Opinions and Polls

Email To Friend

• You can send this article as a link to anyone you feel would be keen to read it. Just fill the form below to send it now.
• Your Email: *
• Your Name: *
• Recipient Name: *
• Recipient Email: *
• Subject: *
• Message: I found a very interesting article on GoaNews.com and thought you will definitely like to read it.
• * mandatory fields

It is not just you and I who are finding Goa's 2017 election hard to guess. The professional psephologists, who conduct 'scientific' opinion polls, are equally divided.

Yesterday's election was difficult to predict. Every party has made confident claims. Union Defence Minister Manohar Parrikar talked of winning 33 of the 36 seats the Bharatiya Janata Party (BJP) is contesting. The Congress is a little more modest; it says it will bag 26 seats.

 The Aam Admi Party (AAP) won't talk numbers, but said on 1 February that "AAP is Sweeping Goa". According to its internal survey, AAP’s vote share in Goa will be 41.9 per cent, BJP 21.5 per cent, Congress 14.4 per cent and MGP 12.2 per cent.

 The MGP-GSM-SS alliance and the NCP are contesting lots of seats, but do not talk of majorities. In fact, the MGP seems confident that there will be a hung assembly, and has asked the BJP to enter into a post-poll alliance in which its leader, Sudin Dhavalikar, becomes chief minister...!

The others — Goa Forward, the Goa Su-Raj Party and United Goans — are contesting only a few seats each. The Goa Vikas Party is a temporary 'dharamshala' for aspiring Congress candidates who were denied tickets, and is just acting as a vote splitter for the benefit of the BJP in Salcete.

Let us leave aside, for a moment, the fact that ever since the state held its first election after statehood, in 1989-90, no party has ever got an overwhelming majority. On the contrary, many elections have thrown up hung assemblies.

 Even the 2012 anti-Congress tsunami gave the BJP, its pre-poll ally the MGP and all the independents it supported — combined — less than 30 seats. So let us first throw into the dustbin any talk of massive majorities in this election.

 It is not just you and I who are finding it hard to guess. The professional psephologists, who conduct 'scientific' opinion polls, are equally divided. Let's take a look at their efforts.

 That's seven polls, apart from the AAP's internal survey. Only two of them predict an absolute majority for the BJP. All the others seem to indicate varying combinations in a hung assembly. And the only thing that all but two agree on is that the AAP will put up a poor performance in terms of seats won.

 I have some experience in conducting and analysing opinion polls, and my observation is that unless the survey work is supervised and carried out painstakingly and meticulously, predictions can go very badly wrong. Conducting a survey by telephone, for example, is standard practice in western countries. But in India it is a highly risky venture.

 Even when the survey part of the poll is done well, what it predicts with reasonable accuracy is just the percentage of vote each party will get. To translate that data into seat predictions is an intensely complicated exercise, especially when there is a new factor in the fray (like the entry of AAP in this election). There's plenty of room for error.

And Indian psephologists have got it badly wrong several times, not least in the 2014 parliamentary poll (when only Kautilya got it right), in Delhi (not a single poll predicted AAP's 67 seats) and in Bihar. Will at least one of them get it right in Goa? We have to wait till 11 March to find out.