Truths, myths about the low voter turnout
Very little is known about the causes and consequences of turnout change. So, caution is advised on drawing broad inferences from any turnout number
The decline in voter turnout in the first two phases of the 2024 Lok Sabha elections — 66.1% and 66.7%, respectively, compared with 69.4% and 69.2% in the 2019 polls — has attracted considerable attention. Several theories on both the causes and consequences of this turnout dip are being discussed.
What has led to the decline? And, which party will benefit or suffer the consequences of the turnout in 2024? More importantly, what does historical data on turnout inform us about Indian politics? The short answer is we do not know much about both the causes and consequences of turnout change. And, we must be very cautious in drawing conclusions.
Quite like any other indicator, voter turnout can’t always rise (or always fall). It is likely to revert to the mean after multiple increases or decreases. The overall turnout increased between 2009 and 2014 (from 58.2% to 66.4%), and then to 67.4% in 2019. Since 1952, the voter turnout has never increased for three consecutive elections. so, though two-thirds of the Parliamentary Constituencies (PCs) are yet to vote, there is a high probability that the turnout will stagnate at or decline from the 2019 baseline.
The turnout declined by approximately three percentage points combined in two phases over 2019. These are normal fluctuations. To illustrate, the highest decline in turnout, of seven percentage points, was in 1962 (from 62.2% in 1957 to 57.4%), and the highest increase of approximately eight percentage points was in 2014. It is true that in some states and constituencies, the change may be in double digits, but these will be statistical outliers. However, one must also factor in the turnout baseline in different states to make any conclusions based on such large changes. Analysis of election data suggests the turnout rates in north and central states have rarely crossed 70% in the last three decades, whereas the southern and eastern states continuously register above 80%. The migration of a significant portion of the working-age population from Hindi-belt states is an important reason for this gap.
The literature on turnout from various countries indicates that temperature shocks, the length of the election schedule, and voter fatigue or apathy with the political process have very weak correlations with turnout figures. Political competition, i.e., uncertainty around the winner is positively correlated with turnout rates. And, cadre-based parties are much better at mobilising their voters even in low-turnout scenarios.
In India, the turnout rates for assembly elections are higher than those for the general elections. And, in seats where the margins of victory are low, the turnout is higher. However, we do not find any effect of political parties on the turnout. Our knowledge about the variation in turnout rates remains limited because, despite substantial efforts by the Election Commission of India (ECI) to revise the electoral rolls before every election, the presence of names that should not have been on the rolls inflates the denominator and leads to a downward bias for the turnout estimate. This happens because dead voters or those who have permanently moved out of a locality/village remain on the voter list. The National Election Studies (NES) survey conducted by Lokniti-CSDS in 2009 indicated that at least 20% of the registered electors should not have been on the electoral rolls as they either did not live in that locality or were dead. We also need a careful analysis of the effect of Covid-19 deaths on the electoral rolls. There may be state-level variations in the cleaning of the electoral rolls on that account.
Several scholarly analyses of both parliamentary elections as well as assembly elections have shown that there is no relationship between an increase or decrease in voter turnout on the re-election chances of an incumbent party. These findings are robust even after controlling several confounding factors and have been tested using sophisticated statistical tools. So, more than aggregate turnout, what matters for election outcomes is which groups are turning out above/below the average. Researchers rely on survey data to estimate the turnout differential among groups, because the ECI only releases gender disaggregated electors and voters data, not on any other demographic metric, at the constituency level.
To illustrate, the pre-poll survey conducted by Lokniti-CSDS in March 2019 indicated the Bharatiya Janata Party (BJP) had an advantage among the voters who were more likely to turn up at the booths. The data suggested that people from the underprivileged castes, Muslims, and the poor were less likely to turn up to vote in 2019. The pre-poll surveys in 2024 point to a very similar pattern of divergence in turnout preferences among the BJP-leaning and Opposition-leaning voters. This was most likely driven by low uncertainty in the final election outcomes as the voters decided to bear the cost of turning up at the booth based on their ability to influence the results.
It is possible that as the campaign for 2024 progressed after the announcements of candidates (and the war of words between the parties over manifestos), the gap in turnout preferences among the BJP-leaning and Opposition-leaning voters either closed or reversed. This can happen if the organisational machinery of various political parties manage to reach out to their respective voters and convince them of the cost of not turning up. There is no gain in highlighting the differences between both online and offline organisational capacities of the BJP and Opposition parties.
Is the high-pitched negative campaign from both sides post the first phase aimed at increasing the emotive quotient in this election, which has been certainly lower than in the previous two? The emotive elements in the 2014 and 2019 campaigns had a definite role in the BJP’s ability to win seats, with a higher probability where the turnout rise was sharper. How will the decline in 2024 affect the BJP’s chances to repeat its performance? Is the decline higher among the BJP-leaning fence sitters? Is the decline a result of voters’ unhappiness with the BJP government’s performance (but not willing enough to switch sides) or because the party and its candidates have over-relied on Prime Minister Narendra Modi’s ability to draw voters to the booth?
While we will know the results of the 2024 Lok Sabha elections on June 4, any meaningful conclusion on the effect the turnout had on the outcomes would require several days of careful analysis. This would largely depend on the availability of PC-level results broken into assembly segments, and randomly sampled representative post-poll surveys. Till then, we must not get carried away by any turnout-related myths.
Rahul Verma is fellow, the Centre for Policy Research (CPR), New Delhi. The views expressed are personal
Continue reading with HT Premium Subscription
