Afghanistan Analysts Network – English

Political Landscape

The 2018 Election Observed (6) in Herat: Insecurity, organisational shambles, alleged rigging

S Reza Kazemi 16 min

Many have praised the parliamentary elections in Herat province in the far west of Afghanistan as second only to the capital Kabul in terms of turnout. There was indeed considerable enthusiasm and determination to vote from those who could get to the polls, but they were a restricted number, mainly those living in the provincial and district centres. In areas controlled or threatened by the Taleban, the vote either did not happen or was troubled. Elsewhere, numerous administrative and technical shortcomings and serious complaints of rigging hindered Heratis from exercising their franchise. AAN researcher Said Reza Kazemi, who observed the poll, reports that the description by provincial authorities that Herat’s election was ‘good’ suggests they must have very low standards.

Voters waiting for their turn to vote inside Masjid ul-Reza Mosque polling centre, while some are voting in polling stations and candidate agents and observers are either sitting or walking around, doing their work. Photo: Author/2018Voters waiting for their turn to vote inside Masjid ul-Reza Mosque polling centre, while some are voting in polling stations and candidate agents and observers are either sitting or walking around, doing their work. Photo: Author/2018

Reinforced security measures

As election day, 20 October 2018, approached, the Afghan government boosted the strength and presence of its military forces in Herat province. These extraordinary security measures, which lasted for about a day before to a day after the poll, were mostly limited to the provincial capital, Herat city, and district centres because large swathes of territory in the province, particularly areas far from the provincial and district centres, are either contested or partially or completely controlled by the Taleban.

In Herat city, the bolstered security forces were clearly visible. Security forces were stationed at strategic entry points to and key intersections in the provincial capital and checked any suspicious movements. Many other members of the security forces could be seen constantly patrolling in Ranger-type vehicles.Additionally, the Office of the Provincial Police Chief banned the use of motorbikes and rickshaws in the city and the districts leading to it on 18 October. All traffic was banned on election day. As a result, as AAN observed, there were few vehicles in the city and the road leading to it from Guzara and Injil districts on polling day. More vehicles started coming out as the day went on when security measures were somewhat relaxed. The transport ban was a response to a number of small-scale vehicle-borne improvised electronic device (IED) attacks in Herat city in the lead-up to election day. The attacks did not cause many casualties and it is not clear whether or not they were directly linked to the approaching elections (for details, see the author’s pre-election dispatch here).

The Taleban push to disrupt elections in several districts

Although the city of Herat was generally safe around polling day thanks largely to the strengthened security arrangements, conditions were markedly different in several districts. The Taleban, as in previous years, set out to stop as much voting as they could (for a countrywide assessment, see pages 1-6 of this UNAMA report). According to residents and district Independent Election Commission (IEC) officials, the Taleban used a range of measures in several districts of Herat province. They warned local populations not to take part in the elections and threatened to cut off their fingers if they did so. This threat was not carried out, however. Also, as will be detailed below, they abducted figures of influence in various villages in a bid to scare local inhabitants away from the elections (but released them soon afterwards following local mediation), carried out small-scale IED and rocket attacks and used small arms fire, predominantly against polling centres or other nearby compounds and land, rather than people.

However, the Taleban insurgents were unable and unwilling to create a bloodbath. Shedding blood, especially of civilians, destroys whatever local relations they may have gradually established in areas under their influence or control (for a case study of Taleban relations with district residents, read the author’s recent case study on Herat’s Obeh district here and an introduction to the subject here). Examples of how the Taleban operated around elections in five districts of the province give a flavour of their tactics.

In the north-western district of Gulran that borders Turkmenistan in the north and Iran in the west, only one out of seven polling centres opened on election day and, even then, only the centre in the very centre of the district. The IEC official in charge of Gulran told AAN that the Taleban began shelling the district centre from the moment the polling centre opened “early in the morning to around 11 am.” The rockets did not hit the polling centre but amplified a previous Taleban warning to Gulran residents not to participate in the elections. “Nevertheless, polling went on throughout the day in the district centre,” said the official. Although no one was hurt, the Taleban did appear to have affected turnout. Out of around 5,500 registered voters, only about 1,200 cast their votes in Gulran.

The western, mainly desert district of Kohsan through which Herat is connected to Iran via the Islam Qala border crossing is considered a comparatively safe district by Herat standards. There, Taleban insurgents carried out small-scale IED attacks against four polling centres including three schools and one mosque, according to an IEC official based in the district. The attacks injured two people (see this media report), inflicted some physical damage on polling centre buildings and significantly lowered turnout. In one of the polling centres, only 29 people turned out to vote. However, there was much greater voting in the remaining 11 polling centres in Kohsan (15 polling centres in total). All in all, nevertheless, Taleban threats and IED attacks reduced turnout in the district: of around 26,000 registered voters, only about 14,000 people voted.

Security in the central district of Adraskan along the Herat-Kandahar highway was worse, given its adjacency to restive Shindand district and Farah province in the south and Farsi district in the east. According to a UNAMA report (see pages 5 and 6 here), on 19 October 2018, the Taleban “gathered” 45 elders and figures of influence from several villages “with the aim of intimidating the population into not participating in the elections” and released them only after the elections were over in the evening of 20 October. A district IEC official based in Adraskan told AAN that some 25 election employees informed him on 19 October that they would not turn up to work the following day due to Taleban threats. On election day itself, the Taleban fired rockets near to five polling centres. In total, of the six polling centres, only the one in the district centre escaped attack. The attacks injured a soldier and a woman and caused some physical damage. At the end of the elections, of some 6,600 registered voters in Adraskan, only about 1,900 had cast their ballots.

Further south, the insecure district of Shindand, which borders Farah province and has been divided into five smaller districts, also witnessed a patchy, insecure and, in some places, non-existent election, according to a district IEC official who spoke to AAN. In two of the newly-created districts, Pushtkoh and Zerkoh, which are largely under Taleban rule, no election took place. In central Shindand district, all 11 polling centres opened. However, polling in two centres was hampered by skirmishes between the Taleban and Afghan government security forces nearby, which left casualties on both sides. Additionally, at about 08:00 when polling started, the Taleban fired a rocket on Shindand district centre to frighten local residents off the elections. There were no casualties, however. At the end of the day, out of about 8,400 registered voters, only some 5,000 voted in central Shindand district. In the district of Kuhzur, the single designated polling centre managed to open and out of some 1,100 registered voters, only a small group of people consisting of election staff, Afghan government security forces and a few others cast their ballots. In Zavul district, the two designated polling centres opened and out of around 2,600 people who had registered to vote, some 1,000 cast their votes.

That elections were held in troubled Shindand at all came as a surprise, but this has something to do with competing Taleban factions in southern Herat province (read AAN background on the conflict in Shindand here). Local sources in Herat city and Shindand told AAN that followers of the splinter Mullah Rasul group led by Mullah Nangialay allowed elections to be organised in areas under their influence or control (for more on this breakaway Taleban faction, read AAN analysis here). Clashes have continued for several years between Mullah Nangialay’s forces and sympathisers of mainstream Taleban leader Mullah Hibatullah led by Mullah Samad in Shindand. The most recent battle left several members on both factions dead and injured in late November 2018 (see this media report).

The Taleban also threatened the elections in Pashtun Zarghun, a relatively green and fertile, but sparsely populated district in the east of Herat province. A district IEC official told AAN that all 19 designated polling centres opened and of about 18,000 registered voters, about 9,800 people cast their votes. He said that the Taleban, angry at this relatively high turnout, despite their warnings and threats, abducted elders from several villages and only released them following mediation by other local influential figures. Additionally, UNAMA says in its report (see page 6 here) that in Chiworshy village of Pashtun Zargun district, “on 20 October Taliban entered a private company where election materials were stored and burned the materials and the civilian building.”

There were similar security incidents (mostly small-scale rocket and IED attacks) in several other districts: eastern Obeh and northern Kushk-e Kuhna districts (see page 5 here) and northern Kushk-e Rubat Sangi, eastern Karukh and Chesht-e Sharif, western Ghoryan, southern Guzara and central Injil (see here).

Organisational failure across the province

Insecurity was not the only obstacle facing Heratis keen to vote. There was also widespread administrative shambles (for a countrywide assessment, read AAN reporting here , here and here).

The first problem on election day, experienced at least in and around Herat city, was related to access. AAN observed that the transport ban had knock-on effects on both electoral staff and voters. Many IEC employees, who had been mainly recruited from among local school staff, did not manage to arrive at many polling centres on time.

Voters waiting in a long queue to get in Masjid ul-Reza Mosque polling centre to vote. Located in Police District (PD) 3 of Herat city, the entrance to the mosque is surrounded by a concrete blast wall. A young man, echoing the words of several other men who had come to vote in this polling centre, told AAN, “Look at the long queue! It’s frustrating and it’s better that I go because I’ve got other things to do.”

Voters waiting in a long queue to get in Masjid ul-Reza Mosque polling centre to vote. Located in Police District (PD) 3 of Herat city, the entrance to the mosque is surrounded by a concrete blast wall. A young man, echoing the words of several other men who had come to vote in this polling centre, told AAN, “Look at the long queue! It’s frustrating and it’s better that I go because I’ve got other things to do.” Photo: Author/2018

Voters faced a similar challenge. AAN saw large numbers of people walking to and from polling centres in Herat city and the neighbouring district of Injil. Some we spoke to did not know they had to vote in the polling centre at which they had registered; then, if they went to the wrong centre, it was difficult to rectify this mistake because of the transport ban.

Those affected by one of the major administrative failings of this election – the names of some voters who had registered not appearing on the list of the relevant polling centre – also found their attempts to vote exacerbated by the transport ban. Many of those who could not find their names at the centre where they had registered then trailed round several other centres in a sometimes vain attempt to find their names so that they could vote. In the most notorious case of this kind, the IEC sent some voter lists from Bamiyan province to Ghoryan district of Herat (see this media report).

The late arrival of staff and generally slow administration of voting led to large queues of voters forming outside many polling centres. People waited between one and six hours to vote, with some people deciding not to wait or waiting in vain as their polling centre did not open at all. “Look at the long queue! It’s frustrating and it’s better that I go because I’ve got other things to do,” a young man who had approached Masjid ul-Reza polling centre in Police District (PD) 3 told AAN, echoing the words of several other men who came to vote, but then left after seeing the long queue. Some people living nearby, however, paid two or more visits to the polling centre and finally managed to get in and vote when it was less crowded.

Similarly, in Khaja Muhammad Taki High School polling centre in PD 3 and Hatefi High School polling centre in PD 4, many people waited for a long time (an hour or more) to get inside to cast their votes, with some deciding to return home without voting at all. As the day went on, however, queues got shorter, especially from late afternoon onwards.

Such late starts were observed by or reported to AAN all over the province. However, in some places, the delays gave rise to speculation that the administrative chaos was deliberate, a move by the government to disenfranchise voters in particular areas of Herat city and the larger province.

Besides this, there was also chaos inside most polling centres across Herat province, according to various sources. AAN observed:

Voters struggled to find their names on voter lists in the Masjid ul-Reza polling centre because voter lists were either missing or wrong or had been misplaced. A middle-aged man said, “I’ve waited for an hour or so to get inside the polling centre but can’t find my name now that I’m in. They keep sending me to one polling station or another.”

A bustling but problematic polling process in Gawharshad High School polling centre in PD 1 of Herat city. There, a candidate agent told AAN: “Polling began late, voters have to wait for a long time to get in, they find their names with difficulty or don’t at all, biometric devices don’t work and so on. Candidate agents have filled out so many complaint forms that no more is left.” Photo: Author/2018

A bustling but problematic polling process in Gawharshad High School polling centre in PD 1 of Herat city. There, a candidate agent told AAN: “Polling began late, voters have to wait for a long time to get in, they find their names with difficulty or don’t at all, biometric devices don’t work and so on. Candidate agents have filled out so many complaint forms that no more is left.” Photo: Author/2018

Confused voters in Gawharshad High School polling centre in PD 1 also struggled to find their names on voter lists amid a large crowd of candidate agents, who were looking for complaint registration forms that had already run out, and a large number of observers; altogether, the orderly administration of the election was hindered as a candidate agent told AAN:

“Polling began late, voters have to wait for a long time to get in, they find their names with difficulty or don’t at all, biometric devices don’t work and so on. Candidate agents have filled out so many complaint forms that no more is left.”

In some polling centres, especially in and around Herat city, polling continued well into the night to let all those standing in queues cast their votes. A candidate agent in Nasaji polling centre in Guzara district told AAN:

“The polling centre finally opened around 1 pm. There was a long queue outside the centre but many people got tired and went back to their houses. When polling began, my colleagues and I coordinated and encouraged people to return to vote. Polling continued well into the night. The candidate we were working for paid for vehicle owners to keep bringing people to vote and then returning them to their houses. We were in the polling centre till about midnight.”

In the districts, a range of similar administrative flaws were reported by the media (see Killid and page 5 of this issue of Hasht-e Sobh daily newspaper). They included delays in opening polling centres; missing, wrong or misplaced voter lists; inadequate numbers of complaint registration forms; biometric voter verification (BVV) devices unavailable, not working, not being charged or IEC staff struggling to operate them and; election materials not provided on time or not provided at all, especially in polling centres located in outlying areas.

The provincial IEC tried to address the inadequacies. According to provincial IEC head Ahmad Shah Qanuni, who spoke to AAN, a 10 to 12-member operational team at the provincial IEC headquarters was busy particularly throughout election day, attempting to resolve the myriad of problems. However, the organisational and technical shortfalls were numerous and simply unmanageable for the provincial IEC. “Technical problems played a greater role in disenfranchising voters in Herat than did insecurity,” was the frank admission of provincial IEC head Qanuni in an interview with AAN. This would appear certainly to be true on election day, itself, although many voters never even got the chance to register because of the Taleban controlling large areas of the province and beyond.

More disenfranchisement due to technical flaws than security incidents

In assessing October’s parliamentary poll, it should be stressed that, months before election day, many Heratis had already been disenfranchised. Out of an estimated one million voters in the province, only around half registered to vote; the shortfall appeared due mostly to security threats and widespread disillusionment with the last fraudulent and controversial presidential elections in 2014 (see the author’s pre-election dispatch here).

The IEC had planned to open 462 polling centres across Herat (see their list here). Prior to election day, it said it would only be able to open 300 of these following an assessment by the Afghan government security institutions (read this AAN dispatch). Then, on election day itself, it opened only 285. On the extra, second day of the elections, the IEC did plan to open the remaining 15 polling centres, but in practice only opened two, one in Kushk-e Rubat Sangi district and the other in Adraskan district (see also this media report). This means that, on the two election days, only 287 polling centres opened in Herat province, 175 polling centres (or almost 40 per cent) fewer than was planned. Only in Herat city and the immediate district of Injil did all designated polling centres open.

About a fifth fewer centres opened in 2018 than in the last parliamentary elections in 2010 when 351 polling centres opened (see their list here) and only 4 per cent of those planned to open failed to do so (see appendix 4 of this report; see also AAN reporting on the 2010 elections in Herat here and here). This is a consequence of the Taleban controlling or influencing more territory than they did eight years ago.

As for the number of people who cast their votes on the two election days, there are no exact figures yet, but available data does indicate that more than 215,000 registered voters did not exercise their franchise. Prior to the elections, the provincial IEC provided AAN with the following voter registration figures (see the author’s pre-election dispatch here):

  • Total: 557,720 registered voters
  • 308,613 men
  • 247,434 women
  • 1,673 kuchis (nomads)

After the election, on 28 November 2018, Qanuni, the provincial IEC head, provided the following figures of those who had voted:

  • Total: 342,225 voters
  • 184,180 men
  • 157,190 women
  • 855 kuchis

215,495 registered voters, almost 40 per cent (38.64 %), did not vote in Herat province. Individual reasons are not known, but potential voters faced technical shambles, closed polling centres and Taleban threats or were disinclined to vote. Compared to the 2010 parliamentary elections, there was actually little difference: overall 348,145 people (198,483 men, 145,050 women, 4,612 kuchis) voted in 2010; this includes both validated (291,625) and invalidated (56,520) ballots. This is just 5,920 more ballots than those who cast their votes in the recent elections.

Alleged rigging

Allegations of fraud have further undermined the parliamentary elections in Herat. As time has passed since the elections, the number of election complaints has steadily gone up: from 649 on 22 October 2018 (see page 5 here), to 792 on 24 October (see here) and to 999 on 28 November, according to the head of the provincial Election Complaints Commission (ECC), Fareshtah Hesham.

Hesham told AAN the provincial ECC had put the 999 complaints into three categories. The first category involves 649 complaints that have been found by the provincial ECC as “undocumented and therefore rejected.” The fact that there has not been much local reaction to the rejection of these complaints could show that the complaints were not solid. The second category comprises 150 complaints about organisational shortcomings such as late starts, missing or wrong or misplaced voter lists and dysfunctional BVV devices on polling day. These complaints have been referred by the provincial ECC to its headquarters in Kabul, for “these problems emanated from the centre [Kabul],” said Hesham.

The remaining 200 complaints are, in Hesham’s words, “evidence-based and serious.” These complaints relate to alleged cases of election manipulation such as ballot box stuffing, the use of fake tazkeras(national ID cards), campaigning for candidates near and even inside polling centres and general illegal interference by different actors in the electoral process. The provincial ECC investigation of these complaints is still under way.

As a result of the third category of complaints, Hesham said the provincial ECC has quarantined ballot boxes from 61 polling centres across the province for re-counting. These include 18 polling centres in Herat city; all polling centres in Chesht-e Sharif, Kushk-e Kuhna, Gulran, Shindand, Adraskan, Zavul and Farsi districts; and some polling centres in Zendajan, Guzara, Injil, Karukh and Obeh districts. The quarantined polling centres indicate that alleged fraud was perpetrated both in relatively safe and central districts such as Injil and insecure and faraway ones such as Farsi and, in other words, across the entire province.

Local observers told AAN that the recruitment of election staff from among the local population contributed to the fraud. For instance, an observer from Kohsan district told AAN that at least two polling centres in the district were staffed by individuals belonging to one kinship group who supported one of the candidates and purportedly engineered the election in favour of him. In other areas, according to local journalists, candidate agents and provincial IEC and ECC heads that AAN spoke to, various actors including locally-hired IEC staff, elders, commanders and candidate agents interfered in the elections in favour of their chosen candidates. Furthermore, the lack of candidate agents and observers in some polling centres situated in far-flung, insecure areas provided an environment conducive for rigging.

In addition, there are concerns among some local activists that electoral fraud at the central Kabul level could have undermined Herat province’s parliamentary elections even further. Some candidates immediately left Herat for Kabul following the elections, which contributed to speculation among journalists and local activists including observers and civil society representatives that they are allegedly “trying to increase their votes through the electoral bodies or are concerned about election manipulation by the electoral bodies” (see page 5 here). These local activists (see page 5 here) have further alleged that some candidates intentionally declared their victory prematurely by throwing parties in order to make Kabul-level vote manipulation possible later on. The provincial IEC and ECC have tried to assure the populace that due process of law was being followed, but the credibility of this assurance is under question. For example, on 7 December 2018, 25 parliamentary election candidates in Herat called for the complete invalidation of votes in the province, citing what they called “widespread electoral fraud and violations” including by the provincial IEC (see here).

In her interview with AAN, the provincial ECC head Hesham said she fully expected the announcement of preliminary results – which came out on 17 December (1) – would bring to them an even greater number of complaints.

Conclusion: what kind of an election did Herat have?

Enthusiasm by Heratis to take part in the elections was evident from the long queues of voters that formed outside polling centres in various parts of Herat province. Yet, many Heratis could not exercise their franchise, meaning the elections were far from representative of the population across the province. Insecurity primarily caused by the Taleban meant that polling was mostly limited to provincial and district centres; this was largely a vote of Herat’s urban population. That widening urban-rural divide, part of a countrywide pattern (read this AAN dispatch), raises questions as to how representative returning MPs to Afghanistan’s next parliament will be of the population as a whole. A vote constrained by insurgency was then further chipped away at by technical deficiencies. Even in safe areas, not every Herati managed to exercise their franchise.

Those twin problems – insecurity and organisational failure – were reflected in the voting statistics, the 175 polling centres that did not open and the 215,495 registered voters who did not vote. Given that half of Herat’s estimated eligible voters had already not or could not register, the figures look even worse. Actual voter turnout on the day was 342,225, or about a third of the estimated electorate.

These shortcomings were compounded by the 200 serious complaints of election rigging that cast serious doubts on how free and fair Herat’s parliamentary elections were. Although these complaints are still being investigated by the provincial ECC, convincing electoral stakeholders, particularly ‘losing’ candidates, to accept the final results could be a very hard task, given the messiness of these elections. It could have consequences for public order, given the commonplace practice of the ‘ill-used’, but influential, to bring supporters out onto the streets and block roads in and around the city of Herat and elsewhere. More importantly, these problematic parliamentary elections could further undermine public trust not just in the capacity of the electoral institutions but also in Afghanistan’s ‘democratic’ processes more broadly. This is particularly worrying given the approaching 20 April 2019 presidential elections.

Despite all these issues, however, according to the provincial authorities, Herat had a ‘good’ election. Muhammad Asif Rahimi, the outgoing provincial governor, said the elections were “relatively safe.” The Taleban failed to prevent the poll, he said, and the people of Herat, particularly the women, “created an epic by their high turnout” (see page 5 here). In her interview with AAN, the provincial ECC head, Hesham, put it more bluntly, “Compare Herat with Ghazni [province] that had no elections: the elections [in Herat] were good and things went safe and sound (bakhair ter shod).”

The enthusiasm on the part of those voters who could vote has to be acknowledged. Where voters could vote they largely did. However, many could not participate. They were let down by violent insurgents and incompetent authorities and, quite likely, cheats. This poses a serious question: how representative, free and fair can even a ‘good election’ in Afghanistan in current times be?


Edited by Danielle Moylan and Kate Clark


(1) The IEC announced the preliminary election results for Herat province on 17 December 2018 (see here). 161 candidates (their full list here) campaigned for Herat’s 17 seats in the parliament (five reserved for women). According to the preliminary results, of the 17 leading candidates (the first 12 are men and the last five women), eight are sitting MPs and the remaining nine new faces – who are referred to in bold:

  1. Habib ul-Rahman Pedram, 16,796 votes
  2. Muhammad Reza Khushak Watandost, 16,140 votes
  3. Omar Nasir Mujaddedi, 10,929 votes
  4. Haji Muhammad Sadeq Qaderi, 10,171 votes
  5. Munavar Shah Bahaduri, 8,324 votes
  6. Nesar Ahmad Faizi Ghoryani, 7,612 votes
  7. Naqibullah Arwin, 7,306 votes
  8. Ghulam Faruq Majruh, 7,004 votes
  9. Ustad Hamidullah Hanif, 6,946 votes
  10. Haji Shahpur Popal, 6,920 votes
  11. Sayyed Azim Kabarzani, 6,582 votes
  12. Al-Haj Ghulam Faruq Nazari, 6,263 votes
  13. Rahima Jami, 3,686 votes
  14. Masuda Karukhi, 2,822 votes
  15. Simin Barekzai, 2,323 votes
  16. Nahid Ahmadi Farid, 1,959 votes
  17. Ustad Shirin Shahabi, 1,737 votes


2018 elections election parliamentary election