Parakala Prabhakar 
Looking at the discrepancy between votes polled and the votes retrieved from the Electronic Voting Machines and counted, from the data released by the Election Commission of India for the 2024 general elections and the recently concluded assembly elections, one cannot avoid the strong suspicion that there is something seriously wrong with the poll body’s functioning, accountability, and transparency.
Being a constitutional and public body charged with the responsibility of conducting free and fair elections, it is incumbent on the EC to clarify its position on the discrepancies cogently, logically, and with robust documentary evidence.
Anything short of that would raise serious doubts about the integrity of our electoral process. The matter goes straight to the heart of our democratic polity and could even raise nasty suspicions that electoral mandates in the country are being stolen.
Doubts were raised about the integrity of poll data by civil society groups like Vote for Democracy (VFD) and Association for Democratic Reforms (ADR), as well as the main opposition party in the country. The EC, however, chose to dismiss the misgivings, describing them at various points as ‘insinuations’, ‘unfounded allegations’, and worse, ‘malicious’. The EC congratulated itself and self-certified the conduct of elections as conforming to ‘gold standards.’ It ignored the concerns aired in the Memorandum submitted to it by Citizens Commission on Election earlier. It seems to think that elections are an affair only between itself and political parties. Citizens’ concerns for transparency do not seem to matter to it.
For the moment, let us examine in some detail the Maharashtra state assembly voting data released by the EC. The concerns that emerged regarding the 2024 Lok Sabha polls and Haryana assembly elections are not dwelt on in detail here.
Maharashtra went to polls in one phase, on November 20 this year. A release by the EC on November 20, at 6:14 pm, (No. ECI/PN/163/2024) says:
“Polling concludes peacefully in assembly elections in Maharashtra and Jharkhand.” It then adds, “All 288 ACs in Maharashtra vote in single phase with a turnout of 58.22% as of 5 PM.” From this percentage given by the ECI one arrives at the gross number of votes polled as 5,64,88,024. The release also says that “Turnout figures will be further updated at – 11:45 pm.”
Then in a release issued at 11:53 pm (ECI/PN/164/2024), on the same day, the ECI says:
“Polling for 288 Acs in Maharashtra and 38 Acs in Jharkhand which recorded an approximate voter turnout of 65.02% and 65.45% respectively, as per updates till 11:30 pm…”
The revised figure of 65.02% for Maharashtra amounts to 6,30,85,732. The difference between the figure given at 6:14 pm and at 11:53 pm on November 20 amounts to 65,97,708 in gross numbers.
On page two of the same release (issued at 6:14 pm on November 20), the ECI says:
“Polling is still going on in certain polling stations where voters were waiting in queue before the close of poll hours” (emphasis added).
Do note that the ECI itself says voters were waiting at certain polling stations. Not in all the polling stations.
And then, finally, on November 22, a few hours before counting, the Voter Turnout App of the ECI which gave constituency-wise polling data for all the 288 assembly constituencies put the total gross votes polled at 6,40,85,095 – up from 6,30,85,732 – and the final polling percentage at 66.05%.
So, from 58.22% at 6:14 pm on November 20, polling percentage goes up to 65.02% at 11:53 pm, and finally before counting it settles at 66.05%.
In total, the jump is 7.83%. In gross numbers, the vote surges from 5,64,88,024 at 5:00 pm according to the 6:14 pm release to 6,30,85,732 at the close of day according to the 11:53 pm release of ECI. And finally, on November 22, a few hours before counting, the gross voting again goes up to 6,40,85,091. That is an increase of another 9,99,359 votes.
All these hikes add up to a staggering 75,97,067 votes.
It needs to be seen if the hike is on account of extraordinarily large numbers of people waiting in the polling stations’ premises. We will return to this in a while.
The EC handbook lays down that all voters who reach the polling station premises before the closing time are allowed to vote. There is no time restriction for them, and they will be allowed until the last voter has exercised their right.
The EC’s Handbook for Presiding Officers 2023 (Section 7.1.1) deals with such situations. It says:
“…that all electors, who are present at the Polling Station at the hour appointed for the close of poll, should be permitted to cast their votes. In such [a] case, poll has to be continued beyond the appointed closing hour till everyone has voted.”
Let us examine the data from Nanded Lok Sabha constituency which went for a byelection. The polling figures for that Lok Sabha constituency and the nine assembly constituencies within its purview present a huge conundrum. The ECI should be able to help us solve it.
For Nanded Lok Sabha election, the total votes polled in the EVMs are 12,93,624. Postal ballots are 15,996. They together total to 13,09,620.
But the polling figure for the nine assembly constituencies (bearing numbers from 83 to 91), for assembly candidates is 19,35,744 in EVMs. Postal ballots are 22,296. That adds up to 19,58,040.
Therefore, the difference between votes cast separately, but in the same voting enclosure in the same room of the polling station, for the Lok Sabha by-poll and for the nine assembly constituencies within its purview is 6,42,120.
The Congress candidate for the Nanded Lok Sabha polled 5,78,264 EVM votes and 8,524 postal ballot votes, totalling to 5,86,788.
The BJP candidate polled 5,78,921 EVM votes and 6,410 postal ballot votes totalling to 5,85,331.
The Congress candidate won the seat by a thin margin of 1,457. However, all the nine assembly seats were won by the alliance led by BJP.
The gross voting in all the nine assembly constituencies under Nanded Lok Sabha was 19,35,744 in EVMs, and 22,296 in postal ballots. This totals to 19,58,040. The difference, therefore, between the EVM votes polled for assembly and Lok Sabha is 6,42,120. This is how many more EVM votes were recorded for assembly constituencies.
The disparity detected in this case raises troubling questions.
Of the 19,35,744 voters who exercised their franchise for assembly, only 12,93,624 chose to vote for the Lok Sabha. That means about 32% of those who voted for assembly candidates abstained from voting for the Lok Sabha candidates. That amounts to 6,42,120 voters.
If that is the case, we need to have a record of that.
The EC manual lays down that if a voter enters the polling station, has their signature/thumb impression taken, has their finger marked with indelible ink, and then refuses to vote, that refusal needs to be recorded by the presiding officers in the remarks column of Form 17A, appending their full signatures. And that number must necessarily be reflected in Part 1 of Form 17C under Rule 49(O).
That means, the ECI ought to be in possession of 6,42,120 refusals by voters who said no to voting for Lok Sabha candidates after having exercised their franchise for the assembly election.
Did the ECI have those refusals recorded?
Can it then make those forms that recorded the refusals available to the public for inspection?
When polls to both assembly and Lok Sabha are held together, historically, there has never been a disparity as high as nearly 32%. It is, however, possible that a voter prefers one party or candidate for the assembly and another one for parliament. But never has there been such a staggeringly high disparity between the gross votes cast for assembly and parliament in an election held together, where votes are cast in the same polling station with the voting machines or ballot boxes for the Lok Sabha and assembly polls positioned quite near to each other.
To give another example: Let us now look at the constituency-wise votes polled as released by Maharashtra chief electoral officer on the Voter Turnout App for numbers of Akot assembly constituency. It says that in that constituency, total votes polled in EVMs were 2,12,690.
At counting, however, the number of votes retrieved from the EVMs turned out to be 2,36,234. That is a massive disparity of 23,544 votes. The BJP candidate won the seat by a margin of 18,851.
In the entire state of Maharashtra such discrepancy between votes polled as announced by the Voter Turnout App of ECI and the number of votes retrieved from the EVMs at counting centres can be found in at least 95 constituencies. In these constituencies, the strike rate of BJP and its allies is high.
Recall what the first release on November 20, issued at 6:14 pm said. It said, voters were still waiting at certain polling stations. That surely must mean, not in all the booths were the voters waiting. So, the distribution of nearly 76 lakh votes uniformly across all the 1,00,186 polling stations at the rate of 76 votes per booth cannot be a sustainable explanation. It does not hold water. But the Chief Election Officer of Maharashtra in his interview to The Hindu (dated December 2) suggests this improbable and absurd possibility to explain away the huge surge of nearly 76 lakh votes.
This leads us to ask the question: which are the polling stations at which voters waited in queues after the polls officially closed?
We must note here that the first press release by the ECI on November 20 at 6:14pm said: “Metropolitan voters fail again…” It means that the ECI on its own admitted that places like Mumbai and other cities in the state did not witness any huge surge. It also means that the huge surge, if at all, happened in the non-metro constituencies. But the Maharashtra CEO’s claim in The Hindu interview is contrary to this.
There is a set procedure of recording the exercise of franchise by those who enter the polling station premises before the polls close officially and wait in queues even after the official closing time. The ECI’s Handbook For Presiding Officers, 2023 (Edition – 2) lays down that the presiding officer shall:
“Distribute to all such electors slips signed by you in full, which should be serially numbered from serial No.1 onwards according to the number of electors standing in the queue at that hour. The last elector should be given slip no.1 and next voter in front of him/her shall get slip no.2 and so on. Continue the poll even beyond the closing hour until all these electors have cast their votes. Depute police or other staff to watch that no one is allowed to join the queue after the appointed closing hour. This can be effectively ensured if the distribution of slips is commenced from the last elector standing in the queue and proceeded backwards towards its head.”
In the light of this norm clearly laid down in its own Manual, it will not be difficult for the EC to tell us at how many booths voters had waited in queue after the official closing of the poll. It can also tell us exactly at which booths (i.e. their locations) they waited, and their serial numbers. This way, the size of the surge and its locations can surely be accounted for and verified beyond doubt. Mere assertion without evidence from EC officials does not wash.
The EC cannot just claim any number of voters as having waited to cast their votes. There is another guardrail procedure to prevent any fraudulent or whimsical claim.
In the guidelines on the ‘Videography of Critical Events & in Polling Stations,’ it is clearly laid down that among other things like positioning of voting compartment and presence of polling agents, video should be made of:
“Voters waiting outside at the close of [the] scheduled hour of poll and the last voter in queue.”
Because of the slips that are mandatorily distributed to the waiting voters, their number can be easily available from each polling booth. Videography could verify if they were actually present. The guidelines also lay down that the faces of people being videographed in this process should be discernible.
Video evidence combined with the number of slips distributed to waiting voters can make verification of the last-minute turnout figures unimpeachable. The EC can surely do this in the interests of transparency and accountability.
But is it willing and prepared to do it?
If it is willing to submit itself to this verification, we can conclude that it does not have anything to hide and be sure that there was no manipulation. Otherwise, suspicions about the EC’s integrity will only deepen.
In the 2024 Lok Sabha and assembly polls, wherever abnormal vote hikes took place, the outcome favoured the NDA. When there was a low or negligible hike, it fared badly.
In Uttar Pradesh, for example, the hike in the Lok Sabha poll was 0.21%, 0.34%, 0.23%, 0.01% and 0.25% for phases 3, 4, 5, 6, and 7, respectively. With this negligible hike, NDA numbers dwindled to 36 from the previous tally of 62 seats.
In Haryana assembly polls, in 10 districts which had over 10% vote hike, the NDA wrested 37 seats out of 44. In the rest 12 districts where the hike was tiny, NDA won only 11 seats out of 46. In that state, a 2.25% vote hike took place 50 hours after the end of polling time, and just 13 hours before counting, at 6:39 pm on October 7. Can it be a mere coincidence that 24 more seats went to BJP and two to Indian National Lok Dal after this last-minute hike?
In the first phase of Jharkhand assembly polls, the detected vote hike is 1.79%. The BJP won 17 out of 43 seats that went to polls in that phase. In contrast, in the second phase that had only a tiny hike of 0.86%, the party won a mere seven out of 38 seats.
It is strenuous to even the most credulous citizen to believe that these are mere coincidences. A similar pattern was evident in the Jammu and Kashmir assembly polls too.
Can we expect all the right and convincing answers from the ECI?
On that depends the legitimacy and integrity of the political mandate. On that, too, hinges peoples’ trust in our electoral democracy.
If the ECI cannot, or does not, address these concerns and provide robust proof that there are no irregularities, doubts will linger about the sanctity of the mandate in Maharashtra.
There is a high probability that the government that takes oath in Maharashtra will be seen as one born of a stolen mandate. If someone alleges that the political leadership that is going to sit in the Mantralaya is illegitimate, it will be difficult to refute that allegation.
The burden of proof to refute such allegations, therefore, rests on the EC.
Parakala Prabhakar is a political economist and author of The Crooked Timber Of New India.