How (not) to deploy an electronic voting system
We must not let the trauma of the 2013 election fiasco drive us headlong into another potential disaster
The writer is Research Associate at Newcastle University, UK. His research topics include next-generation voting systems, cryptocurrencies and security for emerging networks
The judicial commission inquiry on election rigging is underway and all else election-related has been pushed into the background. However, we can expect the issue of electoral reform to resurface soon along with the all-important question of electronic voting machines (EVMs). The Election Commission of Pakistan (ECP) is currently in the process of formulating its recommendations to the Parliamentary Committee on Electoral Reform (PCER) on the matter. As ECP Secretary Ishtiaq Ahmed Khan has noted, the next general elections may likely mark a switch to electronic voting.
An earlier article of mine outlined the multitude of security problems with EVMs and noted that, in the current Pakistani climate, these machines actually exacerbate the threat of rigging. Here we argue that if we are to commit to a mass EVM deployment, it must be undertaken very carefully with years of pilot projects, trial runs and research studies. In other words, it would be a very very bad idea to deploy a full-fledged electronic voting system in time for the 2018 polls. To make this case, we look at other countries that have experimented with EVMs and ask — what sort of challenges can we expect? What could go wrong? What are the worst case scenarios?
Germany presents a good example of the legal issues ahead. The Bundestag elections in 2005 saw a large deployment of about 2,000 machines in five different states, catering to some two million voters. In 2009, amid growing public distrust of EVMs, political scientist Joachim Weisner and his son, physicist Ulrich Weisner mounted a legal challenge. They argued that the average voter could not verify the inner workings of these machines and therefore needed to place “blind faith” in the technology. And, it had been demonstrated that these machines could easily be tampered with. The Federal Constitutional Court of Germany agreed in 2009, and ruled that electronic voting, as used in the last 10 years was unconstitutional. This marked Germany’s return to paper-based voting.
While this court order does not outlaw all forms of electronic voting but just the current incarnation of EVMs, it does make for an interesting reading and contains subtle insights into the nature of democracy and what election transparency truly means. Arguing that “all essential steps in the elections are subject to public examinability”, the court says that EVM “only meets the constitutional requirements if the essential steps of the voting and of the ascertainment of the result can be examined reliably and without any specialist knowledge of the subject … The very wide-reaching effect of possible errors of the voting machines or of deliberate electoral fraud make special precautions necessary in order to safeguard the principle of the public nature of elections.” The ruling concludes with: “In a republic, elections are a matter for the entire people and a joint concern of all citizens. Consequently, the monitoring of the election procedure must also be a matter for and a task of the citizen. Each citizen must be able to comprehend and verify the central steps in the elections.”
Such legal challenges to EVMs on the basis of poor security and insufficient voter confidence are fairly common worldwide and have put the brakes on electronic voting in many countries. Considering the technical issues ahead, Kenya is a good example of how not to deploy EVMs. The 2007 Kenyan paper-based elections had been highly controversial with widespread electoral fraud triggering mass riots and a round of vicious ethnic cleansing. The police shot hundreds of protestors, some 1,200 people were killed, half a million fled their homes, and ultimately former UN secretary-general Kofi Annan had to step in and negotiate a fragile peace.
Given this backdrop, the shiny new electronic voting system deployed for the 2013 Kenyan presidential elections had a lot riding on it. This new system was reportedly “tamper proof”. Biometric scanners would check voter fingerprints against voter lists in real-time. Election results would be transmitted from polling stations via SMS to the tallying centre and show up right away on a big screen. A European Union (EU) observer noted that this system was more modern than anything implemented in the EU. However, this was a perfect storm. In several instances, fingerprint scanners refused to identify legitimate voters. Election staff forgot PIN numbers and passwords to access the EVMs. Several polling stations did not have electric sockets and batteries started to die out. When the technology failed, several polling centres reverted to back-up paper ballots (which ended up delaying the announcement of election results by a week). A single server, meant to collect results from 33,400 polling stations via SMS, collapsed under the strain. The telco, Safaricom, had advised the election commission to test-run the system thoroughly beforehand but it had only done one small last-minute trial. Returning officers were ultimately chauffeured and even flown down to the tallying centre in Nairobi to communicate results for their polling station in person. Thankfully, there was no violence. Michela Wrong, writing for The New York Times, dubbed this nightmare the “School Socket Syndrome”, essentially “the screech that is heard when modern technology is suddenly grafted onto a developing state’s crumbling infrastructure”. In countries like ours, such ad hoc collapse scenarios are unfortunately all too real.
And these technical problems can have political implications. In December, Poland’s electronic voting system suffered major technical glitches during local elections, delaying results, and leading to widely unexpected outcomes. An estimated 60,000 people marched in protest, including extremist nationalist groups. Polish courts were flooded with more than a thousand legal challenges contesting election results.
What these examples hopefully make clear is that we must not let the trauma of the 2013 election fiasco drive us headlong into another potential disaster. It would be delusional to consider large-scale EVM deployment by 2018. We should start with baby steps: we will most certainly encounter unexpected contingencies which will necessitate a rethink of strategy. If there are any show-stoppers (there will most certainly be legal challenges), it is prudent to let them play out before we invest in tens of thousands of expensive machines. There needs to be vigorous public discussion and citizen education programmes. We need to see how voters react to these machines. Trials should be transparent to observers and media personnel. If we’re truly serious, we should even commission studies by election security professionals and legal experts.
The case of Ireland presents another cautionary tale against rushing into electronic voting. Head of government, Bertie Ahern, impressed with EVMs used in India and France, famously declared that Ireland would become “the laughing stock of Europe because voters still used pencils in elections”. In 2002, Ireland spent 54 million euros in procuring and storing 7,500 EVMs to update their electoral system. An independent commission then, quite predictably, discovered that these machines had dismal security. The machines were never used and they languished in storage, costing Irish taxpayers 140,000 euros a year. In 2009, it was decided that it was too expensive to upgrade them. In 2012, the government finally sold the machines off for scrap for a paltry 70,000 euros — a rate of nine euros per machine — a fitting conclusion to an epic scandal.
Published in The Express Tribune, May 21st, 2015.
An earlier article of mine outlined the multitude of security problems with EVMs and noted that, in the current Pakistani climate, these machines actually exacerbate the threat of rigging. Here we argue that if we are to commit to a mass EVM deployment, it must be undertaken very carefully with years of pilot projects, trial runs and research studies. In other words, it would be a very very bad idea to deploy a full-fledged electronic voting system in time for the 2018 polls. To make this case, we look at other countries that have experimented with EVMs and ask — what sort of challenges can we expect? What could go wrong? What are the worst case scenarios?
Germany presents a good example of the legal issues ahead. The Bundestag elections in 2005 saw a large deployment of about 2,000 machines in five different states, catering to some two million voters. In 2009, amid growing public distrust of EVMs, political scientist Joachim Weisner and his son, physicist Ulrich Weisner mounted a legal challenge. They argued that the average voter could not verify the inner workings of these machines and therefore needed to place “blind faith” in the technology. And, it had been demonstrated that these machines could easily be tampered with. The Federal Constitutional Court of Germany agreed in 2009, and ruled that electronic voting, as used in the last 10 years was unconstitutional. This marked Germany’s return to paper-based voting.
While this court order does not outlaw all forms of electronic voting but just the current incarnation of EVMs, it does make for an interesting reading and contains subtle insights into the nature of democracy and what election transparency truly means. Arguing that “all essential steps in the elections are subject to public examinability”, the court says that EVM “only meets the constitutional requirements if the essential steps of the voting and of the ascertainment of the result can be examined reliably and without any specialist knowledge of the subject … The very wide-reaching effect of possible errors of the voting machines or of deliberate electoral fraud make special precautions necessary in order to safeguard the principle of the public nature of elections.” The ruling concludes with: “In a republic, elections are a matter for the entire people and a joint concern of all citizens. Consequently, the monitoring of the election procedure must also be a matter for and a task of the citizen. Each citizen must be able to comprehend and verify the central steps in the elections.”
Such legal challenges to EVMs on the basis of poor security and insufficient voter confidence are fairly common worldwide and have put the brakes on electronic voting in many countries. Considering the technical issues ahead, Kenya is a good example of how not to deploy EVMs. The 2007 Kenyan paper-based elections had been highly controversial with widespread electoral fraud triggering mass riots and a round of vicious ethnic cleansing. The police shot hundreds of protestors, some 1,200 people were killed, half a million fled their homes, and ultimately former UN secretary-general Kofi Annan had to step in and negotiate a fragile peace.
Given this backdrop, the shiny new electronic voting system deployed for the 2013 Kenyan presidential elections had a lot riding on it. This new system was reportedly “tamper proof”. Biometric scanners would check voter fingerprints against voter lists in real-time. Election results would be transmitted from polling stations via SMS to the tallying centre and show up right away on a big screen. A European Union (EU) observer noted that this system was more modern than anything implemented in the EU. However, this was a perfect storm. In several instances, fingerprint scanners refused to identify legitimate voters. Election staff forgot PIN numbers and passwords to access the EVMs. Several polling stations did not have electric sockets and batteries started to die out. When the technology failed, several polling centres reverted to back-up paper ballots (which ended up delaying the announcement of election results by a week). A single server, meant to collect results from 33,400 polling stations via SMS, collapsed under the strain. The telco, Safaricom, had advised the election commission to test-run the system thoroughly beforehand but it had only done one small last-minute trial. Returning officers were ultimately chauffeured and even flown down to the tallying centre in Nairobi to communicate results for their polling station in person. Thankfully, there was no violence. Michela Wrong, writing for The New York Times, dubbed this nightmare the “School Socket Syndrome”, essentially “the screech that is heard when modern technology is suddenly grafted onto a developing state’s crumbling infrastructure”. In countries like ours, such ad hoc collapse scenarios are unfortunately all too real.
And these technical problems can have political implications. In December, Poland’s electronic voting system suffered major technical glitches during local elections, delaying results, and leading to widely unexpected outcomes. An estimated 60,000 people marched in protest, including extremist nationalist groups. Polish courts were flooded with more than a thousand legal challenges contesting election results.
What these examples hopefully make clear is that we must not let the trauma of the 2013 election fiasco drive us headlong into another potential disaster. It would be delusional to consider large-scale EVM deployment by 2018. We should start with baby steps: we will most certainly encounter unexpected contingencies which will necessitate a rethink of strategy. If there are any show-stoppers (there will most certainly be legal challenges), it is prudent to let them play out before we invest in tens of thousands of expensive machines. There needs to be vigorous public discussion and citizen education programmes. We need to see how voters react to these machines. Trials should be transparent to observers and media personnel. If we’re truly serious, we should even commission studies by election security professionals and legal experts.
The case of Ireland presents another cautionary tale against rushing into electronic voting. Head of government, Bertie Ahern, impressed with EVMs used in India and France, famously declared that Ireland would become “the laughing stock of Europe because voters still used pencils in elections”. In 2002, Ireland spent 54 million euros in procuring and storing 7,500 EVMs to update their electoral system. An independent commission then, quite predictably, discovered that these machines had dismal security. The machines were never used and they languished in storage, costing Irish taxpayers 140,000 euros a year. In 2009, it was decided that it was too expensive to upgrade them. In 2012, the government finally sold the machines off for scrap for a paltry 70,000 euros — a rate of nine euros per machine — a fitting conclusion to an epic scandal.
Published in The Express Tribune, May 21st, 2015.