New Delhi: In a setback to the Election Commission (EC), its pilot poll conducted on Sunday to establish a paper trail for electronic voting machines (EVMs) reported significant errors.
Preliminary results of the EC pilot poll indicated discrepancies between votes polled in EVMs and the paper trail, according to three people involved and familiar with the testing process. Two of them are EC officials who confirmed the mismatch, but did not give any more details. EC will release a comprehensive report on the pilot poll in a few days.
“Even a difference of one vote is not acceptable,” said one of the EC officials, who, like the other EC official familiar with the matter, asked not to be identified given the controversial nature of the findings.
To be sure, the discrepancy does not necessarily vindicate the stand of critics who have argued that EVMs can be manipulated, but raises questions on the efficacy of the back-up system that EC was considering to enhance transparency in the electoral process.
According to an analysis by the Citizens for Verifiability, Transparency and Accountability in Elections (VeTA), an activist group campaigning against EVMs, almost one in 20 votes polled in Delhi, one of the four places where the pilot poll was conducted, didn’t have a corresponding paper ballot. VeTA’s representatives were invited to be part of the election process.
“This definitely is some sort of embarrassment for us. However, these are not issues that cannot be resolved. They are...technical problems which are not difficult to sort,” said the second EC official.
Several political parties, including the main opposition Bharatiya Janata Party (BJP) and the Communist Party of India (Marxist), too, claim EVMs are not tamper-proof and have been demanding a paper backup.
To assuage them, the trial —conducted in Leh (Jammu & Kashmir), Thiruvananthapuram (Kerala), Jaisalmer (Rajasthan) and in Delhi—tested the voter verifiable paper audit trail prototypes made by Bharat Electronics Ltd and Electronics Corp. of India Ltd. The pilot took place in Meghalaya on Tuesday.
The system on trial comprises an interface that connects an EVM to a printer and has a list of candidate details corresponding with the EVM. When a person votes for a candidate on the EVM, a paper ballot with a serial number, name and symbol of the candidate will be printed.
“There were 35,791 votes polled in Delhi, each of which had two paper backups. So of what should have been around 70,000 paper trails, around 3,500 were missing. This means there was an error rate of 5%,” said G.V.L. Narasimha Rao, president of VeTA, and a member of BJP’s electoral reforms committee.
He, however, reiterated that these were largely technical snags and could be corrected. According to Rao, the problem was primarily due to erratic printers and poor management.
For the trial, two kinds of printers were used. In one, the printer was sealed and inaccessible to the voter with a transparent window in the front. Once the vote is cast, the printed ballot remains in front of the transparent window for 5 seconds for the voter to verify it, after which it automatically falls into a sealed box. In the other system, there is an open printer and the voter gets a thermal printout of the ballot, which he then has to drop into a sealed ballot box.
In the pilot, the paper ballots were counted and cross-checked with votes recorded in the EVMs. The error rate was 1-5% for the closed printer and 7-12% for the open printer, according to Rao.
“Of the around 35,000 votes cast in Delhi, differences have emerged for a significant number, largely due to jamming of printers and more discrepancies were found in the open printer system,” the first EC official said.
The pilot was conducted ahead of crucial assembly elections next year, including in Uttar Pradesh and Gujarat.
“The system needs to be tested in a real election, like forthcoming by-elections, since EC was unable to simulate an election in the pilot,” said Rao.