SC Notifies EC on Plea Seeking Cross-Verification of Votes in 30% of All Polling Booths

The petition said the current verification process on just one polling booth per assembly constituency was “manifestly arbitrary, irrational and unreasonable".

New Delhi: A Supreme Court bench, headed by Chief Justice Ranjan Gogoi, has issued a notice to the Election Commission seeking its response to a plea that demanded counting and cross-verification of Voter Verifiable Paper Audit Trails (VVPAT) in at least 30% of all polling stations within an assembly constituency.

The petition was filed by former civil servant M.G. Devasahayam, former IFS officer Kalarickal Pranchu Fabian and retired bank officer Thomas Franco Rajendra Dev. It sought directions to the EC to make it mandatory for the returning officer to order hand-counting of all the VVPAT slips in a constituency if the margin of victory was found to be less than 3%. It also demanded mandatory hand-counting of slips even if, irrespective of the margin of victory, the discrepancy between the EVM and VVPAT tally was over 1%.

EC stand based on analysis by two statisticians

The petitioners also noted that their representatives met the Chief Election Commissioner on August 24, 2018 and then again on October 4 to raise the demand. However, the commission justified its decision to only randomly cross-verify one polling booth per constituency based on a research paper by noted statisticians Dr Ajay G. Bhat and Dr Rajeeva Karandikar.

The petitioners submitted before the apex court how several “false assumptions and pitfalls” were relied on by Bhat and Karandikar in their analysis. “The same have been summarised in a note prepared by Dr S.K. Nath, former Director-General, Central Statistical Organisation, Government of India,” they submitted.

Petition cites ex-CSO chief to seek higher VVPAT verification

Referring to Nath, the petitioners said he came to the conclusion that “the sampling size of one booth per constituency only gives a 60% confidence value of a mischief-less and/or a flawless, unbiased election and that to achieve a 95% to 99% confidence level – as is necessary for the functional integrity of the electoral process – at least 10-30% of all the EVMs/polling stations must be selected randomly for cross-verification and tallying EVM counts.”

Also read: Telangana Polls: Mass Deletion of Voter Names Reported, EC Admits Mistake

Raising a question of law, the petition contended that such cross-verification and counting of VVPAT slips was essential in the interest of upholding democratic principles. They said such principles demand that there should be transparency in elections so that “the general public can be satisfied that their vote is correctly recorded and counted.” Also, they said, “all essential steps in the elections ought to be subject to public scrutiny.”

Elections should be perceived as free and fair

The petitioners said ordinary citizens should be “able to check the essential steps in the election process without special expert knowledge”. Finally, they said, there “ought to be verifiability in the counting of votes and ascertainment of the results reliably – without any special technical knowledge.”

The petitioners said “elections must not only be free and fair, but also be seen to be free and fair.”

‘Low cross verification violative of Article 14’

The petitioners said the EC’s decision “to confine such a cross verification exercise to a statistically insignificant number of approximately only one polling station in each constituency, regardless of the number of polling stations in such a constituency,” was “manifestly arbitrary, irrational and unreasonable.”

As such, they said, this decision of EC was also “inter alia, in violation of Article 14 of the Constitution of India.”

The petition said: “India is the largest electoral democracy in the world with more than 900 million registered voters.”

It said the constitutional mandate of superintendence, direction and control of elections to parliament and state legislative assemblies is vested with the EC.

SC directions on VVPATs came in 2013 

The petitioners recalled that in 2013, in Subramanian Swamy v Election Commission of India, the apex court held that in the interests of transparency and voter satisfaction, it is imperative that the EC implement a mechanism of VVPATs. This was necessitated so that a voter could satisfy himself that his vote has been accurately recorded and that should there be any question as to the accuracy of the electronic count, the paper count would be available to corroborate the same.

Therefore, the petitioners submitted that the larger principles of structural due process, transparency, voter confidence and satisfaction require the EC to conduct random cross-verification of EVM counts with VVPAT counts in order to not only detect any technological mischief or mal-programming of EVMs, but also deter such mischief.

VVPATs verified for only 0.3% EVMs in Gujarat, 0.9% in HP polls

They said the EC recognises this and has been implementing such a random cross-verification process in the recent state elections. The petition pointed out that while VVPAT verification was carried out during the Gujarat and Himachal Pradesh assembly elections in December 2017, only 0.4% and 0.9% of the EVMs in the two states were covered.

As such, they submitted that “in order to detect and deter any such mischief or bias in the EVM process, at least 30% of all polling stations in a constituency must be chosen randomly for EVM counts to be cross-verified with VVPAT counts.”

Also read: How Did the EC Link 300 Million Voter IDs to Aadhaar In Just a Few Months?

Stratified sampling of votes urged

The petitioners also stated that “it is also desirable that a stratified sampling be performed with a random sample of one or more polling stations drawn from each of the following ‘strata.”

They suggested that the polling stations could be divided on the basis of classification such as “urban (middle/ upper class); urban (poor/slums); semi urban or rural”. Likewise, they said they could be classified on the basis of “Dalits/minorities; remote hilly/desert/forest areas.”

Finally, they also suggested that the classification could also be on the basis of turnout: “Those with very heavy voter turnout (>80%); with moderate voter turnout (50%-80%); with low voter turnout (<50%); those for which a large number of complaints have been received, etc.”