Technology has failed Kenyans in the 2013 general election. Over the past few months, election officials and their friends in the media have raised public hopes for a fair election by hyping measures to modernize the voting system. But it's possible that these new reforms could instead become the cause of increased tensions.
In some areas, the much-anticipated biometric voter identification kits only started arriving in the early morning hours on election day. Given that polling was slated to start at 6:00 A.M., this left polling agents with hardly enough time to test the equipment, much less to charge its batteries. As a result, it came as little surprise when the kits began to fail just a few hours into voting. There were a number of reasons cited, including low battery life, the lack of electricity in certain areas, and polling officers' difficulties in accessing the central system.
As a result, many polling centers around the country abandoned the biometric kits in favor of the manual register. As a back-up, the manual register, which is supposed to exist in each polling place, is a useful instrument, providing voters' identification numbers and photographs. Complete reliance on it, however, especially in this election, is questionable. The biometric kits were brought in early on in the process and first used to register voters. The machines record voters' biographic information, photographs and -- most critically -- their fingerprints. On election day, the machines were meant to use this biometric data to positively identify registered voters and reject those who had not registered, ultimately providing a check on ballot-box stuffing.
Of course, the manual register should provide a similar check, for it includes everything on the biometric kit except the fingerprints. The problem, however, is that the manual register is vulnerable to unscrupulous individuals, who could much more easily manipulate the paper list of eligible voters. This is especially problematic in party stronghold areas, where public pressure to produce a victory for the candidate who comes from that area can bear down on polling officers.
These issues highlight a bigger problem as well: the lack of a final, published register of voters. As of election day, the final register had not been released, despite promises to the contrary made by the national election commission, the IEBC (Independent Electoral and Boundaries Commission). The lack of the final register begs some important questions. Were there problems with the provisional register? If so, what were they and why were they unable to be resolved in time for elections? And most importantly, was it the provisional register or the final register that was used during elections? If it was the former, is it possible that ineligible people were able to cast ballots?
Early on Tuesday morning, the IEBC also announced that its electronic results transmission system (ERT), which was supposed to have allowed polling officers to use mobile phones to immediately transmit results over a secure network, had experienced a "glitch." In a public press briefing, the IEBC said that the problem, which was related to insufficient server space to accommodate all the incoming results, had been resolved.
The ERT system had been in the hot seat in the days leading up to the election. Tests of the system had revealed deficiencies, and Kenyans were wondering what provisions had been made in case of the system being hacked or in the case of system failure. The "glitch" was thus unsurprising to many. This was compounded, however, by the incredibly slow pace of incoming results. If the error had been resolved, why were only 40 percent of polling centers reporting results more than 24 hours after the official close of voting? And even more inexplicable was the fact that results of many other, lower-level races were still streaming in from around the country. If presidential tallies were supposed to completed and reported first, why were there so many non-presidential results coming in while the presidential results were still at just a trickle?
The slow pace of incoming presidential results and questionability of the ERT system on Tuesday seem suspiciously related to the virtually unchanging difference in the margin of support for the top two contenders, Uhuru Kenyatta and Raila Odinga. For most of Tuesday, the IEBC's live stream showed Kenyatta having garnered around 54 percent of votes and Odinga having won about 40 percent of all votes. This came into sharp focus when it became clear that the IEBC was calculating percentages based on the total number of "valid" votes, rather than the total number of votes cast. There has been concern that many votes were non-justifiably rejected, confusion that voting reforms were hoping to avoid.
In order to win outright in the first round, a successful presidential candidate needs to win an overall majority of the vote and at least 25 percent of the vote in each of more than half the country's 47 counties. Calculating the top two contenders' support based on the total votes cast, as required by the constitution, showed Kenyatta with about 50.1 percent of the vote and Odinga with 39.5 percent of the vote.
Tuesday night, in an attempt to offer a public explanation, IEBC Chairman Ahmed Isaak Hassan announced that the traditionally used "returning officers" from around the country would be in Nairobi on Wednesday to personally file results. He added that the percentages of votes for the presidential candidates would be calculated based on the total number of votes cast.
The announcement was controversial, as it reduced Kenyatta's margin (as of late yesterday evening) to a very bare majority, which could easily change as more results come in. It also raises more questions about the ERT system. Now that it has been abandoned, it is important to know what happened. What were the problems with the system and why could they not be resolved? What checks are there on the manual forms, now that the check of the electronically-transmitted results is unavailable? Given the IEBC's multiple assurances that the ERT system was fool-proof, Kenyans deserve a comprehensive explanation as to what happened.
Technological failures do not necessarily mean a less credible election. Much will be revealed in the coming days, especially with regard to how stringently (or not) the law was followed around the country. Let's hope that even without the much-touted technology, Kenyans will be able to rest assured that the election results reflect the public will.
Seema Shah is a political analyst based in Nairobi
Photo by CARL DE SOUZA/AFP/Getty Images
Transitions is the group blog of the Democracy Lab channel, a collaboration between Foreign Policy and the Legatum Institute.