‘Hidden’ voting preferences

On May 4, former Tamil film actor C. Joseph Vijay and his Tamilaga Vettri Kazhagam (TVK) emerged in the Tamil Nadu Assembly elections as the single largest party. The magnitude of the victory was widely unexpected, and dislodged the Dravidian duopoly in the State since 1959.

While the reasons for this win are still falling in place — a need for change and reform and fatigue with corruption seem to have been decisive forces — it also appears more than a few voters expected the Dravida Munnetra Kazhagam (DMK) to win even as they wished to shrink the size of its majority, to send a message of sorts.

Eventually, of course, the TVK won, securing 108 seats in the 234-seat Assembly and with almost 35% of the vote share.

A rational choice

While the complete political psychology of Mr. Vijay’s win is still developing, we know the TVK did subtract more from the All India Anna Dravida Munnetra Kazhagam vote base. The erstwhile actor also had many vocal supporters, especially among the State’s youth. Many parents were even cowed by their starstruck and adamant children, often well below voting age, insisting they vote for the TVK irrespective of their own inclinations.

In the mix of possibilities, the idea that one would vote for another party assuming a second party would win anyway stands out as a self-negating, even unpredictable, proposition.

However, this way of deciding is actually rational and well-documented in political science and economics. The basic idea comes from the fact that people do not just vote for whom or which ideology they prefer. They also vote based on whom they believe ‘can’ win, how they believe others will vote, and what social rewards (or punishments) they associate with each choice. As a result, if voters believe one party is ‘destined’ to win, they could also believe they can defect to another party — i.e. voting for different reasons while not changing their expectations. When that idea becomes sufficiently widespread, the party they actually voted for could win.

Of course, psychology alone cannot explain what happened in Tamil Nadu. But the way psychologists, sociologists, and economists have explored these possibilities may describe a part of why the State voted the way it did.

Mediated by the media

In the 1920s, the U.S. social psychologist Floyd Allport described a concept called pluralistic ignorance. It says that if each person in a group privately believes something but incorrectly assumes most other people do not believe that as well, they can behave on the basis of a collective opinion that they have failed to read. Scholars have also found that any society whose people think like this is actually less stable than it looks. When private preferences diverge far enough from the public one, even a small trigger — like a rupture in the appearance of consensus — can cause people to quickly and progressively reveal their true choices as they see others doing so.

The TVK’s 34.92% vote share could suggest a reservoir of suppressed private preferences that did not always surface in mainstream discourse — or could it? Since the verdict on May 4, many reporters have recalled people on the ground telling them they would vote for the TVK — even as many in the media reported the DMK would emerge victorious.

They may have struggled to reconcile two conflicting impressions. On the one hand, they encountered enthusiasm for a two-year-old party and its charismatic leader. On the other, they weighed the DMK’s formidable organisational machinery and its entrenched presence in Tamil Nadu politics, plus the common assumption that parties with deep cadres and alliances generally prevail. Many observers may have resolved this contradiction in favour of what they already ‘knew’ about how elections usually work. After all, if such intuitions prove unreliable in one major election, they could unsettle the observers’ confidence in others as well.

Ultimately, those declarations in the press supporting a DMK victory could also have lowered the perceived cost of voting for an alternative. In Tamil Nadu, one popular media narrative of the TVK — that it was a celebrity vanity project, different in spirit from the campaigns of M.G. Ramachandran and J. Jayalalithaa — could also have made Mr. Vijay et al. seem like an alternative through whose bid to power individual voters could send a ‘message’.

Diverging preferences

Voters do not always infer information directly. They also survey the (apparent) behaviour of others. Each voter’s confidence in the DMK’s invincibility could have been abetted partly by observing what seemed to be confidence among others, leading some people to discount their own private impressions. This then could have created a feed-forward loop in which the strength of the private preference diverges significantly from that of the public one, until a voter reaches the ballot.

Scholars have sometimes modelled this as a phenomenon called a threshold effect: each voter has a personal point at which they will act on their private preference. When enough people cross that threshold at the same time, the electoral outcome can shift nonlinearly, meaning it could appear sudden and disproportionate to what people have been seeing around them until then.

In a 1995 book titled Private Truths, Public Lies, the Turkish-American economist Timur Kuran argued that individuals routinely misrepresent their private preferences in order to conform with public preferences, which he called preference falsification. And as falsified preferences accumulate over time, they could stabilise a misleading public consensus.

The political scientist Anthony Downs helped put together a rational explanation of voting in which people weigh the expected value of their vote, including the small chance that it could be decisive. In a political climate where the DMK’s victory could have seemed like a given, for example, voters might try to improve the value of their vote, as they see it. Based on this idea, around three decades later, Australian economist Geoffrey Brennan and his American peer Loren Lomasky argued that because the chance of any one vote being pivotal is really small in a large election, voters also use the (private) ballot as a place where they can express themselves — an interpretation some scholars have also applied to the Brexit referendum.

Defect without consequence

In fact, the duo contended that the lower a person perceives the value of their vote to be, the more sincerely they will vote for whom they actually want to vote for. In this reading, some of TVK’s voters could be said to have voted expressively. Some scholars have described similar situations in terms of the ‘wasted vote’ — in Tamil Nadu, not wasted because it went to the TVK but because it would have been of ‘little use’ to the DMK.

In his noted work on public goods and collective action, the economist Mancur Olson found that individuals who are all interested in a common collective good will still also under-contribute towards the goal as long as they can benefit from the contributions of others. So for instance, if a group of DMK supporters believed that the DMK was set to win, one individual in the group may have also believed that as long as the others voted for the DMK, she alone could vote to ‘send a message’ without consequence. The defector’s decision is actually rational because, according to the information that she had, the cost of defection was zero.

As analysts continue to make sense of the TVK’s unprecedented victory, it will be interesting to watch whether any of these reasons match what played out on the ground on April 23.