Electoral reforms are back in vogue, with a particular emphasis on electronic voting machines (EVMs) and internet voting for overseas citizens. Both technologies have been around for some decades now, but have been dogged by a troubled history of security flaws and vulnerabilities, some of an extremely serious nature.
As a result, an interesting paradox has emerged: even as developing countries — such as Namibia, Nigeria, Kenya, and Bangladesh in recent years — are eagerly hopping on to the EVM bandwagon, technologically advanced nations — including the Netherlands, US, Canada, Norway, Germany and Ireland — are rejecting machines en masse and reverting to paper elections.
The nub of the matter is that EVMs were originally designed to automate elections, not secure them.
Traditional voting machines are no panacea for electoral fraud. EVMs are great at tallying thousands of votes in minutes but, as numerous researchers have demonstrated time and again, are notoriously poor at preventing fraud. In certain cases, machines are actually significantly more vulnerable to rigging than paper.
Electronic voting machines can have numerous benefits but are no panacea for electoral fraud. In fact, they may be more vulnerable to rigging than paper. Their deployment should not be rushed
Def Con, the world’s leading security conference, now organises a yearly ‘Voting Village’ event, where it is routine to witness experts hack into voting machines within mere minutes. It is no wonder that voting machines are controversial.
India, a global leader in EVM deployment, has staunchly resisted this trend. This is because of two main reasons. First, the Indian EVM is a small homegrown piece of self-contained hardware, more akin to a calculator than a computer, and is therefore considerably harder to hack. Second, since 2010, the Election Commission of India (ECI) started to augment EVMs with paper trails of votes. The logic here is that, if the EVM is compromised, a paper back-up of the voting record is still available to tabulate results.
But even in India, a heated backlash against EVMs has erupted over the last two years, since the BJP came to power in a significant election upset. In just the last couple of months, the Congress Party has started a concerted campaign to get the ECI to switch to paper ballots, because EVMs supposedly no longer inspire confidence in the electorate.
Former minister and Congress leader Navjot Singh Sidhu insists that BJP would not win a single seat in Punjab if EVMs are replaced with paper. Other parties in Punjab are also calling for a return to paper. These cases are of particular interest to us, because India is a role model for the developing world with regard to EVMs.
For a thorough accounting, one can turn to the Citizens Commission on Elections (CCE). This Indian civil society organisation, headed by senior judges, economists, engineers, and activists, recently published the first of a series of authoritative and critical reports, calling on the ECI to revamp the vaunted Indian EVM system, and deploy new technology to provide auditing and transparency capabilities to citizens. Given the shiny aura of the Indian EVMs, such a call would have been unthinkable a decade ago.
In this comprehensive document, they raise several points of concern: for instance, local and international experts have demonstrated publicly that Indian EVMs are not ‘unhackable’. Malfunctions which inexplicably flip votes are not uncommon. For instance, in polls in 2018, the ECI had to call for 15 percent extra EVMs, to compensate for faults because of extreme heat, light and dust.
More worryingly, machines can easily be accessed by unauthorised or dishonest actors. Paper trails can be switched just as easily as ballot boxes. The CCE cites reports of machines being moved into strong rooms just before counting of results, and how machines are transported in unofficial vehicles without security officers and proper protocol.
Machines have been known to be lost or stolen. In a prominent security lapse, EVMs were found in a hotel in Bihar in 2019. Activists have launched public interest litigation after discovering that a staggering 2 million EVMs may be unaccounted for in the ECI’s inventory.
These scandals aside, a more essential contribution of the Citizens Commission report is to include depositions in their report — by a host of impressive local and international domain experts — which deflate many of the vaunted security claims of the ECI. Machine audits and safety protocols publicly touted as rigorous turn out to be insufficient and ad hoc.
The CCE notes that there is no vigorous dispute resolution protocol in case of VVPAT (voter verified paper audit trail) problems. ECI members themselves do not possess strong credentials in information security and lack the means to verify the software on the machines. Most importantly, the machines are not auditable — they are literally black boxes, their inner workings are not transparent to election officials or the average voter.
The major security guarantee is the paper trail verification protocol. This suffers from small sample sizes and flawed statistics. In 2019, 21 Indian political parties had to approach the Supreme Court to mandate more rigorous checks on the paper trail, which they believed was currently little more than an ‘ornamental’ exercise.
What are we to learn from this situation? That application of election technology is not straightforward at all. It doesn’t make all problems disappear. Sometimes it even creates new problems.
If implemented successfully, the prospects can be ground-breaking. Research indicates that EVMs in India and Brazil have transformed political and social landscapes significantly for the better. Machines are more user friendly than paper ballots, marginalised and uneducated groups in particular find it easier to register their choices, and their participation has measurably increased.
Reduction in electoral fraud has loosened the stranglehold of corruption, made elected officials more responsive to the public in some regions, and resulted in significant cost savings. These factors are reflected in some very interesting development trends.
For instance, data from certain states in India show that the provision of electricity increased in correlation with deployment of EVMs. In Brazil, reduction in election spending was redirected to healthcare, where Princeton University researchers discovered that infant health outcomes improved in correlation with phased deployment of voting machines.
But many an authoritative report cautions that new election technology has to be implemented cautiously and systematically, with an eye for every little detail, and adapted carefully to every country’s unique ground realities.
There are numerous cases of deployments gone sour, of political expediency trumping common sense, and obscene amounts of money being wasted, leading to embarrassing and costly white elephants. We would be wise to return to a compelling report on the topic of electoral reforms in Pakistan, authored in 2011 by Democracy Reporting International, a Berlin-based NGO.
The authors of the report urge the need for careful deliberation on EVMs, stakeholder consultation, and extensive pilot-testing and small-scale deployments. They note that a successful large-scale rollout typically takes from four to eight years. And new technology should certainly not be introduced in the middle of a major election.
More importantly, they caution against the tendency to think of EVMs as ‘silver bullets’, which will somehow magically fix all the complicated systemic and human problems in our electoral system. The authors are very explicit: “Do not introduce technology to compensate for poor procedures.” Problems like vote-buying, voter intimidation, corrupt polling staff, etc. may be much better addressed with legal solutions.
These are very real concerns that demand a frank and forthright dialogue. We, in Pakistan, have yet to start this dialogue in earnest. This was the key mistake made with the internet voting solution we developed for overseas citizens in 2018. This ambitious and hastily assembled system completely overlooked the research literature and global internet-voting experiments, and was plagued with basic and critical security flaws. We must not repeat this mistake.
Election technology is an uphill climb. It is certainly a climb worth making — but only with our eyes fully open.
The author teaches at NUST. He has a postdoc in election security and advises the government and the ECP on election technology.
He can be reached at firstname.lastname@example.org
Published in Dawn, EOS, March 28th, 2021