The West may be reeling from such revelations, but the vulnerability of election technology is well known in other parts of the world. In 2016, Andrés Sepúlveda revealed how, beginning in 2005, he manipulated elections for a variety of clients across the South American continent, ultimately having a hand in nine different countries’ polls. Just before the 2016 election in the Philippines, hackers attacked the country’s election commission website and compromised personal data of approximately 70 million people. In India, the election commission has challenged hackers to infiltrate Indian election technology, after calls to replace electronic voting machines with paper ballots.
Is technology losing its luster? Clearly, technology offers no fool-proof solution to the problems of electoral malpractice and fraud. It can, however, ease the administrative burden of conducting elections, making activities like vote collation, counting and tallying less onerous. With the correct safeguards in place, technology can also promote public trust in the electoral process.
Kenyan experience with election technology shows both sides of the story. In 2013, many expected the election, which was governed by a set of new electoral laws and a new, internationally-renowned constitution, to mark the beginning of a new political era. Technology was at the heart of this new electoral environment, and it was Kenya’s harbinger of hope. Biometric voter identification, electronic voter identification and electronic results transmission were meant to address many of the country’s past problems; the new tools were designed to ensure a credible and verifiable voters’ register, prevent problems like voter impersonation and multiple voting, and address the risk of results tampering.
Unfortunately, however, technology did not live up to its promises. Despite early indications that all was not well, including severely delayed procurement, inadequate time to train clerks in the use of the machines and problematic test results, the IEBC moved forward. By the time voting was complete, voters around the country had been faced with dysfunctional and/or nonfunctional voter identification machines, a frozen ‘live stream’ of results and the eventual collapse of the entire electronic results transmission system. The IEBC announced that polling station tally forms would have to be manually transported to Nairobi before results could be determined. Instantaneously, the past was the present; the election would be determined by a manual count of thousands of tally forms, the validity of which was in the hands of those officers charged with their transport. When the manual forms were examined, observers and analysts found that they were riddled with myriad problems and errors. In the aftermath of the election, other problems came to light, the most notable of which was the alleged ability of certain actors to access the IEBC’s electronic results system. When it became clear that the same server was being used to host TNA and IEBC data, public confidence in the legitimacy of the results plummeted.
This electoral cycle is not marked by the same hopeful sentiment of 2013. Instead, domestic and international watchers have voiced significant fear of violence in 2017, especially in relation to gubernatorial and other local races. The fear is palpable, which makes credible, verifiable and transparent systems even more critical.
Unfortunately, it seems that few lessons have been learned. In this election cycle, procurement is again an issue. After canceling the tender for its integrated elections management system (IEMS), the IEBC has awarded a direct contract to the same company who supplied the contentious BVR kits in the last election. This decision alone threatens public confidence in the integrity of the process. Moreover, the process has been so delayed that the IEBC has requested that it be permitted two extra months to finalize the deal. Perhaps most telling are the findings from domestic observer groups who monitored the voter registration process. InformAction, for instance, found that voter registration was marked by widespread problems with the biometric voter registration kits. Some kits suffered temporary and fixable problems, but others failed to work at all. Voters who were previously registered suddenly found that they were absent from the register, and voters who were registering for the first time found that they were already on the list.
There were also highly disturbing indications that technology has failed to address the problem of voter impersonation. Some individuals who went to register found that they were already registered but in areas that were not at all related to their places of residence or employment. Such cases indicate that people’s ID numbers are being acquired and used to strategically populate the register. Clearly, technology has limited utility in preventing fraud if the rest of the electoral system is problematic.
The problems have sparked contentious debate and deliberation in Kenya. One response has been to amend the electoral law and allow for the use of ‘complementary mechanisms’ for voter registration, voter identification and results transmission, in case technology fails. Since these mechanisms have not been clearly defined in the law, however, it remains to be seen what specific mechanisms will be used as back-ups, and whether they open up the process to further manipulation. On the other hand, the last phase of voter registration included the re-emergence of the notorious and much-criticized ‘green books,’ a collection of notebooks in which registrants’ information is manually recorded. It seems, then, that the IEBC plans to use such manual tools.
While it is of course necessary to have back-up mechanisms in place, given Kenya’s particular history and context, it might be a wiser choice to have non-manual forms of back-up processes. Given the limitations of the current context, however, parties, observers and other stakeholders should consider training their agents and observers in the workings and use of technological tools. These agents should be on the ground to monitor the way in which technology works (or doesn’t) on Election Day and through the counting and announcement of results. Future revisions of the electoral law should include extensive stakeholder consultation, especially with observers who have witnessed the problems, and should include a clear definition of what technological and manual mechanisms will be used.
Technology may have the potential to transform elections, but the fulfillment of that potential depends largely on the surrounding electoral context. Regardless of the sophistication of the tools in use, political will and broad stakeholder commitment are the keys to credible elections.