India’s Lok Sabha elections kicked off last week, with the first of seven phases that will carry on until May 19. On that day we will get the first indications of how voting has gone, since exit polls can only be published after voting in all phases is complete. Counting will begin on May 23 and the official verdict will become clear.

So what are political analysts and political junkies going to do between now and May 23?

One option is to listen to the claims of political parties themselves. But, with research suggesting that Indian voters are influenced by who they think is likely to win in an election, even trusted political sources are often attempting to convey the impression that their party has done well in the hope of creating “hawa”, the chatter that suggests which way the wind is blowing.

The other option is to look at turnout figures – the only officially published data that we have access to once voters have actually gone out and pressed the EVM buttons. The Election Commission provides official numbers of how many people voted after every phase. Analysts seize on it, hoping to derive some sort of political data from the numbers.

The turnout in the first phase on April 11 was 69.43%, just marginally lower than the figures from 2014, although there were much larger variations if you look a little more closely. All 17 Lok Sabha seats in Telangana, for example, saw a drop in turnout, while only two out of eight seats in Uttar Pradesh had a higher turnout than 2014.

What does turnout tell us, though?

High turnout = anti-incumbency?

The conventional wisdom is a very simple formula: high turnout means anti-incumbency, which is voters coming out in droves to kick out a sitting government. This should conversely suggest that low turnout is a pro-incumbent indicator.

Unfortunately, there is no evidence to suggest that this wisdom holds true. In 2018, Milan Vaishnav and Jonathan Guy published a paper in the journal, Studies in Indian Politics, which looked at data from 18
major Indian state elections between 1980 and 2012.

“Despite the popularity of the notion that citizens come out to the polls in greater numbers when they are motivated to punish the incumbent government, our analyses of three decades of electoral data uncover no such relationship,” the paper concludes.

Psephologist-turned-politician Yogendra Yadav, looking at national election data, arrived at the same conclusion, saying that the erroneous belief probably dates back to the general elections of 1967 and 1977, which saw anti-Congress waves.

BJP needs high turnout?

What if the question was narrowed a little more to account for specific political parties? In a new book about Indian elections and polling, NDTV’s Prannoy Roy and Dorab Sopariwala look at whether turnout might say something about the relative success of the Bharatiya Janata Party.

Analysing results from the Lok Sabha elections of 2004, 2009 and 2014, they conclude that “in each one of the three consecutive elections, the BJP+ performed much better in constituencies which had a low turnout.”

The authors suggest that one reason might be the intrinsic character of cadre-based parties, in this case counting the Rashtriya Swayamsevak Sangh as the BJP’s cadre, which are able to mobilise votes even if voter enthusiasm on average is low. “The key for the BJP is when low self-motivation amongst voters is widespread and leads to a low turnout; that’s when the BJP does best,” the authors say.

Modi wave?

What complicates this is the fact that 2014 was a “wave election”, which means turnout was high across the board, indeed the highest India has ever seen in a Lok Sabha election.

As political scientist Neelanjan Sircar has pointed out, the places that saw the highest spike in turnout also corresponded to BJP victories, suggesting that a significant chunk of the new voters were BJP supporters, or specifically voting for Modi.

Sircar wrote: “When many new voters were mobilised, the BJP was very likely to win the constituency. Voters also specifically came out to vote in constituencies where the BJP was competitive, and more so in constituencies in which the BJP was engaged in head-to-head battles with Congress.”

For Sanjay Kumar of the Centre for the Study of Developing Societies, this means that a lower turnout compared to 2014 should trouble the BJP, as was the case in six out of the eight Uttar Pradesh seats in the first phase, all of which it won five years ago. “I assume that in the first round a higher turnout would spell an advantage for the BJP in BJP-ruled states and a lower turnout might translate into the opposite for this party,” he wrote. “If that happens, the BJP would be down six in UP in the first round.”

BJP fine with low turnout?

But this contrasts with the analysis by Roy and Sopariwala earlier, which suggested that BJP tends to do well with low turnouts.

Meanwhile, Centre for the Study of Developing Societies-Lokniti pre-poll survey appeared to show that BJP voters were much more enthusiastic about going out and voting:

“In other words, NDA [the BJP-led National Democratic Alliance] voters are more likely to turn out and vote and non-NDA voters are less likely to turnout to vote,” the report said. “It is clear that if the UPA [the Congress-led United Progressive Alliance] and the non-UPA parties do not get their act together, the levels of eagerness among their supporters is not likely to increase as voting day approaches.”

So what does all of this tell us?

Simply put, the analysis is mixed.

What we know from the first phase is that there does not seem to be a wave, in the sense of a higher turnout compared to 2014 although it may remain at the same level as then.

Analysts seem to believe that lower turnout, particularly in the Uttar Pradesh seats, is bad news for the BJP. Yet historical trends and the Centre for the Study of Developing Societies voter enthusiasm data suggests the BJP will perform better in a low turnout situation, though that is more clearly established in places where the saffron party is up against the Congress and may be complicated by the new alliance arithmetic of Uttar Pradesh.

As almost all will say though, turnout itself tells us nothing unless you also analyse it alongside more contextual factors specific to a constituency, as Kumar and Dipu Rai among others have attempted to do for the first phase. As more voters go out to push buttons, there will be more such analysis, but we will only know how accurate this will be once the votes are counted on May 23.