Delhi Polls: Mismatch in Votes in 68 Seats; EC Terms it Minuscule

Discrepancy in votes polled & counted in 68 constituencies in Delhi Assembly polls; EC calls it ‘minuscule’ error.

The Electronic Voting Machines (EVMs) were introduced by the Election Commission of India to bring transparency and accuracy to the election process.

But is this really happening? Are EVMs and VVPATs (Voter Verifiable Paper Audit Trail) recording votes with accuracy?

In the past, The Quint, through its two reports, brought to fore the discrepancies in the data on votes polled and counted in 2018 Madhya Pradesh Assembly elections and 2019 Lok Sabha elections. But the EC had brushed aside the discrepancies on grounds that the data released was “provisional” and not final.

To find out if there were any discrepancies in EVM votes polled and counted in 2020 Delhi Assembly elections, The Quint analysed the Form 20 data of all constituencies. Form 20 provides details about the number of votes counted, number of votes received by each candidate in each polling station, total number of registered voters, number of NOTA votes and tendered votes in each constituency. The Form 20 data is compiled by the election officers.

We found that there was mismatch between the number of EVM votes polled and votes counted in 68 out of 70 constituencies in Delhi. Notably, out of 68 constituencies, in 47 constituencies, the mismatch in number of votes was more than 50.

• The maximum surplus of EVM votes was recorded in Karol Bagh Assembly constituency. The number of EVM votes polled in this constituency was 1,07,228 or 60.44% and the number of EVM votes counted was 1,08,339 leading to a surplus of +1,111 EVM votes.
• The highest deficit in EVM votes was recorded in Najafgarh Assembly constituency. The number of EVM votes polled in this constituency was 1,62,206 or 64.41% and the number of EVM votes counted was 1,61,194, leading to a deficit of 1,012 EVM votes.

Three Constituencies Recording Surplus Votes

Three Constituencies Recording Deficit Votes

Election Commission’s Response

The Quint reached out to Delhi’s Chief Electoral Officer with its query over the discrepancies. The response of Delhi CEO can be summarised in the following three points:

1. There were “very minor difference in voters’” turnout data and votes counted from EVM on counting day. The differences were minuscule in nature and “occurred mainly due to inadvertent human/clerical error in data entry”.
2. In four Assembly constituencies, “the difference (in data) occurred due to non-counting of votes polled in EVM owing to non-retrieval of result from EVM or non-clearance of mock poll data.”
3. “There was no case of any objection from the counting agents or the candidates in respect of any mismatch in polled votes registered in Control Unit of EVM and Form 17C on the day of counting, which shows that there was no discrepancy between the two.”

Interestingly, all three points do not address the issue of mismatch and can be easily debunked.

Could ‘Minuscule’ Difference Affect the Verdict of 2 Constituencies?

EC’s claims that the difference in voters’ turnout and counted data is ‘minuscule’ in nature. But could this ‘minuscule’ difference have affected the verdict of at least two Delhi Assembly constituencies, where the winning margin was less than 1000 votes?

• One, in Bijwasan constituency, Aam Aadmi Party (AAP) candidate won by 753 votes, while the runner-up was the candidate from BJP. There was a deficit of 239 votes in the data.
• Second, in Laxmi Nagar constituency, BJP candidate won by 880 votes, while the runner-up candidate was from AAP. The discrepancy here, too, was a deficit of 86 votes.
• Experts say, that in both the cases, the runner-up candidates have a strong case to file an election petition in the court based on the Form 20 data.

Polling Stations Manned With Four Persons

Each polling station is manned by four election officers comprising a Presiding Officer or the head of the polling station, and three other polling officers. On an average, not more than 1000 voters are listed in one polling station because the VVPAT cannot record more than 1200 votes.

Now, if one goes through the Form 20 data, one will find that in 95% polling stations, less than 500 votes were cast.

Not just that, as per the Election Commission’s manual, the Presiding Officer is supposed to record voter turnout or votes polled in his polling station every two hours in his diary, and simultaneously inform the Returning Officer of the constituency about the same.

When the election process is well-defined, then why do we witness discrepancies in the votes polled and counted?

“Form 20 is issued by the Commission after due diligence of the officers involved in the election process. At least, I have not witnessed such discrepancy when I served in EC. Clerical error can happen in a few constituencies but not in almost all. If the discrepancy is due to technical error then EC should address it immediately rather than coming up with lame excuses.”

Former employee of Election Commission of India

Despite Clerical Mistakes, EC Says Data on Votes Polled to be Accurate

The EC’s announcement, declaring the voter turnout data for the Delhi Assembly elections was delayed by 24 hours. EC’s explanation for the delay in the announcement was that it ‘wanted to avoid speculation.’

“They did not want to speculate and wanted to give exact figures... So, returning officers worked throughout the night to check data to ensure it is accurate. It has taken a little bit of time but, in data entry, it is very important to ensure accuracy,” said Ranbir Singh, Delhi’s Chief Electoral Officer.

Given that the EC ‘ensured accuracy’,  the mismatch in the data comes as a surprise. Moreover, how could clerical mistakes happen in not one or two, but 68 constituencies?

EC had earlier said that there was a delay in the announcement of votes polled data because they wanted to ‘ensure accuracy’. They are now saying that the inaccuracy in votes polled and counted data is because of clerical mistakes. At least one of the two claims appears to be misleading.

If EVM Data Not Retrieved, Then Why VVPAT Not Counted?

The Form 20 – which provides polling station-wise details of votes counted – shows that EVMs of all polling stations were counted. Clearly, no EVM was set aside. Nor were votes not counted because of some technical error. Further, mock poll votes were also not deleted.

Secondly, only 50 votes are cast in EVM-VVPAT during the mock poll on the day of polling. In that case, the surplus votes shouldn’t exceed 50 votes. But in  three constituencies, the surplus votes exceeds the 50 votes-count. How, then, will Election Commission explain this?

Thirdly, if EVM data of four constituencies was not retrieved or counted, then as per the rules, VVPAT slips should have been counted. Was this done? If not, why?

What about the deficit in EVM votes count? Were votes deleted from the EVMs after polling was over, resulting in the deficit of votes?

“This discrepancy between votes polled and counted continues unabated, which means that there are some infirmities in the system that the EC is either not able to or is unwilling to identify and deal with. It is actually very necessary to solve this mystery once and for all.”

Jagdeep Chhokar- Member, Associated Democratic Reform

Based on The Quint's report on the mismatch in votes polled and counted in 2019 Lok Sabha election, a petition was filed in the Supreme Court. The Court issued notice to the Election Commission of India seeking an explanation. The matter is still pending.

Inspite of a petition in the SC, why is the EC not ensuring accuracy of data in the election process?

The Quint has written to the office of Delhi CEO, seeking further answers about their response on the issue. We will update the article as and when we receive their reply.