls have failed in India. First, the greater willingness of voters of the perceived winning parties to be polled. Second, the diversity of India extends to every village, so it is almost impossible to ensure that any sample size is completely representative and we are forced to suffer the consequence of the belief that wrong data is better than incomplete data. That means when ‘learned’ analysts, highly rated statisticians and media-savvy social scientists pore over exit poll figures to give you insights on how caste and community influenced the “result”, they are largely an exercise in fiction. The pollster’s job is also getting more and more difficult with the Indian polity becoming increasingly fragmented with many local parties emerging as formidable entities at the national level. The characteristics of the electorate are just too many to be incorporated in any statistical model.
Besides, there are practical problems, too. A researcher for an exit poll firm who spends the morning outside a polling booth, for instance, may meet voters predominantly from one community or voting group, whereas an opposing block may vote late in the afternoon — after work — in large numbers.
There isn’t any doubt that exit polls give us a treasure trove of information about the voting behaviour of subgroups of the electorate but they are after all imperfect instruments. Meanwhile, switch on your TV this evening – a popcorn entertainer is assured at least.