VVPAT slip-based audit: EC’s chance to prove prudent or blinkered
By Mohd Haleem Khan
Once I know the proper question, I could solve the problem in less than five minutes,” Albert Einstein pronounced. “It’s not that they can’t see the solution. They can’t see the problem,” GK Chesterton echoed. By these dicta, one can’t escape the conclusion that Indian Statistical Institute (ISI) did not ‘know the proper question’ or ‘see the problem’ when it submitted its report to the Election Commission of India (ECI).
On April 8, the Supreme Court (SC) disposed of the writ petition(c) No. 273/2019 Chandrababu Naidu & others versus Union of India & another. The court noted “the ECI contends that a query had been posed to the ISI, namely, what would be the reasonable sample size of Polling Stations where VVPAT slips verification is required to be carried out to achieve the object of establishing the credibility and integrity of the electoral process.” The SC further noted the ECI averment: “In response, the ISI had submitted an elaborate report, the crux of which is that verification of VVPAT paper trail of 479 (randomly selected) Electronic Voting Machines (EVMs) would generate over 99% accuracy in the election results.” The court took on record: “as per Guideline No. 16.6 i.e. verification of VVPAT paper trails of one Assembly Constituency or Assembly Segment in a Parliamentary Constituency would involve verification of VVPAT paper trail of 4,125 EVMs instead of 479 EVMs which is eight times more than what has been reported by the ISI.”
It is inconceivable any statistics expert can suggest 99% accuracy by taking out more than 88% (479 out of 4,125) assembly segments out of the VVPAT slip based audit. The ECI and ISI somehow failed to comprehend each other’s jargon while trying to define the problem. If one were looking into the infallibility of EVMs per se, the sample size of 479 of the total number of EVMs being used might have been arrived at rightly. But if the problem is, to quote BJP spokesperson GVL Narasimha Rao, “the distrust among political leaders of all hues in voting machines is so high that most losers are wondering if they had been unfairly defeated in polls,” then ISI should have eschewed from seeing the individual assembly segment wise problem in aggregate. It should have suggested statistically consistent sample size assembly segment wise. The number of EVMs to be audited must depend on the total number of machines used in that assembly segment after factoring in the impact of polling booth numbers being more or less.
The Supreme Court took on record following guideline: “…16.6.5. The following procedure shall be followed for the conduct of draw of lots:
1. a) White colour paper cards of postcard size shall be used for conducting the draw of lots.
2. b) Total number of such paper cards should be equal to total number of polling stations in the Assembly Constituency.
3. c) The paper cards shall have pre-printed Assembly Constituency/ Assembly Segment number, AC/AS name and date of polling on the top, and the polling station number in the centre. Each digit of polling station number shall be at least (1 inch by 1 inch) size and printed in black ink.
4. d) The paper cards to be used for draw of lots should be four-folded in such a way that polling station number is not visible.
5. e) Each paper card shall be shown to the candidates/their agents before folding and dropping in the container.
6. f) The paper cards shall be kept in the big container and must be shaken before picking up 01 (one) slip by the Returning Officer. …”
Even the most finicky losing candidate will find it difficult to find loopholes in the above mechanism. The most ludicrous allegation against the returning officer, however, could always be contrived. Somehow the pooling booth where no rigging took place got picked up in the draw of lot, the poor looser may intone. My purpose of writing this is not to add credibility to such far-fetched insinuations of bias. It is to search for a logically sound and transparent way of saving the election process from political brinkmanship. The solution must upfront save the returning officer from allegations of bias and partisanship.
The Supreme Court has observed: “at the very outset the Court would like to observe that neither the satisfaction of the Election Commission nor the system in vogue today, as stated above, is being doubted by the Court insofar as fairness and integrity is concerned. It is possible and we are certain that the system ensures accurate electoral results. But that is not all. If the number of machines, which are subjected to verification of paper trail, can be increased to a reasonable number, it would lead to greater satisfaction amongst not only the political parties but also the entire electorate of the Country. This is what the Court should endeavour … In this regard, the proximity to the Election schedule announced by the ECI must be kept in mind.” The Court concluded “having regard to the totality of the facts of the case and need to generate the greatest degree of satisfaction in all with regard to the full accuracy of the election results, the number of EVMs that would now be subjected to verification so far as VVPAT paper trail is concerned would be five per Assembly Constituency or Assembly Segments in a Parliamentary Constituency.”
It is important that the ECI does not feel circumscribed by the validating observations of the SC. It should travel enough distance to achieve the larger objective of enhancing bipartisan faith in our election process. “…to greater satisfaction amongst not only the political parties but also the entire electorate of the Country,” to quote the SC. Instead of selection of five booths for VVPAT slip based audit by draw of lots by the returning officer, the ECI must seriously consider options that insulate the returning officer from allegations of bias and partisanship—real or imagined.
Lest this opportunity to mitigate the vulnerability of the returning officer and eschew cost and time consuming process laid down in Guideline 16.6.5 is frittered away, the Guideline 16.6.1 and onwards, which in any case are going to be revised in view of the SC decision, be thoroughly revised. After the completion of the last round of counting of votes recorded in the EVMs for identifying the five polling stations per assembly constituency/segment, the names of two polling stations from the leading candidate and three names from the immediately behind trailing candidate be invited. The candidates/or his agent must furnish the same to the returning officer concerned, in front of the general observer appointed by the commission for that constituency. The returning officer shall give a written intimation regarding the invitation of names of five polling stations for verification of VVPAT slips to the candidates/their election agents well in advance. Since the names of polling stations going through VVPAT slip audit are given by the two candidates who emerged as number one and two in machine-based counting, it shall deny all opportunity to cast aspersions on EVMs. The objective of establishing the credibility of the electoral process shall be achieved and the transparency of the election process enhanced. It shall make the confidence-building measure participatory, too.
(The author retired as secretary to the government of India, ministry of finance. Views are personal)