Afghanistan Analysts Network – English

Political Landscape

Afghanistan’s 2019 Election (15): Administrative and technical problems in Kabul city

William Maley 9 min

Continuing our reports of how the 2019 Afghan presidential election went, we take a close look at the voting in Kabul city. With the most registered voters of any city or province by a long way, Kabul is one of the most electorally-significant. With its mixed population, the voting intentions of its population are difficult to predict. William Maley, Farkhondeh Akbari, Srinjoy Bose and Nishank Motwani,* who were special guests of the Independent Election Commission (IEC), bring us their observations of the vote and report multiple technical and administrative failures.

An empty polling centre in Qala-ye Fathullah in central Kabul. Photo: Norwegian Afghanistan Committee.

Electoral observation, the focus of much attention in recent times, (1) can serve a number of specific purposes. It can function as a confidence-building measure for voters in a tense environment, especially if observers offer a channel through which voter concerns can be forwarded to those with the capacity to address problems that might have emerged as part of an electoral process. In certain circumstances, it can work against the practice of electoral fraud, although determined fraudsters may have many tools they can use and countering fraud can be a difficult undertaking. (2) Perhaps most usefully, it can permit the identification of problems in either the design or the implementation of an electoral process so that these can be considered, addressed and hopefully rectified by the responsible authorities. In that sense, serious electoral observation can and should contribute to the fostering of electoral integrity, a concept now the focus of significant professional discussion. (3)

The authors, three of whom have observed several Afghan elections previously, visited seven polling centres in the city of Kabul (numbers withheld to protect staff) and made the following observations. (For a map of Kabul’s police districts and brief descriptions, see this AAN guide.

Police District 4 (Taimani)

We arrived at this polling centre well before the scheduled opening time of 7 am. A small number of voters had already arrived and the queue began to grow significantly while we were there. In no sense was the polling centre even close to opening by the scheduled time; the first voter was not admitted until 7:28 am. The furniture to support voting was not in place and staff were attempting to use a kitchen knife to cut through the metal tabs that held in place the lids of the boxes containing election-sensitive material. It was far from clear at this stage who was authorised to do what, as the staffer wielding the knife did not have any identifying badge and the IEC vests for staff to wear were still trapped inside the box. There were no signs included in the kits to assist in directing voters to particular stations and the staff were obliged to create them themselves. The polling centre director informed the observers that training had concluded at 2 am on the morning of the election. The staff appeared as frustrated as the voters by the lack of guidance that they felt was inhibiting their ability to meet the expectations of the voters.

While the first voter was admitted at 7:28 am, he was not issued with a ballot paper until 7:45 am and was bounced around several stations trying to discover where his name was listed. He had finally cast his vote and left the centre at 7:59 am. The crowd of voters was becoming increasingly restive as a result of the delays. One voter was in a particularly aggravated state of mind, which was only when he was told the voter list containing his name could not be found and that it would be necessary for him to return to the centre later in the day. A particular problem that emerged in the centre was associated with the Biometric Voter Verification (BVV) devices. On a number of occasions, when a voter’s number was entered into the machine, an error message was generated and it was necessary to reboot the machine before it would accept the number in question (4) This significantly slowed down the processing of voters through the stations.

A different problem emerged in the women’s station. Two separate groups of women each arrived believing they had a mandate from the IEC to staff the centre. A heated exchange broke out between members of the two groups which rapidly acquired an ethnic tinge. This highlights the importance of great precision in the allocation of tasks and responsibilities within an electoral administration. There was also a staffer who plainly could not read names on the voter list; she stated that she had left her spectacles at home by accident.

In this station, some women voters were marking their ballots and placing them in the ballot box, but refusing to accept indelible ink on their fingers on the basis that this would expose them to threats from anti-government elements opposed to the holding of the election; in some cases, they had already cast their votes.

The situation in this polling centre was so unsettling that the observers decided to leave and return later on, at about 10:20 am, to see if things had settled down. On the whole they had, but the number of recorded votes at that stage was low, with the observers’ rough impression being that at that stage of the day, less than 10 per cent of those whose names appeared on the voter list had, in fact, voted. It was not possible to tell whether this reflected a broader pattern of low turnout, or whether voters who had arrived earlier when the situation was chaotic had simply opted not to vote.

Police District 2 (Karte Parwan)

The observers reached this polling centre just as presidential candidate Dr Abdullah was arriving to cast his vote. Media representatives informed the observers that they had received notice about 40 minutes earlier that he would be arriving at the centre. This centre was running much more smoothly than the first and lists of voters were on display at the main entrance. A number of issues arose in the women’s station. One woman objected strenuously to the removal of her niqab for the purposes of biometric identification. There was also a male voter who discovered that his name and identifying details had been recorded on the list for the women’s station; he was allowed to vote, but the centre director took great care to explain in a highly transparent way to representatives of different candidates what had happened and why this was being permitted.

Police District 3 (Silo)

This proved to be one of the quieter centres visited by the observers, but not without interest. One voter, known personally to one of the observers, was an official of the presidential palace, but although he had registered to vote in the 2018 election and had the appropriate sticker attached to his tazkera, his name could not be found on the printed or biometric lists and he was not permitted to vote. He stayed at the centre until 3:00 pm in the hope of being able to vote, but was unable to do so. A number of other voters were in exactly the same position and the observers also encountered voters who had registered in Baghlan and Kunduz, but had hoped to vote in Kabul on the not-unreasonable basis that the security situation in Baghlan and Kunduz had deteriorated so greatly that they could not safely return to the places where they had registered.

Police District 13 (Dasht-e Barchi)

At this polling centre, there were extensive problems of voters’ names not being recorded on the voter lists, despite the voters having registered and voted at the same venue in the 2018 election. An IEC staff member, present at the polling station since the start of voting, told the observers that approximately 500 people had been sent away because their names could not be found; she had recorded their names and taken photographs of their taskeras with her mobile phone. Candidate agents who spoke with the observers were deeply suspicious about how this had come to pass. Some voters complained that they were treated discourteously or with disrespect when their names could not be found on the lists. Others complained that some staff appeared to lack the literacy skills required to find voters’ names, or find them quickly.

Some voters had been told to wait until 2:30 pm in case instructions came that would allow them to vote at that point. Further enquiries revealed that a numerical code to allow a voter’s details to be entered into the BVV device even if the voter’s name had not been earlier uploaded had been delivered to the polling centre. The intention of the IEC was apparently that such a code could be used only to enable a single voter to vote; but those staffing the centre had realised that since such numeric codes had likely been issued sequentially rather than randomly, it was possible to guess what other such codes might be and use them to accommodate further voters whose names were missing from the list. At a certain point, however, the staff had been instructed to stop using the codes in this way.

In a polling station in Kabul’s Taimani neighbourhood, a staff was attempting to use a kitchen knife to cut through the metal tabs that held in place the lids of the boxes containing election-sensitive material. It was far from clear at this stage who was authorised to do what, as the staffer wielding the knife did not have any identifying badge and the IEC vests for staff to wear were still trapped inside the box. Photo: Authors.


In addition, at this centre, a number of women found that their names had been included in the men’s lists. Those who sought to do so were apparently permitted to vote in one of the men’s stations.

Police District 13 (Dasht-e Barchi)

Staff at this polling centre, in a rather inaccessible part of Dast-e Barchi, also reported that a significant number of voters’ names were missing from lists. The staff had again made use of a numeric code to activate the BVV devices, but in contrast to the experience at the previous polling centre, had received no instruction to cease using the code in the creative way we have described above.

At this centre, one male staffer of the IEC had discovered that his name had been included in the women’s list rather than the men’s list. He chose not to attempt to vote.

Police District 1 (Jada-ye Maiwand)

The observers were informed at this centre that polling had been extended to 5:00 pm, but by the time they arrived, there were no voters at the centre and it appeared that only a small percentage, most likely less than ten per cent, of registered voters had cast their ballots. One party agent at the centre stated that voters who had presented e-tazkeras had been unable to vote because the BVV devices would not recognise them and photocopies of the voters’ old paper tazkeras were not acceptable. However, the observers were not able to verify this directly. On the issue of numerical codes to allow access to BVV devices for voters whose names had not been uploaded, the staff told the observers they had been instructed to use such a code only after 5:00 pm, ie after voting had officially ended, for voters returned who had been turned away earlier.

Police District 10 (Qala-ye Fathullah)

The observers watched the votes being counted. Since the turnout here appeared to be low – as at other centres that the observers visited – the count wound up quite quickly. By 6:05 pm, just over an hour after the ballot closed at 5:00 pm, it was completed. It appeared to be conducted in a professional and transparent way, with candidate agents able clearly to view what was happening and with the staff explicitly highlighting the steps they were taking in the course of the count. Final result sheets were also prepared in a professional and transparent way.


The observers had noted at around midday that the IEC, had decided that those whose names did not appear on either printed or biometric lists could vote if they had their stickers. Carried to its logical extreme, this would imply the abandonment of the 2018 system whereby eligibility to vote depended upon the appearance of the name of a voter on a voters’ list and reversion to the pre-2018 system in which eligibility to vote depended upon the production of a document by a voter.

The problem of voters’ names missing from lists was widely flagged following the 2018 election, but seems to have recurred. What in effect amounts to the disenfranchisement of some voters on account of logistical or administrative failures requires urgent and immediate remedy. When voters have the courage to go to the polls in the face of grave security threats, they are entitled at the very least to expect that their commitment will not be undermined by problems of management and organisation.

Edited by Kate Clark

* Professor William Maley is Professor of Diplomacy at the Asia-Pacific College of Diplomacy, the Australian National University.

Farkhondeh Akbari, is a PhD candidate at the Asia-Pacific College of Diplomacy, the Australian National University where she is focusing her research on peace settlement negotiations with the Taleban in Afghanistan and Khmer Rouge in Cambodia.

Dr Srinjoy Bose is Assistant Professor in Politics and International Relations at the University of New South Wales, Sydney.

Dr Nishank Motwani is a Visiting Fellow at the Australian National University.

The authors would like to thank the Independent Election Commission (IEC) and its staff and the Independent Directorate of Local Governance (IDLG) for their help in making the observations of the election possible.


(1) See Eric C. Bjornlund, Beyond Free and Fair: Monitoring Elections and Building Democracy (Baltimore: The Johns Hopkins University Press, 2004); Susan D. Hyde, The Pseudo-Democrat’s Dilemma: Why Election Observation Became an International Norm (Ithaca: Cornell University Press, 2011); Judith G. Kelley, Monitoring Democracy: When International Election Observation Works, and Why It Often Fails (Princeton: Princeton University Press, 2012).
(2) William Maley and Michael Maley, ‘Appraising Electoral Fraud: Tensions and Complexities’, Conflict, Security and Development, vol16, no.6, 2016, pp653-671.
(3) See Pippa Norris, Why Electoral Integrity Matters (Cambridge: Cambridge University Press, 2014); Pippa Norris, Richard W. Frank, and Ferran Martínez i Coma (eds), Advancing Electoral Integrity (New York: Oxford University Press, 2014); Pippa Norris, Why Elections Fail (Cambridge: Cambridge University Press, 2015); Pippa Norris, Strengthening Electoral Integrity (Cambridge: Cambridge University Press, 2017).
(4) Whether such technology offers an appropriate tool for addressing challenging issues of electoral integrity is a more complex matter requiring further analysis. See Joel D. Barkan, ‘Technology is Not Democracy’, Journal of Democracy, vol24, no3, 2013, pp156-165; Ali Adili, Afghanistan elections conundrum (21): Biometric identification likely to spawn host of new problems (Kabul: Afghanistan Analysts Network, 19 October 2018); Nic Cheeseman, Gabrielle Lynch and Justin Willis, ‘Digital dilemmas: the unintended consequences of electoral technology’, Democratization, vol25, no8, 2018, pp1397-1418.



election Election 2019 Presidential election