(This story was first published on 10 June and is being republished from The Quint's archives in light of the Election Commission of India's tweet calling reports on such discrepancies a 'design to discredit largest elections ever held in the history of mankind'. If you want us to do more such investigative stories on transparency and voter rights, please become a member and support our work.)
5,54,598 votes polled by the Electronic Voting Machine (EVM) in 362 constituencies across states were trashed by the Election Commission of India (ECI) as results for the 2024 Lok Sabha elections were declared on Tuesday, 4 June.
Additionally, the ECI also registered a surplus of 35,093 EVM votes in 176 constituencies.
An investigation by The Quint into two sets of data released by the poll body — first on voter turnout or the absolute number of votes polled on the EVMs and the second on the number of EVM votes counted in each constituency on the result day — has found discrepancies in 538 out of 542 Lok Sabha constituencies.
Which means that in these constituencies, the number of votes polled on the EVMs did not match the number of EVM votes which were counted on the result day. (This comparison does not include postal ballots, since the turnout data only considers EVM votes polled.)
In at least 267 constituencies, this difference was more than 500 votes.
Let's Understand With The Help of Examples
In Tamil Nadu's Tiruvallur constituency, which voted on 19 April in the first phase of polling, 14,30,738 EVM votes were polled as per turnout data released by the ECI on 25 May. On the counting day (4 June) 14,13,947 EVM votes were counted — a deficit of 16,791 votes.
In Assam's Karimganj constituency, which voted on 26 April in the second phase of polling, 11,36,538 votes were polled as per the ECI data. And again, on the day of the result (4 June) 11,40,349 votes were counted — a surplus of 3,811 votes.
Copies of phase-wise turnout data released by the ECI can be accessed here.
While there's been no specific explanation by the poll body on what caused this mismatch between the two sets of data, the Uttar Pradesh (UP) Chief Electoral Officer took to X (formerly Twitter) to explain the cause of discrepancies in UP.
"The difference can arise between votes polled and votes counted because there are certain polling stations whose votes polled are not counted as per the extant protocol issued by the Commission and provided in various Manuals and Handbooks," the CEC wrote.
He further explained the two scenarios in which the number of votes counted can be less than the number of EVM votes which were polled.
First, where the Presiding Officer by mistake fails to clear the Mock Poll data from the Control Unit before starting the actual poll or he fails to remove Mock Poll slips from the VVPAT before starting the actual poll. And second, where the total votes polled in the Control Unit does not match the record of votes in Form 17-C prepared by the Presiding Officer who records incorrect number by mistake.
For context, on the day of polling, mock polls are held at 5 am to check whether the EVMs are functioning correctly. Normally, mock polling sees 5 votes polled for each candidate, the results of which are checked later. Ideally, polling officers are required to clear the Control Unit before actual polling starts.
Form 17C, which the CEC mentioned in his post, is a polling-booth-by-polling-booth comprehensive voting record. It contains information such as the number of voters assigned to each polling place, the total number of voters registered in a given area, the number of voters who choose not to cast a ballot, the number of voters who are denied the opportunity to vote, the total number of votes polled by Electronic Voting Machines (EVMs), and details regarding ballots and paper seals.
The CEC, however, did not clarify circumstances under which surplus votes can be counted.
Constituencies Where Surplus EVM Votes Were Counted
In Karimganj, where 3,811 surplus votes were counted, Bharatiya Janata Party (BJP) candidate Kripanath Mallah registered a victory with a margin of 18,360 votes.
Jagdeep Chhokar, co-founder of The Association for Democratic Reforms (ADR), a nonprofit organisation working on electoral and political reforms in India, told The Quint that the Election Commission should come up with "constituency specific explanations" for these discrepancies.
"So far, the Election Commission has only given a generic explanation for EVM vote surplus or deficit, that too on Twitter. The poll body needs to give specifics here. This makes an even stronger case for the EC to make form 17C available in public domain. We're not casting doubts on the result of the elections but there needs to a transparent and robust mechanism for counting of votes," Chhokar said.
Among other constituencies where surplus votes were counted are Ongole in Andhra Pradesh, Odisha's Balasore, Mandla in Madhya Pradesh, and Buxar in Bihar.
In Odisha's Jajpur, where 809 surplus votes were counted, BJP candidate Rabindra Narayan Behera won with with a margin of only 1,587 votes.
Constituencies Which Recorded a Vote Deficit
In Assam's Kokrajhar constituency, 12,40,306 votes were polled and 12,29,546 votes were counted — a deficit of 10,760 votes. The United People's Party candidate Joyonta Basumatary won the seat with a margin of 51,580 votes.
Similarly, in Dhenkanal in Odisha, 11,93,460 votes were polled and 11,84,033 votes were counted — a deficit of 9,427 votes.
In UP's Aligarh, where BJP's Satish Kumar Gautam won by a margin of 15,647 votes, 5,896 votes were discarded.
To check the scale of mismatch in your constituency, click on it on the map below or find in the search bar.
(Infogram by: Naman Shah)
'Election Commission Needs To Be More Transparent'
A story by The Quint exposed similar discrepancies in EVM votes polled and counted during the 2019 Lok Sabha election. After the story, The Association For Democratic Reforms filed a petition in the Supreme Court demanding "a court order directing the Election Commission of India (EC) to conduct actual and accurate reconciliation of (votes) data before the declaration of the final result of any election".
On 26 April, the Supreme Court dismissed this and several other pleas seeking complete verification of the Electronic Voting Machines (EVMs) with the Voter-Verifiable Paper Audit Trail (VVPAT) slips.
This after the EC told the Court that "no mismatch was detected" between the votes counted in the EVM and VVPAT slips after they verified over 4 crore slips.
"This whole issue is central to the trust in our electoral process," said transparency activist Anjali Bharadwaj. She is a founding member of the Satark Nagrik Sangathan, a citizens’ group working to promote transparency and accountability in government.
"If there is a mismatch between the data on number on number of (EVM) votes polled and the number of (EVM) votes counted, it is bound to create doubts in the minds of people. This only further ascertains the need for the ECI to upload complete form 17C, including part two which has data on results declared, in the public domain. Technically, there should not be a mismatch because these are the numbers that are read off the machines but if there is a mismatch, EC should come up with granular constituency-wise explanation," she added.
Video Editor: Kriti Saxena
Camera: Swagatam Roy
(The Quint has reached out to the Election Commission of India for a response. This story will be updated as and when we hear from them.)
(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)