The high incidence of glitches in the Voter Verifiable Paper Audit Trail (VVPAT) machines in Monday’s by-elections should be a major cause of concern for the Election Commission of India. Fresh polling had to be ordered in dozens of booths in Kairana and Bhandara-Gondiya in Uttar Pradesh and Maharashtra, respectively, as a consequence. Ever since the implementation of the VVPAT system last year, machine malfunction and subsequent delays in polling have been recurring issues. Close to 4.2% of the VVPAT machines deployed in the Karnataka Assembly elections this month developed glitches during the testing as well as polling processes. The overall fault rate was as high as 11.6% in the by-elections held in four parliamentary and nine Assembly constituencies on Monday. The ECI has suggested that these machines were more prone to malfunctioning due to their sensitivity to extreme weather conditions and exposure to light. It also blamed the relative inexperience of polling officers handling them, compared to the ballot and control units for the electronic voting machines (EVMs) that have been in use for much longer. The technical committee of the ECI is now faced with a challenge to ensure that the VVPAT machines hold up, with the general election due next year in the hot summer months. The VVPAT was added to the EVM to audit the voter tallies stored in the machine. Its universal implementation — which began in the Goa Assembly polls in February 2017 — was deemed necessary as many political parties complained about the possible hacking of EVMs. These complaints lacked any basis, but the VVPAT implementation was hastened to bring back trust in the election process. In all elections where it has been used, the VVPAT tallies have matched with the EVM counts, but for a stray case or two when the VVPAT machine was not reset before polling began.
Inadvertently, the use of these machines, which are adjuncts to the ballot and control units of the EVMs, has added to the complexity of an otherwise simple, single programmable-chip based system, and rendered it prone to more glitches. There is enough empirical evidence to show that EVMs have eased polling and helped increase voter turnout since being put to use. But in using VVPAT machines to reassure sceptics about an election’s integrity, the ECI has introduced a new element, and cost, to the process. Considering these challenges, the ECI should consider deploying the VVPAT machines in a limited, statistically significant, randomly chosen set of polling booths. This will reduce the possibility of glitches affecting the polling process as well-tested machines could be deployed (with enough replacements also handy) to such booths. The current verification process, after all, only involves the counting of VVPAT slips by randomly choosing one booth from each constituency (or segment), and this check should not be affected drastically by the new method.
