Roll up, roll up, punters, the Greatest Show on Earth is about to begin. India’s vast electorate – aside from those disenfranchised for reasons like moving away from home – is about to vote. Though that isn’t the show I meant. I was talking about the discussions. Ah, the discussions. We are in for two months of extraordinary claim and counter-claim, of rickety rhetorical houses built on assumptions of sand, of desperate attempts to bend the narrative one way or another.
And so, gentle reader, this handy guide to electoral myths in 2014. Don’t switch on your TV without it.
1. Myth: Opinion polls reveal something useful. No, they don’t. Not at all. This isn’t necessarily because they sample only a few thousand people in constituencies of millions; that, statistically, doesn’t make that much of a difference. No, the reason they don’t make sense is because those few thousands are not randomly chosen. (And perhaps because people don’t tell pollsters the truth.)
Consider this: in both 2004 and 2009, major polls underestimated the United Progressive Alliance (UPA)’s national seat tally by 30-60 seats, and overestimated the National Democratic Alliance’s by even more. Why? Probably because the Congress’ voters are predominantly lower-caste, poorer, and more rural — and pollsters tend to have fewer such people in their sample. Conduct the following thought experiment: suppose the Congress’ support among such voters has stayed the same, or marginally increased thanks to their higher living standards since 2009; but among richer, upper-caste and more urban voters, they’ve lost considerable support. Would polls underestimate the Congress’ national seat tally by as much? Or, perhaps, by even more?
Of course, this assumes that polls are halfway accurate even given their massive sampling error. But, as Chennai-based data scientist Puram has been pointing out on his blog about the CNN-IBN/ CSDS poll, that is perhaps too strong an assumption. Here’s some of what the IBN/CSDS poll claims: in one month, satisfaction with the UPA increased an extraordinary 32 per cent in Bihar and 15 per cent in Tamil Nadu; in UP, as the Bharatiya Janata Party (BJP)’s vote share declines, it supposedly gains seats; in Maharashtra, Narendra Modi lost nine per cent of his support in one month, while Manmohan Singh doubled his popularity. In the AC Nielsen poll for the same two months, the Congress was shown as gaining a percentage point in vote share — but the pollers said that they were likely to win 10 fewer seats. Is any of that halfway believable? If not, why are you wasting your time thinking about, or watching, polls? Why are respectable people commenting on them on TV?
2. Myth: More first-time young voters than ever, and they’ll make a crucial difference. As Karthik Shashidhar pointed out recently in Mint, the 1989 election actually saw a proportionally higher increase in the number of enrolled voters — as did 1996. In fact, 2009 might have as well, if not for the fact that the Election Commission had been working on cleaning up the rolls. We don’t know if the increase this time is mainly young voters; and, finally, there’s no reason to suppose that a high enrolment necessarily means that there’ll be a high turnout, allowing the first-timers to make a difference. And, above all, there’s no reason to suppose, from history, that younger people vote that differently from their older relatives.
3. Myth: Prime ministerial candidates make a difference to voters. Yes, we know that TV wants this to be true, and we know that Narendra Modi wants this to be true. But TV and Modi also presumably believe opinion polls, even if I don’t and this is what they say: that only 16 per cent of voters take into account who a party’s PM candidate is. This matches what ground-level politicians have always said. By all accounts, other factors matter more: the choice of candidates, local-level alliances, the performance of state governments. Of course, perhaps Everything Is Different This Time.
4. Myth: Narendra Modi will give the economy a bump. One effect of opinion polls showing a Modi wave, when combined with the very large numbers of innumerate idiots in the financial world, is that the markets already seem to be pricing in a Modi premiership. Thus, there might not be a lot of room for them to increase further if Modi actually does win. And, of course, in the longer term, there’s no indication yet that Modi or the BJP will be able to manage the problems the UPA faced any better. After all, the essential model is just the same — friendly to big business, broadly welfarist. Administrative efficiency might work at the state level; national-level growth needs reform. But Modi, like all other Indian politicians, is hesitant to spell out concrete reform. He’s not OK with the Centre changing labour laws, for example. (The only politician who’s declared he’s OK with it is, actually, Rahul Gandhi. One wonders why he didn’t mention it for 10 years.)
5. Myth: Modi needs a wave to win allies. Clearly not; Ram Vilas Paswan – remember, a relatively short time ago he was viciously attacking the Sangh Parivar for “saffron terror” and what-not – has proved that Modi isn’t as unlikely an ally as all that.
6. Myth: The Election Commission is always right. Yes, they keep the polls free and fair. But it’s an irritating Indian tendency to assume that people who get one thing right are, therefore, right about everything. In particular, the EC’s “model code of conduct” is quite ridiculous in parts. Why should they limit the number of cars in a convoy, the excuse used to shut down Arvind Kejriwal’s tour of Gujarat? Why should Union ministers be denied the right to make promises at public rallies? Why should parties not be allowed to distribute T-shirts? There’s a lot of nonsense in there. Let’s not praise it all.
