Assembly elections 2017: A keen analyst or pundit needs to travel extensively, know his sociology

Trying to figure out trends among voters during an election campaign is tough. Trying to predict the results on the basis of an "election tour" is a slippery slope to shame for pundits.

Experienced politicians with an ear to the ground are generally quite quick at assessing the public mood. While traveling on an election campaign, I have overheard such political masters as Atal Bihari Vajpayee and VP Singh remark — as soon as they finish making a speech — hum yahan haar rahein hain (we are losing here).

That sort of feel for the public pulse is rare among ordinary mortals. But it is possible to get a sense of the pulse with sheer hard work and diligence.

Representational image. PTI.

Representational image. PTI.

One of the first things I figured out when I began covering elections was that urban trends were far more difficult to predict than rural ones. For, while urban voters tend to vote as individuals, group patterns tend to hold firm among rural voters.

It helped that I got my early immersion in election coverage in Uttar Pradesh, during campaigning for the 1991 Lok Sabha election campaign. Those elections took place less than a year after the Mandal report had been implemented and the Ram Rath Yatra had tried to mobilise Hindutva consolidation. It was as fierce a campaign as one can imagine.

And if that was not enough, Rajiv Gandhi was assassinated in the middle of campaigning. Part of the state had already voted; the rest was yet to vote. And the elections went on interminably during an extremely hot summer.

While traveling through rural areas, one had to learn to keep in mind the sociological profile of each person one interviewed. I learnt pretty early that interviewing several individuals separately was far more useful than chatting with a large bunch. Tedious, but far more fruitful. One had to figure out not only a person's age and relative socioeconomic status but also his/her caste. In the heartland of 'Bharat', during the Mandal riots, I learnt that caste was an extraordinarily important part of person's sociological profile.

Another thing one learnt quickly was never to directly ask a voter which way they would vote. Rather, one asked what the trends in that area were. Invariably, the interviewee would describe the trends among their own kind of people — ie, among their caste, gender, age and socioeconomic type. One also learnt not to ask which one party was likely to win. It worked much better if one asked what the main fight was — which were the main contenders. More often than not, the interviewee mentioned the party he/she preferred, as also the leader of the pack in that area.

In those elections of 1991, and the next few elections in UP, it became quite that people were by and large voting according to caste affiliations. By traveling over longer distances and interviewing different kinds of people at some intervals, one could figure out general patterns — even call the election. It's very difficult to predict an outcome without that kind of extensive tour, without a keen eye on sociological patterns. Even massive waves such as the ones in Uttar Pradesh and Uttarakhand over the last few years can be tough to see without this sort of rigour.

Some of the most high-profile journalists missed such obvious waves like the one in Andhra Pradesh in 1994 and in Kashmir in 2002. Since the latter was an anti-incumbency wave, it was not clear whether the Congress or the PDP would win. But it was clear that the National Conference was set to lose.

Not every round of elections is so easy to read, even with rigorous hard work. The Bihar election of 1995, for instance, was tough to call, as was the Goa election that just ended.

Given energy, interest and requisite rigour, though, it can be exhilarating to cover an election campaign.


Published Date: Mar 13, 2017 03:52 pm | Updated Date: Mar 13, 2017 03:52 pm