Election Technology,Disputes,and Political Violence in Kenya

Abstract

In Africa, many countries have adopted technology in the conduct of elections so as to improve efficiency and deter malpractice. However, electoral disputes and violence continue to recur even where elections involve use of technology. This article builds on a case study of Kenya to show the limitations of technology use in elections. We argue that use of election technology does not guarantee the credibility of elections; technology has become a ‘black box’, which competing parties exploit to either play victim or declare themselves winners. The paper concludes that election technology is the new frontier for fraud. The manner in which election technology is procured, deployed and utilized is not only technical but also political. Addressing this political problem is imperative to avoid political violence around elections.

Keywords

Election technology political violence disputes electoral integrity Kenya opposition

Introduction

A few weeks before Kenya’s August 2017 elections, one of the country’s most prominent politicians and leader of the opposition, Raila Odinga, was asked by the media whether his party, the National Super Alliance (NASA), had established a parallel vote-tallying centre outside Kenya. Odinga replied that the party had a parallel tallying centre in Kenya and in the clouds (Muraya, 2017). The question of whether NASA had a ‘high tech’ parallel vote-tallying centre followed highly publicized media reports that the party had established such a centre because of fears of the government tampering with the results. Earlier, the police had destroyed a ‘high-tech tallying centre’ set up by NASA (Ombati, 2017). After the destruction of this centre, the opposition continued to build its ‘digital technological defence’ against rigging.

The great stock placed on technology became more evident later in court during the hearing of the case filed by the opposition after losing to the incumbent President Uhuru Kenyatta. The opposition declined to accept the results and challenged the election management body, the Independent Electoral and Boundaries Commission (IEBC), to ‘open’ the computer servers, which would reveal that their candidate, Raila Odinga, had won the election and should be declared president. But the IEBC was reluctant to grant access to its servers, citing difficulties relating to time and the security of their computer servers. In the end, the Supreme Court annulled the presidential election results and called for a fresh election. Raila Odinga and his party declined to take part in the fresh election, demanding a shake-up of the IEBC’s information and communications technology (ICT) infrastructure and computers used in results transmission.

Placing great faith in technology is not new in Kenya or Africa in general. There has been remarkable deployment of technology in elections, specifically use of inventions including electronic equipment, software programmes and new materials. Use of new technologies such as computers, optical scanners, the internet and digital equipment such as biometric devices has been on the increase. By 2018 about half the national-level elections in Africa involved use of digital technology, of which the most widely used was biometric voter registration alongside the growing adoption of biometric voter identification, and electronic results transmission (Cheeseman et al., 2018; Nwangu et al., 2018: 4). Many countries are continuing to adopt technology to improve efficiency and transparency at various stages of the electoral process (Gleb and Clark, 2013; Yard, 2010).

However, studies have cautioned that some of the electoral problems in Africa require political solutions rather than technological interventions. A recent study warns against the fetishization of digital technology (Cheeseman et al., 2018). Other studies warn that election technology should be deployed with caution because it is not necessarily a solution to the political complexities surrounding elections (Barkan, 2013; Cheeseman et al., 2018; Evrensel, 2010; Gelb and Clark, 2013). Still others such as Yard (2010: 10) point out that adoption of technology may lead to passing control of the process from the majority of voters to a select few who are able to understand the workings of the technology. These studies have not highlighted the particular aspects of digital technology that contribute to disputes and electoral violence, or shown how technology has failed to promote integrity in elections. They have also failed to demonstrate the interplay between technology and politics in undermining electoral credibility.

This interplay between technology and politics and how this interplay undermines electoral credibility is the contribution that this paper makes based on the Kenyan experience. Its central argument is that the relationship between technology and politics is more nuanced. Improvements in electoral credibility will benefit from the adoption of technology only if the technology is deployed in a reformed political environment. The paper demonstrates that technology cannot solve problems of electoral integrity if the challenges associated with human intervention remain in place. Electoral technology can reinforce or undermine democratic processes, both of which are dependent on the interests of the incumbents or other powerful groups. The paper is based on research data collected by way of interviews with different informants, and on grey and other literature including laws and court judgements. It focuses on the use of digital technology in Kenya’s elections and brings out the particular aspects of digital technology that exacerbate rather than reduce electoral violence.

The paper is organized as follows: section two focuses on the role of technology in the administration of elections. It provides the theoretical framework for assessing the role that technology can play in influencing the quality of elections. Section three provides an overview of Kenya’s past elections and recommendations for the adoption of comprehensive technology. Section four discusses the challenges in the use of technology in Kenya’s elections, while section five discusses how use of election technology in the 2017 elections caused rather than prevented disputes and violence. Section six identifies new challenges resulting from use of technology in the 2017 elections; this section also discusses the specific aspects of the election technology that occasioned disputes and violence during the 2017 elections. The last section concludes the discussion by emphasising the importance of focussing on the practice of democratic principles to prevent electoral violence.

Technology and difficulties in enhancing credibility of elections

In Africa, there is continued use of digital technology in the electoral process for mainly similar arguments: to reduce the scope for manipulation of elections, improve efficiency, embed transparency in the electoral process and, above all, boost legitimacy of elected governments (Cheeseman et al., 2018; Evrensel, 2010; Gelb and Clark, 2013; Jacobsen, 2020). Use of new technologies, however, has not resulted in greater effectiveness in administration of elections or contributed to greater credibility in the results of elections. Where technology is not well-managed, it has led to electoral disputes and violence (Amoah, 2019; Cheeseman and Klass, 2018). There have been failures of technology due to poor implementation and logistics but also due to deliberate manipulation to favour certain outcomes. This suggests that technology may promote efficiency in the electoral process but it does not necessarily address the political challenges in any given election.

In discussing the role of technology in elections, some studies have been overly enthusiastic about the important role technology plays in improving the electoral process. The studies, especially those using a ‘technological determinism’ perspective, emphasise that technology is ‘an unstoppable force in its own right’ (Garson, 2006: 5) and that new technologies transform societies in many ways (Dafoe, 2015: 1048; Wyatt, 2008: 173). Technological determinism assumes that technologies are developed independent of social and political interests and that they only serve the interests for which they are developed. Election studies from this perspective are emphatic that technology has led to improvements in the electoral process and achievement of better outcomes (Gelb and Clark, 2013; Gelb and Decker, 2012; Harvey, 2003; Golden et al., 2014; Mugica, 2015). There is thus a general fetishization of electoral technology (Cheeseman et al., 2018). Others note that use of technology in elections improves the integrity of the process and could potentially contribute to more stable electoral systems (Hobbis and Hobbis, 2017: 112; Nwangwu et al., 2018: 5).

Studies on how technology is deployed in elections have revealed certain challenges. In their critique of new technologies and elections Cheeseman et al. (2018) and Barkan (2013) point out that election technology arouses uncertainties in electoral politics but fails to resolve these uncertainties. Furthermore, these technologies do not operate independent of socioeconomic and political forces. They are subject to control by the dominant political forces. This limits the ability of technology to solve political challenges that arise in the electoral process. Technological determinism theory, therefore, does not explain how disputes in the use of technology arise; nor does it facilitate predicting how such disputes might arise and the consequence for the electoral process and outcome.

We argue that technology can be used to reinforce the interests of dominant political forces, to promote democratic values, or subvert them altogether. This view draws from reinforcement theory, which emphasises that technology is ‘a tool like any other’ and that ‘like other tools, it tends to reinforce the structural position of the powers that be’ (Garson, 2006: 6). This view stresses that ‘computer technocrats function as the hired guns of the electronic frontier, answerable to the masters who pay them’ (p. 6). This is a pointer that any form of technology is susceptible to manipulation that can result in a wide range of social outcomes (Castells, 2004; Dahl, 1989: 339). It is also an indication that human factors rather than the technology itself are central to how technology achieves the ends (Ezebuenyi, 2014: 52). For this reason technology can constrain or promote change. What it achieves is subject to the interests of the powerful political forces prevailing at a particular time.

Reinforcement theory sees technology as a tool primarily serving the interests of those with the power and resources to employ it for their benefit. Use of technologies can reinforce existing power structures; they can be manipulated to either reinforce democratic tendencies or promote non-democratic ends (D’Agostino et al., 2011: 6). With regard to elections, use of technologies can promote public confidence or damage the credibility of elections altogether. Indeed, Cheeseman et al. (2018) have pointed this out, describing numerous examples of where new technologies have been deployed to maintain the dominance of incumbents but ultimately damage the credibility of the electoral process and the legitimacy of the results. Other recent studies on elections in Africa also show that use of technologies in elections is the new battleground (Amoah, 2019; Debrah et al., 2019). These studies show that manipulation of the electronic computation and transmission of results has been critical in deciding presidential elections. This has meant increased contestations over the procurement of technology, and how it is securely utilized by the election management body in the transmission of results. These contestations result from the recognition that whoever controls the deployment of election technology, and especially the electronic computation of results, may take advantage and turn the technology to their own ends and win. On the other hand, whoever loses control of the electronic computation of results, loses the election (Amoah, 2019: 69). This control oftentimes begins by gaining influence over the electoral management bodies so as to direct decision-making about the technology, its procurement, deployment and utilisation.

We argue here that technology as a tool for use in the electoral process does not have its own end per se. The end it delivers depends on the rationale for its introduction, the nature of the political context in which it is used, and the political forces that gain control over its deployment. It can either be used to advance social change in society or to maintain the ‘status quo and even thwarting democracy’s reach’ (Calista and Melitksi, 2007: 90).

When election technology is well-managed and -implemented, it can reinforce trust in elections and boost voters’ confidence in the electoral process. This in turn improves democratic participation. However, technology is subject to human behaviour and may be manipulated by politicians and election officials to promote the self-interest of powerful individuals and their political parties. In this latter instance it reinforces the divisions in society and provokes rather than solves disputes and attendant violence. This implies that the political context on which technology is deployed matters. Where the framework for political competition is based on democratic values and principles, technology will promote a transparent electoral process. Where the context is one characterised by competition over control of political power to advance ethnic and other sectarian interests, use of technologies may be manipulated by powerful political elites to consolidate their rule by validating elections that are generally conducted on a weak infrastructure. The fact that technology is a tool that can be used to either build trust in the electoral process or reinforce certain outcomes suggests that technology does not in itself contribute to credible elections.

We argue that adoption of technology in Kenya’s elections and Africa in general glosses over the key electoral reform questions and aspects of political competition that are necessary to guarantee credible elections. We also contend that electoral disputes are largely political, and technologies cannot subsume the role of political solutions. New technologies may enhance efficiency in the electoral process but will not address political challenges such as ethnic and other divisions that undermine confidence in the election management bodies. Technology should not be front-loaded in discussions on electoral reforms, rushed, or taken as the panacea for electoral fraud. This is particularly so because adoption of election technology has been characterized by periodic breakdowns, unreliability, rent-seeking and vendor wars, all of which combine to undermine the potential of technology to make elections credible. Its adoption without addressing the fundamental, underlying human-related challenges inherent in elections, as the Kenya experience demonstrates, results in more complaints as opposed to improvements in the conduct of elections.

Kenya’s elections, violence and technology

In Kenya, elections conducted after the return of multiparty democracy in the early 1990s have been characterised by disputes that have led to violence. The post-2007 election violence in particular is a subject of several studies (Cheeseman, 2008; Githongo, 2008; Kanyinga, 2011; Kanyinga and Okello, 2010; Kameri-Mbote and Kindiki, 2008; Mueller, 2008; Wolf, 2009). The immediate trigger for the violence was a dispute over the vote count and tallying of the presidential election results. The opposition, the Orange Democratic Movement, protested at the way the Electoral Commission of Kenya, tallied the results and announced the incumbent President Mwai Kibaki of the Party of National Unity (PNU) as the winner. The opposition leaders insisted that the vote was rigged in favour of the incumbent. The government/PNU insisted that the election was free and fair, and hurriedly swore in Kibaki for a second term as president. The Independent Review Commission for the 2007 elections (the Kriegler Commission) appointed by African Union’s Panel of Eminent African Personalities to investigate the electoral process and outcome concluded that those elections were so tainted at several stages that it was difficult to accurately determine who really won (Republic of Kenya, 2008). It then argued that the integrity of the vote count and tally could be enhanced by use of technology and recommended that such technology be introduced in future elections, especially in the tallying and transmission of results.

The Kriegler Commission expressed the view that technology on its own would address the challenge of fraud at vote counting and reduce the politicisation of vote counting and tallying. This formed the basis for the introduction of technology in subsequent by-elections. In 2010, the country adopted a new constitution with provisions for electoral administration methods that were accurate and verifiable. Based on these, what had emerged as general infatuation with electoral technology now had solid constitutional backing.

Although not expressly required by the constitution, technology was central in the demands for conducting credible elections in Kenya. Consequently, as part of delivering on its constitutional imperatives, the IEBC introduced technology in certain aspects of the 2013 elections. The elections pitted two strong political alliances against each other, mobilized along ethno-regional blocs of almost equal numerical strength. There was, on the one hand, the Jubilee Alliance of Uhuru Kenyatta and his running mate, William Ruto. On the other hand, was Raila Odinga, then Prime Minister, and his running mate Kalonzo Musyoka, then Vice President. The relative equality of these alliances turned the election into a closely contended one. Use of technology was therefore thought to be valuable in determining the accuracy of results.

In introducing technology, the IEBC sought to address two of the greatest shortcomings of the 2007 elections: lack of an accurate voter register, and the lack of accuracy and integrity in the results transmission and collation processes from the polling station all the way to the national tallying centre (Barkan, 2013: 157). For voter registration, a biometric voter registration (BVR) system was introduced. This was to capture ‘biometric data in the form of digitally scanned fingerprints and a digital photo of each person registered’ (Barkan, 2013: 158). By capturing biometric data, it was hoped that fraud, especially in cases where dead voters ‘voted’, would be eliminated. BVR would enable all voters who reported to the polling station to be identified by having their fingerprints scanned and the image compared with that already stored on the computerized register (Barkan, 2013: 158). The identification process was to be undertaken through a biometric electronic voter identification device (EVID or poll book; Bowman and Githaiga, 2015: 280).

The IEBC also introduced technology to address problems around the results transmission system, to ensure fast and accurate transmission of results from each of the polling stations using a dedicated and programme mobile phone (Barkan, 2013: 158). These changes were expected to help improve the credibility of the elections. Technology would ensure an election that was ‘free and fair’ and thus legitimate in the eyes of the people (Barkan, 2013: 158).

Technology challenges and credibility of the 2013 elections

With these interventions, the IEBC proceeded to conduct the 2013 election. The election was generally peaceful. Even though there were instances of disruptive violence in some parts of the country, the peace narrative that underpinned the election muted these concerns (Karen et al., 2014). Use of technology in the 2013 elections, however, did not silence disputes, first over the widespread failure of the BVR kit in many polling stations and subsequently in relation to vote counting and tallying. Despite use of technology, the election results were still contested at presidential and lower levels.

In addition to the dispute over tallying of the presidential election results, there were other issues that bedevilled the application of technology in the 2013 elections. One procurement incident regarding the technology became the subject of political disputes and subsequent litigation. Firstly, at a cost of USD 95 million, procurement was found to be exorbitant (Bowman and Githaiga, 2015: 279) because it provided ‘rents’ to election officials. Secondly, the technology that was procured was complicated and difficult to implement (Barkan, 2013: 160). The requirement for acquiring EVIDs in addition to laptops, which were to be used as BVR machines, presented a logistical nightmare for officers; it also increased the financial outlay. Thirdly, procurement disputes led to the original bids being disqualified for technical deficiency and cost implications. This caused a delay in the voter registration process (Barkan, 2013: 161). With these challenges, technical advisors hired by donors counselled the IEBC to abandon the BVR process and resort to manual registration (Barkan, 2013: 161). However, due to the belief by politicians that technology was the ‘magic wand’ addressing fraud in elections, the IEBC was compelled to acquire BVR equipment from a French company, under an arrangement that involved the Canadian Government (Barkan, 2013: 161).

The technology worked well during the registration of voters. Officers registering voters reported minor challenges that were not widespread. On voting day, however, the domestic observation team, the Elections Observer Group (ELOG), reported widespread failure of the EVIDs. Even though they had been designed to identify voters biometrically on voting day, they largely failed to work with ELOG, putting the failure rate at about half. Polling officials resorted to the use of manual printouts to identify voters. In addition, many of the cell phones meant to transmit provisional results from the polling stations to the tallying centre did not work either due to forgotten passwords, low batteries or data connectivity problems (Barkan, 2013: 62; Bowman and Githaiga, 2015: 280). Technology was meant to improve confidence and trust in election results and in doing so, erase the memories of the 2007 disputes and allegations of fraud. However, failure to provide proof of verifiable results from the computer servers at the national tallying centre undermined that confidence. This failure to verify the results from the server then undermined confidence in vote tallying and in the final result; yet this was what everyone expected to prevent perceptions of fraud. In the end, the IEBC suspended electronic transmission of results and tallying. The commission then asked election officers to deliver the physical election results forms to the national tallying centre (Bowman and Githaiga, 2015: 280). Rather than reduce disputes, technology had now become a new source of contestation at the national tallying centre, with the opposition party accusing the IEBC of rigging the election in favour of the Jubilee Alliance.

While the IEBC declared Jubilee candidate Uhuru Kenyatta as the winner of that presidential election, the leading opposition coalition led by Raila Odinga and a coalition of civil society organisations challenged the electoral process and the results in court. The Supreme Court decided in favour of Kenyatta, thus validating the IEBC’s results. This decision has since been criticised on many grounds (Gondi and Basant, 2015; Harrington and Manji, 2015; Heidi and Nyoike, 2016; Maina, 2013). Some argue that the process was so fundamentally flawed that it is difficult to discern whether the declared presidential results were accurate and lawful. Noteworthy, however, is that the failure of technology during the elections and whether the technology used complied with the constitution was at the centre of the dispute that ended up in the Supreme Court.

Technology failure stimulated more intense disputes at the Supreme Court, particularly because some argued that electoral technology, like other technologies, is rarely perfect. The court in particular disagreed that the failure of technology had affected the credibility of the election. In the view of the court, the technology had not been intended to replace the traditional manual process (Aywa, 2016: 55), since technology can and invariably does fail. To the judges, competing interests involving impropriety and criminality marked the procurement process. They decided that the integrity of elections could not be fully placed on the adoption and successful deployment of technology. In the judges’ opinion, technology could quickly become obsolete and could also fail. More robust measures were required, so that even if manual processes had to be resorted to as a backup (when technology failed, as happened) the credibility of the emergent results would be in no doubt. Evidently, the adoption of technology seemed to have spurred disputes rather than improve confidence and trust in election results.

Technological efficiency and disputes in the 2017 elections

The opposition accepted the Supreme Court decision on the 2013 presidential election but continued to raise concerns about the conduct of those elections and how technology was used. First, opposition leaders raised concerns over the impartiality of the IEBC. They pressed for the disbandment and reconstitution of the commission. Second, they demanded additional electoral reforms, including the use of polling station results as final results. These demands, triggered by technological failure and a lack of confidence in the transmission of results in the 2013 elections, informed further political disputes. The first phase of these demands took a political turn when the opposition called for a referendum to change the constitution, through what they called the Okoa Initiative (Opalo, 2016):, this included, among other things, a call to restructure the IEBC and the election management process (Menya, 2015). This initiative failed when the IEBC argued that the signatures collected to initiate the drive for a referendum did not meet the constitutional threshold of 1 million.

Following the collapse of the referendum push, the opposition resorted to street protests to pressure members of the IEBC to resign. The government responded to the growing protests by establishing a Joint Parliamentary Select Committee (JPSC) composed of members of both the National Assembly and the Senate, drawn from both the opposition and ruling party, to resolve the crisis around IEBC commissioners.

As with previous initiatives, this parliamentary committee had immediate faith in the election technology. The committee recognized that countries around the world had adopted elections technology to various extents (Republic of Kenya, 2016: 134). A survey by the International Institute for Democracy and Electoral Assistance (IIDEA) recorded that approximately 24% of countries were utilizing biometric voter identification at polling stations. The JPSC report, citing the IIDEA study, indicated that while there was adoption of electronic voting technology, only 14% of the countries had progressed as far as having e-voting (Republic of Kenya, 2016: 134). The committee also evaluated the challenges faced in the utilization of technology in 2013 and found that the failure of EVID was as a result of poor staff training. The committee also found that batteries ran out of charge and that only 17,000 of the 33,000 polling stations had electronically transmitted the provisional results before the system failed. After system failure, the IEBC abandoned the electronic transmission of results. Finally, there were inconsistencies in data submitted by political parties – using technology – employing the party nomination software (Republic of Kenya, 2016: 131–2).

Based on the submissions by individuals and groups, the JPSC recommended that technology be used in all aspects of the 2017 elections. It provided broad guiding principles, including that technology be impartial, efficient, simple, accurate, verifiable, secure, accountable and transparent, and that its procurement be undertaken in a transparent manner. The committee also recommended that any new technology be introduced at least a year before elections (Republic of Kenya, 2016: 142). The committee also made suggestions on amendments to the legal framework, to obligate the use of technology in the election process (Republic of Kenya, 2016: 143). In addition, the committee underlined that for the 2017 elections, technology would be used for voter registration, identification and results transmission (Republic of Kenya, 2016: 143) and that the equipment be tested at least 60 days before the general election (Republic of Kenya, 2016: 144).

These recommendations were adopted by Parliament in October 2016, at which time the legislature also adopted amendments to the Elections Act, among other election laws. The amendments underscored the place of technology in elections in Kenya. Politicians in Parliament had the faith that technology would deliver the country from the challenges of past elections. This was akin to what Cheeseman et al. (2018) categorize as fetishization of technology in elections.

The changes were also supported by faith in technology among other players in the electoral process, as the date for the 2017 general election approached. A national election conference convened by the IEBC, the Ethics and Anti-Corruption Commission, and the Kenya National Commission on Human Rights, for instance, underlined the need for the IEBC to allow for independent testing of technology supporting the 2017 elections to ensure it met the constitutional threshold. The meeting urged the IEBC to avoid past problems associated with technology by ensuring early testing of any technology to be deployed in the elections. The meeting stressed the need for rigorous training of IEBC staff on election technology, timely and transparent procurement, public awareness on the technology being deployed, and integration of the technology being deployed – electronic voter registratoin, electronic voter identification and electronic transmission of results – into one. It also requested electrification of polling stations, use of ICT to enable Kenyans in diaspora to vote, and monitoring and evaluation of technology used in the 2017 elections.

Flaws in the 2017 elections technology

Contestations over the legal framework

The process to deliver technology for the 2017 elections was cleared by the JPSC. In line with the committee’s recommendations, Parliament made changes to the Elections Act capturing the importance of technology in Kenya’s elections. The amendment provided for the introduction of an integrated electronic system for BVR, electronic voter identification, and electronic transmission of the results. The law required that the technology be integrated to link within the various components of voter registration, electronic voter identification and electronic transmission of results.

After legislation was enacted to entrench use of integrated technology, the IEBC complained that the law ‘tied their hands’ without any wriggle room should the technology fail. This led to efforts to amend the law to allow for a complementary mechanism. The process to pass the law was, however, rife with contestations between the ruling Jubilee coalition and the opposition Coalition for Reforms and Democracy (CORD) and Democracy. The opposition argued that the changes were intended to compromise the elections. Despite protests, the Jubilee Alliance mobilized its parliamentary numerical strength to amend the Elections Act introducing a new section to allow the commission to put in place a complementary mechanism for the identification of voters and transmission of results. The term ‘complementary’ aroused new controversy because it was seen as undermining the ‘authority’ of technology as the panacea for the country’s electoral challenges. The opposition sought the intervention of the courts (Langat, 2016), arguing that the complementary mechanism had to be one of technology. This demonstrated blind faith in technology as a cure to fraud in elections.

The new law required that a policy on the progressive use of technology be developed. Similarly, the IEBC prepared the Election Technology Regulations, 2017, and submitted them to Parliament for adoption. The regulations covered a wide spectrum of technology application procedural issues, including procurement, deployment, maintenance, testing, audit of technology, information security, data archiving and data recovery. Finally, the Election Offences Act also included offences relating to technology, ranging from stealing electronic equipment, destruction of electronic equipment and information, and manipulation of computer data. Kenya’s legal framework governing the 2017 polls had robust provisions on the use of technology for elections, demonstrating the premium the country placed on its importance in ensuring that those elections were credible and acceptable to the people of Kenya.

Control of procurement

Following the changes to the law, the IEBC started rolling out the election technology. The process, however, provoked controversies that evolved into bitter political conflicts between the two main parties – the ruling Jubilee Alliance and the opposition, NASA. The first dispute arose from the bidding to identify a vendor. The IEBC identified Safran, the company that had supplied the equipment that failed in the 2013 elections. There were multiple objections immediately this was done, including from the opposition NASA, who argued that the company could not be trusted because of its association with past failures. The IEBC argued the commission was justified in picking Safran because the company had the capacity to comply with the tight election timelines and that the company had earlier supplied the BVR kits (Africog, 2017: 15). In addition, the cost of the technology became a point of contestation. Jubilee supported the IEBC argument, which led to the opposition party accusing the government, the IEBC and Jubilee of planning to rig the election using Safran. The IEBC had to cancel the initial contract for the integrated elections management system after objections were raised about the manner in which the process was being undertaken (Africog, 2017: 15).

No sooner had this controversy been resolved than another arose. The commission’s ICT manager, Chris Msando, was found murdered a few days before the elections. Msando had played a key role in developing and rolling out the Kenya Integrated Elections Management System (KIEMS) kits and therefore was central to the employment and use of technology at the IEBC. By this time, everyone in Jubilee and the opposition had come to believe that technology was central to determining the final result and therefore securing it was important. The competing parties believed, and convinced the public, that technology was the game changer in the upcoming elections, with its application determining whether the elections would be free and fair or fraudulent. Msando, the ICT manager, was in charge of ICT security and of the broad election technology. His murder was interpreted by the opposition as being aimed at removing the security of the system to allow election rigging. A local civil society organization’s coalition, Kura Yangu Sauti Yangu (Swahili for My Vote My Voice), lamented the messages that the murder sent, observing that his torture and murder undermined Kenya’s election management body, the IEBC. His murder was never solved. Msando’s death added to the controversy around the application of technology in the elections and ultimately contributed to denting the credibility of the 2017 elections.

Veracity of digitally transmitted results

The election was held in August 2017 and the IEBC declared the incumbent President Uhuru Kenyatta the winner. Immediately after, Raila Odinga, leading the opposition, protested that Kenyatta and others who won the election were ‘computer generated’ leaders and that they were not the people’s choice. Together with others in the opposition, they challenged the presidential election results in the Supreme Court. In their petition, the opposition again focused on the abuse of election technology and complained that the process of relaying and transmitting the results from the polling stations all the way to the national tallying centre was flawed. They claimed that the data were neither accurately transferred from the statutory election forms to the KIEMS kits nor transmitted with an accompanying electronic image of the statutory forms. Consequently, the opposition raised several complaints relating to technology. One of these concerned the challenges around the use of the KIEMS kits and the results transmission processes, which they claimed offended the constitutional process, robbing the elections of credibility. On the other hand, the IEBC and Jubilee argued that the flaws in the electronic transmission of results, could not be the basis for voiding the presidential election because the results had a large margin of difference between the two leading contestants. The court disagreed and underlined the fact that elections were not about numbers alone. The failures of the process, including technology, had the effect of vitiating the election.

The second complaint concerned the failure to electronically transmit results from 11,000 polling stations that were outside the 3G and 4G network range for transmission. These results were manually transmitted. The court did not accept the explanation for this issue or why the IEBC resorted to manual transmission. The Supreme Court continued to emphasise the importance of technology by observing that the failure to access the 3G and/or 4G network was not a technological failure. The court observed that the ‘IEBC’s ICT officials must have known that there were some areas where the networks were weak or totally lacking and should have made provision for alternative transmission’ (Republic of Kenya, 2017).

A third complaint arose during the hearing of the petition at the Supreme Court. The opposition requested access to IEBC computer servers to review the data used to transmit the results. Even though the court gave an order for access to these servers and appointed ICT experts to oversee the process, the IEBC stonewalled, earning the wrath of the court and contributing to the final decision to nullify the presidential election result. The IEBC’s reluctance also triggered protests by supporters of the opposition. In the end, the clarion call to ‘open the servers’ came to define the contestation around technology, with the petitioners holding the belief that those servers held the ’true results’, which would show that the opposition candidate, Raila Odinga, had won the presidential election. The Supreme Court in the end found that the IEBC had violated several technology-related directives. These included the failure to provide an internal firewall configuration; failure to provide certified copies of penetration tests; and failure to supply the specific locations of each KIEMS kit using the General Packet Radio Service, an electronic system that allows for sending and receiving information using cell phones. The court also found that there was no record of login trails of users and equipment into the IEBC servers. The login trails of users and equipment into the KIEMS database management systems, and the administrative access log into the IEBC public portal between 5 August 2017 to 28 August 2017 were also lacking. This raised doubts about the accuracy and verifiability of the transmission system: important aspects of a credible election.

The court reached the decision that non-compliance with the technology-related requirements for the transmission of results had affected the integrity of the results. It was especially critical of the IEBC for its failure to allow petitioners in the case to access the logs to help determine whether or not their systems had been hacked and presidential election results altered. Second, the IEBC was also indicted for refusing to grant access to its servers holding the electronically transmitted results forms. There was also the related concern of the accuracy of the results displayed on IEBC’s public portal and its website. As a result of these concerns, the court, by a majority decision of four judges to two, concluded that the elections were not conducted in accordance with the Constitution of Kenya and further that the elections were marred with illegalities and irregularities. The court annulled the August 2017 presidential election.

Audit of computer servers

Flaws in the use of technology played a significant role in the court’s decision to annul the August 2017 presidential election. The development and actual employment of election technology had become part of the political conflict from the outset. Competition and attendant conflicts over the control of the procurement of technology intensified. This in turn undermined public confidence in the IEBC and employment of the technology itself. The fresh presidential election held on October 26, 2017 was, therefore, not sufficient resolution for the grievances voiced by the opposition. The failure to grant access to the servers when the court ordered so, and when the opposition supporters organised protests to demand that the servers be opened, became the basis for more disputes. The police violently dispersed the protesting groups, which led to an intensification of violence in areas where the opposition enjoyed great support. This made it difficult to hold elections in opposition-dominated areas where protesters argued that their leader, Raila Odinga, had won the August 8, 2017 election; that is why the IEBC refused to open the servers (Kahura, 2017; Oduor, 2017). The protests intensified when the election was held on October 26, 2017 and the Supreme Court subsequently declined to annul the result. Protests continued thereafter. The opposition regrouped and, in the end, called for a big public rally to swear in their leader, Raila Odinga, because ‘he had won the August 2017 presidential election.’ On 30 January, 2018, in a rally organised by the opposition, Odinga was installed as ‘the people’s president’. Again, the government reacted to this with increased violence by arresting certain leaders allied to the opposition on trumped-up charges. What began as a challenge to the veracity of technology turned into a major political crisis. The crisis ended on 9 March, 2018 when the president and the leader of the opposition publicly embraced each other and shook hands to demonstrate a unity of purpose after meeting discreetly to discuss the crisis. This ‘handshake’ marked the beginning of a new relationship between Kenyatta and Odinga.

Conclusion

Kenya has adopted use of new technologies to improve the credibility and transparency of elections following the recurrence of political disputes and election-related violence. The 2017 election in particular witnessed increased violence after the opposition protested against what they considered to be manipulation of the electronic transmission of results to favour the incumbent. Use of election technology has tended to reinforce the political divisions that characterize Kenya’s electoral context. Technology reinforced these divisions and provoked more disputes despite the existence of the transformative 2010 constitution that was passed against the backdrop of the 2007 post-election violence. The adoption of new technologies became an issue of contestation during different phases: from the development of the law, the procurement process, tallying and results, transmission of results, and in litigation. Disputes intensified because both the opposition and the ruling party sought to gain control over use of technologies, but the ruling party already had the IEBC on its side. For the opposition, this relationship meant that the ruling party and the IEBC would use technology in the transmission of results to reinforce the ruling party’s position and therefore that of the government.

Technology does not make elections any better politically; use of technology does not guarantee the credibility of elections. It may improve efficiency, especially by ensuring modernization of equipment for use in the election process, but it has become a ‘black box’, which few election players have the technical competence to ascertain how it operates. Political parties compete to exploit this black box because whoever gains control of it may turn it to their advantage to win an election.

The discussion has shown that technology cannot compensate for a lack of political commitment to electoral integrity. It is employed to address technical and administrative challenges but its use is undermined by competing political interests. Technology is, ironically, the new frontier of electoral fraud; it has helped seal the traditional loopholes of election rigging, but it is also a black box whose operations are understood by a minority.

The context on which election technology is employed matters. The introduction of electoral technology when there is poor electoral governance, deep ethno-political divisions and a lack of trust does not address the challenges that lead to electoral violence, it exacerbates the conditions for violence. As Yard (2010: 151) aptly puts it, any successful partnership between elections and technology must prioritise practice of ‘principles of fairness, equality, broad participation, and transparency.’ Use of election technologies where these democratic values are not deeply embedded in the political context will always lead to perceptions that technology will reinforce the interests of those in power and help them to consolidate their power and exclude others. Addressing these political challenges will help improve electoral governance and ensure that technology, when applied, will reinforce democratic reforms rather than create opportunities for electoral fraud.

Footnotes

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Karuti Kanyinga

Author biographies

Collins Odote is a Senior Lecturer and Director at the Centre for Advanced Studies in Environmental Law and Policy, University of Nairobi, and co-editor of Balancing the Scales of Electoral Justice: Resolving Disputes from the 2013 Elections in Kenya and the Emerging Jurisprudence (Nairobi: International Development Law Organization and Judiciary Training Institute, 2016).

Karuti Kanyinga is Research Professor of Development Studies at the Institute for Development Studies (IDS), University of Nairobi, and co-editor (with Duncan Okello) of Tensions and Reversals in Democratic Transitions: the Kenya 2007 Elections (Nairobi: Society for International Development and Institute for Development Studies, University of Nairobi, 2010; and co-editor with Nic Cheeseman and Gabriel Lynch (2020) The Oxford Handbook of Kenyan Politics. Oxford: Oxford University Press.

References

Africog (2017) Ready or not? An Assessment of Kenya’s Preparedness for the 8 August 2017 Elections. Nairobi: African Centre for Open Governance (Africog).

Amoah

(2019) Sleight is right: Cyber control as a new battleground for African elections. African Affairs 119(474): 68–89.

Aywa

(2016) A critique of Raila Odinga vs IEBC decision in light of legal standards for genuine presidential elections in Kenya. In: Odote

Musumba

(eds) Balancing the Scales of Electoral Justice: Resolving disputes from the 2013 elections in Kenya and the Emerging Jurisprudence. Nairobi: International Development Law Organization and Judiciary Training Institute, pp.46–77.

Barkan

(2013) Kenya’s 2013 elections: Technology is not democracy. Journal of Democracy 24(3): 156–165.

Bowman

Githaiga

(2015) Information Technology, the state and the voter in the Kenyan election of 2013. In Njogu

Wekesa

(eds) Kenya’s 2013 General Election: Stakes, Practices and Outcomes. Nairobi: Twaweza Communications Limited, pp.274–291.

Calista

Melitski

(2007) E-Government and e-governance: A convergence of public sector information and communication technologies. Public Administration Quarterly 31(1/2): 87–120.

Castells

(2004) The Information Age: Economy, Society and Culture, Vol. 2, The Power of Identity. Malden: Blackwell.

Cheeseman

(2008) The Kenyan elections of 2007: An introduction. Journal of Eastern African Studies 2(2): 166–184.

Cheeseman

Klaas

(2018) How to Rig an Election. New Haven and London: Yale University Press.

10.

Cheeseman

Lynch

Willis

(2018) Digital dilemmas: The unintended consequences of election technology. Democratization 25(8): 1397–1418.

11.

Dafoe

(2015) On technological determinism: A typology, scope, conditions and a mechanism. Science, Technology & Human Values 40(6): 1047–1076.

12.

D’Agostino

Schwester

Carizales , et al. (2011) A study of e-government and e-governance: An empirical examination of municipal websites. Public Administration Quarterly 35(1): 3–25.

13.

Dahl

(1989) Democracy and Its Critics. New Haven: Yale University Press.

14.

Debrah

Effah

Owusu-Mensah

(2019) Does the use of a biometric system guarantee and acceptable election’s outcome: Evidence from Ghana’s 2012 election. African Studies 78(3): 347–369.

15.

Evrensel

(2010) Introduction. In: Evrensel

(ed.) Voter Registration in Africa: A Comparative Analysis. Johannesburg: EISA, pp.1–56.

16.

Ezebuenyi

(2014) Cyber-politics: Analysis of new media and political information management interface for electoral participation in Nigeria. COOU Interdisciplinary Research Journal 1: 47–56.

17.

Garson

(2006) Public Information Technology and E-Governance: Managing the Virtual State. Sudbury, MA: Jones and Bartlett.

18.

Gelb

Clark

(2013) Identification for development: The biometrics revolution. Working Paper No. 315. Washington, DC: Centre for Global Development (CGD).

19.

Gelb

Decker

(2012) Cash at your fingertips: Biometric technology for transfers in developing countries. Review of Policy Research 29(1): 91–117.

20.

Githongo

(2008) Riding the tiger. Journal of Eastern African Studies 2(2): 359–367.

21.

Golden

Kramon

Ofosu

(2014) Electoral fraud and biometric identification machine failure in a competitive democracy. Available at: https://columbiacpseminar.files.wordpress.com/2015/04/golden-kramon-ofosu.pdf (accessed 13 August 2019).

22.

Gondi

Basant

(2015) Judicial activism versus restraint: Assessing constitutions and judicial reforms in Kenya’s 2013 Elections. In: Otieno

(ed.) New Constitutions, Same Old Challenges: Reflections on Kenya’s 2013 General Elections. Nairobi: Uraia Trust and Society for International Development, pp. 73–98.

23.

Harrington

Manji

(2015) Restoring the Leviathan? The Kenyan Supreme Court, constitutional transformation, and the presidential election of 2013. Journal of Eastern African Studies 9(2): 175–192.

24.

Harvey

(2003) The Fetish of Technology: Causes and Consequences. Macalester International 13(7). Available at: https://digitalcommons.macalester.edu/macintl/vol13/iss1/7.

25.

Heidi

Nyoike

(2016) A new dawn postponed: The new constitutional threshold for Valid Elections in Kenya and Section 83 of the Elections Act. In: Odote

Musumba

(eds) Balancing the Scales of Electoral Justice: Resolving disputes from the 2013 elections in Kenya and the emerging jurisprudence. Nairobi: International Development Law Organization and Judiciary Training Institute, pp.78–113.

26.

Hobbis

(2017) Voter integrity, trust and the promise of digital technologies: Biometric voter registration in Solomon Islands. Anthropological Forum 27(2) 114–134.

27.

Jacobsen

(2020) Biometric voter registration: A new modality of democracy assistance? Cooperation and Conflict 55(1): 127–148.

28.

Kahura

(2017) The mystery of the missing servers: Were the August 8 elections predetermined? Available at: https://www.theelephant.info/features/2017/10/25/mystery-of-the-missing-servers-were-the-august-8-elections-predetermined/ (accessed 13 June 2020).

29.

Kameri-Mbote

Kindiki

(2008) Trouble in Eden: How and why unresolved land issues landed ‘Peaceful Kenya’ in trouble in 2008. Forum for Development Studies 35(2):167–193.

30.

Kanyinga

(2011) Stopping a conflagration: The response of Kenyan civil society to the post-2007 election violence. South African Journal of Political Science 38(1): 85–109.

31.

Kanyinga

Okello

(2010) Tensions and Reversals in Democratic Transitions: The Kenya 2007 General Election. Nairobi: Society for International Development (SID) and Institute for Development Studies, University of Nairobi.

32.

Karen

Gibson

Long

(2014) Voting behaviour and electoral irregularities in Kenya’s 2013 elections. Journal of Eastern African Studies 8(1): 153–172.

33.

Langat

(2016) CORD legislators delay special House session to protest changes to election law. Available at: https://www.nation.co.ke/news/politics/cord-delay-house-session-protest-changes-election-law/1064-3493316-g72ct0z/index.html

34.

Maina

(2013) Verdict on Kenya’s presidential election: Five reasons the judgment fails the legal test. The East African. Available at: http://www.theeastafrican.co.ke/OpEd/comment/Five-reasons-Kenya-Supreme-Court-failed-poll-petition-test/434750-1753646-tm96jl/index.html (accessed 14 May 2019).

35.

Menya

(2015) Cord proposes radical changes in Okoa Kenya drive. Daily Nation. Available at: https://mobile.nation.co.ke/news/Cord-Okoa-Kenya-Referendum-Constitution-Amendment/1950946-2609180-format-xhtml-vnx4x1/index.html (accessed 12 June 2020).

36.

Mueller

(2008) The political economy of Kenya’s crisis. Journal of Eastern African Studies 2(2): 185–210.

37.

Mugica

(2015) The case for election technology. European View 14: 111–119.

38.

Muraya

(2017) NASA Parallel tallying centre in Kenya and clouds, Says Raila on solo debate. Available at: https://www.capitalfm.co.ke/news/2017/07/nasa-parallel-tallying-centre-in-kenya-and-clouds-raila-says-on-solo-debate/ (accessed 13 June 2019).

39.

Nwangu

Onah

VCO

Otuh

(2018) Elixir of electoral fraud: The impact of digital technology on the 2015 general elections in Nigeria. Cogent Social Sciences 4: 19.

40.

Oduor

(2017) Kenyan opposition NASA disputes election results on delayed transmission. The East African, 12 August. Available at: https://www.theeastafrican.co.ke/news/ea/Kenya-opposition-disputes-elections-results/4552908-4054958-lx7y33z/index.html (accessed 10 July 2019).

41.

Ombati

(2017) Gunmen raid presidential tallying centre, steal computers. Available at: https://www.standardmedia.co.ke/article/2001250385/gunmen-raid-nasa-presidential-tallying-centre-steal-computers?fb_comment_id=1131682910264756_1131724703593910 (accessed 12 June 2019).

42.

Opalo

(2016) Constitutional amendment through popular initiative: Tentative lessons from the ‘Okoa Kenya’ campaigns. Constitution Net, 26 January. Available at: http://constitutionnet.org/news/constitutional-amendment-through-popular-initiative-tentative-lessons-okoa-kenya-campaign (accessed 12 June 2020).

43.

Republic of Kenya (2008) Report of the Independent Review Commission on the General Elections Held in Kenya on 27th December 2007. Nairobi: Government Printers.

44.

Republic of Kenya (2016) Report of the Joint Parliamentary Select Committee on Matters Relating to the Independent Electoral and Boundaries Commission (IEBC). Nairobi: Parliament of Kenya.

45.

Republic of Kenya (2017) Presidential Petition No.1 of 2017: Raila Amolo Odinga and Another v Independent Electoral and Boundaries Commission & 2 others; & Attorney General and Another. Nairobi: Supreme Court of Kenya.

46.

Wolf

(2009) Poll poison? Politicians and polling in the 2007 Kenyan election. Journal of Contemporary African Studies 27(3): 279–304.

47.

Wyatt

(2008) Technological determinism is dead; long live technological determinism. In: Hackett

, et al. (eds) The Handbook of Science and Technology Studies. 3rd ed. London: The MIT Press, pp.165–180.

48.

Yard

(2010) Direct Democracy: Progress and Pitfalls of Elections Technology. Washington, DC: International Foundations of Electoral Systems. Available at: https://www.ifes.org/sites/default/files/20111026_direct_democracy_progress_and_pitfalls_election_technology_yard_0.pdf (accessed 12 May 2019).