A research paper written by an economics professor at Ashoka University suggesting that the Bharatiya Janata Party (BJP) won several seats in the 2019 general elections through electoral manipulation has triggered a political row. The controversial paper by Sabyasachi Das, Assistant Professor of Economics at Ashoka university, was published in July this year. But days after it circulated on social media, Ashoka University distanced itself saying the paper wasn’t published in an academic journal yet. The paper is titled ‘Democratic Backsliding in the World’s Largest Democracy’, referring to the phenomenon of a country sliding from a democratic to an autocratic way of functioning. Noting that the BJP won a disproportionate share of closely contested seats in the 2019 general elections – most of them in states where the BJP was in power at the time – the study looks at the possibility of electoral fraud and makes several damning suggestions.
Das concludes that his findings are consistent with electoral manipulation at the stage of voter registration as well as at the time of voting and counting. He also concludes that the results suggest targeted electoral discrimination against Muslims.
Many academics and others have condemned Ashoka University’s stand to distance itself from its own faculty member when it comes to work that is critical of the ruling BJP. “Ashoka values research that is critically peer-reviewed and published in reputed journals… Social media activity or public activism by Ashoka faculty, students or staff in their individual capacity does not reflect the stand of the University,” the university said in its statement.
To begin with, Sabyasachi notes that in the 2019 general elections, in constituencies that were closely contested between a BJP candidate and a rival, BJP won disproportionately more seats than it lost. He notes that such a pattern was never observed in past elections, neither in favour of the BJP nor the Congress. He also notes that these ‘disproportionate’ BJP wins in the closely contested seats were mainly in states ruled by the BJP at the time.
There are two possible reasons for this, Das says — either the BJP committed electoral fraud, or it was able to accurately predict the closely contested seats and mobilise party workers to campaign more intensively, in a phenomenon known as precise control. “[BJP] may have been able to exercise precise control in 2019 since it had significantly built up its organisational capacity in several states, subsequent to its 2014 general election victory,” he notes.
The paper examines data from the National Election Survey (NES) of 2019 which has microdata on election campaigning by political parties, to understand if the party managed to pull off precise control by campaigning more aggressively in the closely contested seats. By measuring door-to-door campaigning in a representative sample of the constituencies, Das says he found that neither the BJP nor any other party campaigned significantly harder in constituencies that BJP barely won.
To investigate election manipulation at the level of voter registration, the study looks at the growth rate of the electorate across constituencies and finds that it was significantly lower in closely contested constituencies with a higher Muslim population, suggesting that there was targeted disenfranchisement of Muslim voters, particularly in Muslim-dominated constituencies where the BJP’s chances were slim.
To examine manipulation at the voter turnout level, the paper looks at constituency-level EVM turnout data from two different official versions — the final count of EVM votes polled as initially released on the Election Commission of India (ECI) website, and the number of EVM votes actually counted. These two figures did not always match. “When the media pointed out the discrepancy, the ECI removed the earlier figures from its website. I access copies of the earlier turnout data to measure discrepancy,” says Das.
Das found these discrepancies to be wider in seats where BJP won by a close margin, particularly in BJP-ruled states. Noting that the magnitude of data revision, however, was smaller than the margin of victory in most of these cases, Das presents more evidence to suggest that the alleged manipulation was likely to have been done locally at the polling stations.
Looking at the ECI-assigned counting observers who monitored the counting of votes in each constituency, Das says he found abnormalities in constituencies with observers from the State Civil Services (SCS), particularly in BJP-ruled states. Das says that state government-appointed SCS officers are more likely to be “politically pliable” and therefore facilitate manipulation of vote counting.
Das also looks at polling station-level election results and finds that in constituencies where BJP won closely, there was a higher spike in BJP’s vote share at that station relative to its voter share in that whole constituency, at stations with a high turnout — suggesting that stations where the turnout was unusually high, the ‘excess’ votes mostly went to the BJP. The spike in BJP’s relative vote share is also higher in constituencies where the EVM data (on votes polled and counted) show a higher discrepancy, the paper says.
Noting that support among Muslims for the BJP was consistently low across the 2014 and 2019 general elections, Das points to abnormalities in the closely contested constituencies with a high Muslim electorate to suggest manipulation.
Finally, it says that based on back-of-the-envelope calculations, in closely contested seats with a winning margin of less than 5%, BJP had “excess” wins ( relative to the benchmark of 50% chance of winning) in about 11 constituencies (the likelihood varying from 9 to 18 seats), and notes that this also was not sufficient to stop the party from forming the government. BJP won very comfortably in 2019 with 303 out of 543 seats.
“The tests are…not proofs of fraud, nor does it suggest that manipulation was widespread. Proving electoral manipulation in a robust democracy is a significantly harder task that would require detailed investigation of electoral data in each constituency separately,” Das concludes.