Afghanistan Analysts Network – English

Political Landscape

Afghanistan’s 2019 Election (14): In Herat, power and comms failure, district insecurity and low turnout

S Reza Kazemi 23 min

There was a slight increase in campaigning in the last few days ahead of the poll, especially in the provincial capital, but E-Day itself suffered from electricity and telecommunication failures in many parts Herat province. Worse, the Taleban launched attacks on areas near polling centres in districts across the province with a view mostly to intimidating voters into staying at home. Widespread voter disillusionment, particularly since the previous mismanaged parliamentary elections in 2018, all contributed to a significantly low turnout: with about 20 per cent of registered voters casting their ballots, according to the latest figures released by the central Independent Election Commission (IEC) in Kabul. Reporting on both macro-provincial and micro-level of a few polling centres in Herat city, AAN researcher Reza Kazemi details how the poll went off in this populous and therefore electorally-significant province.

An IEC employee takes a photo of a voter by using a BVV device in Khaja Muhammad Taki High School polling centre. Photo: Said Reza Kazemi.

Herat elections: security

  • Electricity was cut off on the eve of the elections in northern districts and parts of Herat city. It seems the Taleban interrupted electricity to disrupt the elections, particularly intra-IEC communications.
  • In addition, telecommunications were down before, during and after the polling day. Various theories have emerged as to why this happened: Taleban disrupting the elections, the government wanting a news blackout of any attacks, or one or a number of actors providing conditions for possible fraud.
  • There were a number of clashes in different districts around polling day, with casualties for both the Taleban and government. However, for the most part, the Taleban did not directly target polling centres and seem to have wanted to warn off and scare potential voters, rather than kill them.

Herat elections: polling

  • Technically, the balloting was carried out much better than in 2018 parliamentary elections; there were fewer problems with the Biometric Voter Verification (BVV) devices, voters lists and staffing at the polling centres.

Herat elections: turnout

  • Initial data released by IEC central shows 253 polling centres were open, of which 220 reported votes had been cast and the remaining 33 that there had been no voting. The IEC originally planned that 300 polling centres would open; after a security assessment ahead of the elections, this dropped to 268. On E-Day, only 253 centres actually opened.
  • The latest IEC central data shows around 130,000 people cast their ballots, out of a total number of 571,000 registered voters (almost 23 per cent). Although Herat’s turnout is the sixth largest nationwide, it was significantly lower in this election than previous ones; in the 2018 elections, for example, some 340,000 people voted (almost 60 per cent).
  • This data should be taken with a grain of salt. The IEC, both its headquarters and provincial office, has yet to release turnout figures per polling centre. There is also a caveat about the provincial Election Complaints Commission (ECC) claim that no fraud took place, because as time passes, more problems may come to light, even though the deadline for lodging complaints has already expired.

Pre-polling: uptick in campaigning, boost in security and breakdown in telecommunications

Although campaigning was mostly insipid in Herat (read the author’s pre-election dispatch), it became busier in the last few days before the poll when candidates and their local supporters rushed to woo a reluctant voter population. More campaign billboards were put up, more rallies were held and one more candidate, Abdullah Abdullah, showed up in town. (1) [corrected on 5 October 2019 at 19:30]

This last-ditch push was partly facilitated by Herat’s parliamentarians and other figures of influence; they at last came out in support of specific presidential hopefuls. In his campaign rally in Herat city, for example, Abdullah was surrounded by four Herat MPs, Habib ul-Rahman Pedram, Ghulam Faruq Majruh, Qazi Nazir Ahmad Hanafi and Sayyed Azim Kabarzani. Some other influential figures simultaneously gathered and met Ashraf Ghani in the capital Kabul to back his re-election bid; among them were, for instance, Mawlawi Khodadad Saleh, the head of the ulama (religious scholars) council in western Afghanistan, and MP Muhammad Reza Khoshak Watandost. Based on ethnic, political party and other links with presidential aspirants or their running mates, this emerging alignment has been a product of behind-the-scenes quid pro quo negotiations.

As campaigning moved towards a slightly bustling tempo, there was also increased security posturing. A day before the poll, Herat city and district gates were closed to heavy vehicles such as trucks. On the evening before the elections, the provincial government banned all motorised traffic in Herat city, except for vehicles that had received special permits. As more troops were dispatched to secure elections in the districts, fewer were seen by AAN at the entry points to and intersections in Herat city. An officer from the provincial National Directorate of Security (NDS) office told AAN:

The government has sent more security forces, even traffic police officers, to districts such as Zenda Jan and Ghoryan [in the west] to provide security for the election there. It’s because Herat city itself is safe and there isn’t a need for lots of troops given that its surroundings are well-protected.

For their part, Taleban vowed to disrupt elections in Herat. Two Independent Election Commission (IEC) employees in Gulran district in the north and Shindand in the south told AAN, for example, that local Taleban had warned them and their colleagues not to work and the people in general not to vote, saying polling centres were military targets. In Gulran, warning notices appeared on some mosque and school walls, threatening to bomb polling sites and cut off the fingers of those who voted. In the same district, the Taleban even went door to door in some neighbourhoods, telling people to keep away from the polling centres. In Shindand, an IEC employee told AAN:

A few days ago, some people who had held a campaign event for a candidate were abducted by the Taleban. Their fate is not known yet. The Taleban have warned the local population not to take part in the elections. Some people have bought whatever foodstuff they need for a couple of days as the Taleban might not let anyone get out of their areas for the period around the elections.

Telecommunications also became a major battleground. In most districts in Herat, the service has been stopped from about 6 pm to 6 am the following day for several hours (eg Kushk in the north, Ghoryan in the west, Adraskan in the south and Pashtun Zarghun in the east). It has been worse in some districts like Zavul in the south where services are off from about 9 am to 3 pm and then again from about 6 pm to 6 am the following day. Before polling, however, there were rumours that mobile phone service – and electricity – might go off all day long to undermine the elections, especially by disrupting intra-IEC communications needed to manage the elections and transfer polling centre data from the field to the provincial IEC office in Herat city and then on to IEC headquarters in Kabul.

The rumours did indeed become reality. On the night of 27 September, hours before polling commenced, an improved explosive device (IED) destroyed electricity transmission towers in Kushk district in the north, cutting off electricity supply imported from Turkmenistan. This left parts of the province (northern districts and parts of Herat city) without electricity for several days, before, during and after polling. At the same time, there was a telecommunication breakdown as all mobile network operators became dysfunctional around the polling day. Although the power shutdown seems to have been carried out by the Taleban, there were different theories over who caused the mobile phone service breakdown. Some say it was the Taleban wanting to disrupt the elections. Others wonder if it was the government trying to prevent reporting on security incidents that might have discouraged voters. Another theory voiced is that a news blackout would provide conditions for fraud, especially in more distant, less secure districts.

Since information is still emerging about how elections have gone off across Herat province, the remainder of this dispatch is organised in two sections. First, it looks at the polls in Herat province with focus on one polling centre in Herat city on E-Day. Second, it briefly returns to the bigger provincial picture, assessing the quantity, quality and security of the recent elections.

Polling: a detailed view from Herat city and one polling centre in particular

On polling day, roads were shut to traffic. The city was quiet and there was not the usual hustle and bustle on the streets, particularly in the busy Badmurghan Square. The police were stopping anyone who had, as usual, come out on their motorbikes. The road in which Masjed ul-Reza polling centre was located in police district 3 was blocked by two vehicles that had been parked at both ends, and there was a police Ranger vehicle parked in front of the entrance to the mosque where some eight police and national security officers had been assigned to provide security for polling. Another armed police officer stood on the roof of the mosque. The wall of the mosque had already been reinforced with a concrete blast wall a couple of years ago to protect it from attacks that have targeted Shia places of worship in Herat during the last few years (more on these attacks in the author’s previous dispatch). The mosque is also used as a high school called Shahid Sultani. Some 20 people were waiting outside the centre before polling began.

Masjed ul-Reza polling centre had nine male polling stations (five in the courtyard and four in classrooms in the basement) and eight female polling stations (all in the basement and separated by two doors, of which one connected to the courtyard and the other to a small alley where there was a special entrance for women to enter the mosque). This author mostly accessed male polling stations and could access female polling stations only at the end of polling when candidate agents and observers were briefly allowed by the IEC to go there to take photos of polling station result sheets.

There were many candidate agents and a few observers. In the male polling stations, this author counted 11 candidate agents for Ashraf Ghani, eight for Abdullah Abdullah, two for Ahmad Wali Massud, one for Nur ul-Haq Ulumi, one for Gulbuddin Hekmatyar and one for Sayyed Nurullah Jalili. In addition, three observers (one each for Election Complaint Commission (ECC), Free and Fair Election Forum of Afghanistan (FEFA) and Afghanistan Civil Society Forum organisation (ACSFo)) were counted. As for candidate agents in female polling stations, this author heard from a male candidate agent working for Abdullah that “candidate agents working for Ghani are more than those working for Abdullah. They’re about twice more, both in male and female polling stations.” He said they were in contact with Abdullah’s female candidate agents because “we received food for lunch from our campaign headquarters and gave them their share of the food.” The male candidate agent working for Jalili said there were also two female agents for his candidate at the site.

First, IEC employees read out ballot box lock numbers in each polling station to candidate agents and observers and showed them that biometric voter verification (BVV) devices were empty of any pre-existing data and working properly. In polling station 01, this author recorded these five ballot box lock numbers: P590611, P590612, P590613, P590614 and P590615. When ballot boxes were locked and everything else was ready on the IEC side, voters were let to come in to cast their ballots at 07:08 am.

IEC employees are reading out ballot box lock numbers and then locking ballot boxes in the presence of candidate agents and observers. Photo: author


They were the first people to come in to cast their ballots in Masjed ul-Reza polling centre in police district 3 of Herat city. Photo: author

At 07:40 am, there was no longer any queue at the polling centre, while voters kept coming in small numbers. It was quiet inside too. Not having had breakfast, many people inside the polling centre including IEC staff, candidate agents and observers bought food like cakes, biscuits and beverages from a grocery store that was open on the street outside. “One cannot fight on a hungry stomach,” the ECC observer said, referring to the famous Afghan proverb that one needs a full stomach to do business.

There was no queue in front of Masjed ul-Reza polling centre, although voters were coming in small numbers. Photo: author


This clearly contrasted with the 2018 parliamentary elections when there was a very long queue in Masjed ul-Reza polling centre throughout the day. Photo: author

By 08:04 am, even fewer voters were trickling in to vote. Most IEC staff were unoccupied. A man on crutches came in and was directed by the queue officer to one of the male polling stations in the basement. He did not find his name there and was redirected to one of the polling stations in the courtyard. He was tired and already had sweat on his forehead from walking up and down stairs. He finally found his name and cast his ballot. While leaving the polling centre, he told AAN:

I’ve come to vote with difficulty. You see I can only walk with crutches. People say, ‘We don’t go to vote because our votes don’t decide our president.’ They say, ‘It’s fraud that decides.’ Some others say, ‘It’s the USA that decides.’ But I hope my ballot and the ballots cast by others who have come out to vote today will be the only [factor] determining who our next president will be.

An elderly voter told AAN:

This year, IEC employees are well-trained. They’re more familiar with using the computer [BVV device]. I wished they had been so in last year’s election. I live nearby and, thank God, there aren’t many problems like insecurity in our area.

As it was not a busy polling day, IEC employees, observers and candidate agents present in the polling centre had ample time to talk to one another. A candidate agent said the IEC was better organised this time and thought more people would come as we moved towards afternoon, while the ECC observer thought prime polling time had already elapsed and turnout would be low in the city and much lower in the districts:

In this election, the IEC is prepared and everything is working well. A voter’s name is found easily and soon. The biometric device is working properly. Voters don’t spend a lot of time waiting to come in to cast their votes. But the problem is that there aren’t many people turning out to vote. People aren’t interested in getting out of their houses. There’s no queue. There is no bustle. Things might change, though, as we move towards noon as people might assess how elections go in their areas and then decide to get out and vote. Also, yesterday was Friday [weekend] and today is also off. Some people might have slept longer and they might get out of their houses later.

– A candidate agent

I’ve just come from Jami High School polling centre and Khwaja Muhammad Taki High School polling centre in the city. There weren’t any queues there, either. In all previous elections, early morning was the prime time when large numbers of voters came out to cast their ballots. It’s very different this time. It’s hot these days in Herat and I don’t think many will turn out in the heat of the afternoon. If this is the turnout in the city, turnout is likely to be much lower in the districts.

– ECC observer

There was little to no conversation about polling in the districts because the mobile networks were down.

In this polling centre, AAN saw several voters who went directly to the piped water area of the mosque to wash off the ink on their fingers after casting their ballots. When a voter was washing off the ink, an IEC officer told him not to do so for candidate agents and observers might object to it. The voter told him he was leaving for Kandahar by road the following day and he did not want to be stopped by the Taleban in their checkpoints. “They will cut off my finger if they see the ink,” he said. Far from condemning the voter, a candidate agent gave him advice. He suggested to him to use “even liquid bleach or acetone to clear off the ink.” The voter said he did not care about anything but washing the ink off. This brought to life the fear from the Taleban for taking part in the elections and the risk some Afghans decided to take to support their young democracy, despite all its shortcomings.

Although the IEC handled the elections well this time, there were still some problems. At around 09:30 am, five people who had come to vote were not allowed to do so because their names were not found either on the voter lists or the BVV devices of the polling stations. This was while, as AAN observed, their tazkeras (national ID papers) had stickers that showed they had registered at the polling centre. Around 10 am, a female voter’s name was found in a male polling station where she was directed and allowed to cast her vote. In the afternoon, the author came across a voter who had come from a faraway part of the city to cast his vote in Masjed ul-Reza polling centre. “I came here because my name wasn’t found in the polling centre I’d registered and they told me to come here to vote,” he said. A similar case relayed to AAN by a reliable source came to the notice of the author in Nassaji polling centre in Guzara district where a voter, not finding his name at the local polling centre where he was registered, decided not to go to the polling centre in the city where his name had, by mistake, ended up.

From about 11:00 am to 12:00 noon, AAN briefly observed polling in Khwaja Muhammad Taki High School polling centre, also located in police district 3 of Herat city. There were similar observations there. A FEFA observer said voter numbers had been very small since early morning when the polling centre opened. There was a queue of 30 people at the beginning, according to a candidate agent working for Abdullah, but there were no queues thereafter. A candidate agent working for Ghani said the BVV devices worked well, IEC employees were more skilled than last time and there were more polling stations, but only small numbers of people had turned out to cast their ballots. AAN saw several polling stations (eg polling station 09) in which IEC employees had nothing to do because there were no voters during the time AAN was observing. The polling centre officer there informed other officers of a central IEC decision made in Kabul that voters whose names were not found on the voter list but were found in the BVV device and vice versa were eligible to vote.

An IEC employee takes a photo of a voter by using a BVV device in Khaja Muhammad Taki High School polling centre. Photo: author

AAN returned to its observation of Masjed ul-Reza polling centre in the afternoon, from 01:25 pm onwards. As with breakfast, lunch was a concern for those working in the polling centre. While some campaign headquarters brought food for their candidate agents in this polling centre, others were paid money to prepare food on their own. Some IEC employees had engaged in andiwali (literally meaning comradeship), a practice of raising money from one another, paying, preparing, cooking and eating food in a shared room. So they had their lunch in a room in the mosque. They even, one after another, left their duty, made ablutions (wuzu) and said their prayers in the prayer hall that had not been designated for polling. This was easy to do because they were so few voters coming in.

The IEC staff, candidate agents and observers who had stayed in the polling centre since early morning were slowly becoming more familiar with one another. Among the many discussions they had was why fewer people were coming to vote in this election than the past, particularly the previous elections in 2018. A candidate agent said candidly:

People are disappointed because they question if it’s their votes that determine the president of Afghanistan. The ones already in office did little for the people and broke their hearts. The new ones aren’t known much. So the people ask why and for whom we should vote. I wouldn’t have voted if I weren’t a candidate agent. I only voted to have my finger inked so there won’t be any issues with the campaign headquarters when I get back later today or early tomorrow.

As the polling centre was not busy throughout the day, there was ample time for often frank conversation among IEC employees, candidate agents and observers including in the courtyard of Masjed ul-Reza Mosque. Photo: author


This clearly contrasted with the 2018 parliamentary elections when it was bustling inside Masjed ul-Reza polling centre throughout the day. Photo: author

By 02:30 pm, half an hour before polling was originally planned to end, of about 6,500 registered voters in Masjed ul-Reza polling centre, some 1,300 (20 per cent) had come and cast their ballots. In terms of gender, of some 3,300 male and 3,200 female registered voters, about 840 men (26 per cent) and 460 women (14 per cent) had turned out to vote in nine male and eight female polling stations in this polling centre, as shown respectively in charts 1 and 2 below.

Chart 1: Total votes for and votes cast in the nine male polling stations (numbered 01-09) in Masjed ul-Reza polling centre by 02:30 pm, half an hour before the originally planned end of polling. Chart: author


Chart 2: Total votes for and votes cast in the eight female polling stations (numbered 10-17) in Masjed ul-Reza polling centre by 02:30 pm, half an hour before the originally planned end of polling. Chart: author

Around 03:00 pm, voting was extended for two hours even though there was no queue in the polling centre. An IEC employee said light-heartedly, “In the past, polling was extended because people were waiting outside to come in to cast their ballots. Now polling is extended in the hope people will come, especially those who might have been asleep in the afternoon.”

By 04:00 pm, most people present in the polling centre were impatient and tired. Some candidate agents even told IEC employees to close polling and begin counting as so few voters were coming in. An IEC officer asked them to be patient as “one more hour will also pass.”

Around 04:30 pm, there was a wake-up call to those in the polling centre. A post began circulating on Facebook, especially among users in Herat, showing people in Masjed ul-Reza polling centre: four candidate agents were sitting around, playing games on a smartphone, and an IEC employee sleeping on his forearm on a table, surrounded by other employees looking bored. Some IEC employees and candidate agents began searching for the person who had posted this on Facebook. When they found the man, a candidate agent himself, he evaded questions about his actions, slowly sneaked out of the premises and then ran away, although he later came back and others, busy with recording vote counts, did not grill him. Before that, however, a heated discussion followed among IEC employees, candidate agents and observers, one of whom thought this was “a bad thing” for that candidate agent to do as it would “disgrace” them, the elections and especially the mosque – a place considered sacred.

At 05:06 pm, the gate of the mosque was finally closed and polling came to an end. In each polling station, IEC employees read out the number of votes and the names of the last people who had cast their ballots. Then, three ballot box locks were broken and ballots were taken out for counting. Ballots were sorted by candidate numbers on the ballot paper and were then counted for each candidate. Once this was done, the IEC employees matched the votes to the number of votes recorded in the BVV device and the ballot papers taken from ballot paper bundles. Wasted and invalid ballot papers were also recorded. Vote counts were then recorded in result sheets, copies of which were shared with candidate agents and observers. Finally, materials including ballot papers, unused ballot paper bundles and result sheets were placed into a protection bag that was then sealed. The protection bag was then put into the ballot box that was locked with three new locks (P590606, P590607 and P590608 for polling station 01). Counting in both male and female polling stations ended around 07:00 pm and a truck was waiting outside the mosque to transport the election materials to the provincial IEC office.

IEC employees are sorting and counting ballots cast for each candidate in polling station 01 of Masjed ul-Reza polling centre after polling came to an end around 05:00 pm. Photo: author

By the end of polling, in Masjed ul-Reza polling centre, of about 6,500 registered voters, some 1,570 (24 per cent) had voted, meaning that around 270 more people (4 per cent) turned out to vote during the two-hour extension period. In terms of gender, of around 3,300 male and 3,200 female registered voters, some 980 men (30 per cent) and 590 women (19 per cent) had turned out to vote in the nine male and eight female polling stations, as illustrated respectively in charts 3 and 4 below. On average, 108 ballots were cast in each male polling station and 74 in each female one.

Chart 3: Total votes for and votes cast in the nine male polling stations (numbered 01-09) in Masjed ul-Reza polling centre by end of polling at about 05:00 pm. Chart: author


Chart 4: Total votes for and votes cast in the eight female polling stations (numbered 10-17) in Masjed ul-Reza polling centre by end of polling at about 05:00 pm. Chart: author

Post-polling: low turnout, technical issues and district security incidents

Due to the electricity and telecommunication outages described above, no one including the provincial IEC office had any idea about turnout in Herat province, even by 1 October, two days or so after polling had ended. According to this media report, there were no communications between Herat city and 11 out of Herat province’s 19 districts throughout polling day. The disruption caused by the breakdown in power and telecommunication services was summed up by an IEC staff member in nearby Zenda Jan district to the west of Herat city who spoke to AAN on 1 October:

All telecommunications excluding [the public] Salaam network went down on the night before the elections. On polling day, Salaam was also disconnected. Although we were in Zenda Jan, which is near to Herat city, we didn’t have any contact with the IEC office in Herat city throughout the polling day. Our communications resumed around 01:20 pm the following day [on 29 September]. First it was Salaam that got reactivated and Roshan and Etisalat followed suit a couple of hours later in the afternoon. On E-Day, some of the polling centres also lacked electricity to power the devices that printed the BVV barcodes. We had power banks, but the printers could only work by getting connected to direct electricity current. We somehow managed to solve this problem by arranging for and using diesel generators in those polling centres.

In a press conference held by the provincial government the day after the elections (29 September), Daud Seddiqzad, provincial IEC head, said election materials had, by then, only come in from the police districts of Herat city and nearby districts of Enjil and Guzara. Even after all election materials reportedly arrived in the provincial IEC office a couple of days later, in an interview with the private Herat-based Asia TV, Seddiqzad gave no turnout figures and just said he “[doesn’t] have any information.” He later told Hasht-e Sobh that “it was up to the central IEC to make turnout figures available to the media.” Similarly, the head of the provincial ECC office, Fatema Baqeri, told AAN on 2 October that it lacked any data on either the numbers of polling centres open on the polling day or of people who turned out to vote.

When the central IEC in Kabul released preliminary provincial turnout figures, including for Herat, (2) there was increased questioning from local observers and activists in Herat about why the provincial IEC has continued refusing to release turnout figures with some even calling it “an effort to engineer election results.” According to the data released by the central IEC, 253 polling centres were open in Herat province for the 28 September 2019 presidential elections, of which 220 reported polling and the remaining 33 reported no polling. In the 220 polling centres, around 130,000 people (almost 23 per cent) turned out to cast their ballots, out of a total number of 571,000 registered voters. It was originally planned that 300 polling centres would open; after security assessment ahead of the election, this dropped to 268. On E-Day, only 253 actually opened (details on Herat’s registered voters and polling centres in the author’s pre-election dispatch). Nevertheless, the veracity of these figures can not yet be assessed because the IEC, neither the headquarters in Kabul nor its field office in Herat, have yet started to release results per polling centre.

Going by the preliminary figures released by the central IEC (see analysis here), although Herat’s turnout is the sixth largest nationwide, it is significantly low in this election in the province, compared, say, to the chaotic parliamentary elections in 2018 in which some 340,000 people voted (details in the author’s dispatch on Herat’s 2018 elections here). The total provincial turnout of almost 23 per cent based on preliminary figures seems not very different from the detailed case of the one polling centre described above that recorded an end-of-the-polling turnout of 24 per cent. This is surprising as Masjed ul-Reza is in a safe part of the city; one would expect it to be far higher than average.

Ali Jan Fasihi, an official of Afghanistan Civil Society Forum organisation (ACSFo), which fielded about 400 election observers in Herat province, told AAN that their reports indicate that “votes cast in polling stations ranged from fewer than 100 to up to 140” and that “fewer women than men turned out to vote in Herat province.” The Masjed ul-Reza polling centre lies in the middle of what Fasihi thought was the range of voting patterns in Herat province, ie an average of 108 votes per male and 74 per female polling station.

There is already intense speculation in town about why there was such a low turnout in this election in Herat province (see, for instance, this media report). Technical problems do not seem to have had much impact on turnout. In a conversation with AAN, the provincial ECC head, Baqeri, said they had registered some 490 complaints but emphasised “they are all technical and can’t lead to invalidation of all votes cast in specific polling centres. “We haven’t received any complaints on fraud.” In particular, Baqeri referred to BVV devices not working properly and mistakes in voter lists, as a result of which “some people have been prevented from exercising their right to vote.” However, there is a caveat here because as time passes, more problems may come to light, even though the deadline for lodging complaints has already expired.

By contrast, security was a major factor keeping voters at home to the extent that some insecure districts registered almost no votes at all. Examples are Koh-e Zur in the south and Farsi in the east, where around ten people turned out to vote in each district, according to an ECC observer based in Koh-e Zur and an IEC employee based in Zavul district which neighbours Farsi – both had talked to the IEC officer for Farsi. “The people who voted in Koh-e Zur were the polling centre employees,” the ECC observer relayed. Both Koh-e Zur and Farsi are insecure with vast swathes of territory under Taleban control or influence.

Turnout varied but was generally low in all the remaining 15 districts, excluding the two districts of Zir-e Koh and Pusht-e Koh in the south where there has been no voter registration. Here is a sample of three more districts, according to AAN’s conversations with IEC employees based in those districts.

In Pashtun Zarghun district in the east, out of around 10,000 registered voters, some 2,000 cast their ballots in 13 polling centres. In Zavul district in the south, there was a higher turnout where out of about 3,500 registered voters, some 2,300 voted in two polling centres. One reason for this is the huge Shindand airbase: of those who voted in the district, some 1,000 were military working in the airbase, said an IEC employee in conversation with AAN. There are also three villages near the airbase. Finally, in the nearby district of Zenda Jan, out of around 17,000 registered voters, some 4,600 voted in 13 polling centres, an IEC employee based there told AAN:

At first, I thought this was a low figure but when I later discussed this with my colleagues from other districts, they told me this turnout was in fact a big achievement in this election for the province.

The Taleban launched several attacks, mostly small arms fire and rocket propelled grenades, on areas near polling centres as well as on checkpoints, vehicles or convoys of Afghan government security forces on the polling day in different districts including Obeh in the east, Shindand in the south and Gulran, Kushk and Kushk-e Kohna in the north. It seems, however, that they mostly aimed at intimidating voters because their attacks were not directed at polling centres or seemingly intended to kill or injure voters and election workers. The only exception to this pattern occurred in Islam Qala, on the Iranian border in Kohsan district in the north-west, where two motorcyclists opened fire on civilians near a polling centre, wounding five of them.

For its part, the security forces, according to Governor Abdul Qayyum Rahimi in a press conference a day after the elections (29 September), “rapidly reacted to defeat the enemy’s conspiracies, killing about 25 armed Taleban members and injuring many others on the day of the elections.” He also claimed the security forces had prevented “tens of cases of explosions and terrorist attacks on the day of the elections.” Shindand also witnessed a Resolute Support airstrike on Taleban positions as well as a violent clash between the two breakaway Taleban factions there; the latter incident reportedly left 20 dead among members of the two Taleban groups (AAN background on the Shindand conflict here).


The IEC was better organised compared to the 2018 parliamentary poll, although there were still some technical issues related to the operation of BVV devices and the accuracy of voter lists that prevented some people from voting, especially in the morning. Electricity was cut off, especially in the north, and telecommunications broke down across the province around the polling day. The Taleban seem to have interrupted electricity, but various theories have emerged as to why mobile networks broke down: Taleban disrupting the elections, the government wanting a news blackout of any attacks, or one or a number of actors providing conditions for possible fraud.

All indications point to a significantly low turnout in the 28 September 2019 presidential elections in Herat province. One main factor for this was insecurity. There were a number of clashes in different districts around polling day, leaving casualties for the Taleban and government, the two key parties to the conflict. However, for the most part, the Taleban did not directly target polling centres and seem to have wanted to warn off and scare potential voters.

However, the biggest obstacle facing this election in Herat was citizens’ lack of interest in voting. Over 50 residents of Herat city including registered voters, candidate agents, observers, students, rickshaw drivers and shopkeepers told AAN the low turnout was the people’s response to the ways in which Afghanistan has been governed in the last years. It showed, as they viewed it, a lack of confidence, as well, that votes will actually determine who the next president will be. Other citizens who did cast their votes believed they could influence the result, sending up a flicker of hope in Afghanistan’s democracy, despite all its failings.

Edited by Kate Clark


(1) Five candidates had earlier come for campaigning: Ashraf Ghani, Faramarz Tamanna, Gulbuddin Hekmatyar, Nur ul-Haq Ulumi and Sayyed Nurullah Jalili (details in the author’s pre-election dispatch).

(2) Hasht-e Sobh, a private daily paper based in Kabul, 10 Mizan 1398/2 October 2019, page 3. Similar data has been released by other media (see, for example, this report). The latest figures released on 3 October 2019 can be found on the Facebook page of one of the IEC commissioners here.


Election 2019 Elections Herat Presidential elections