Afghanistan Analysts Network – English

Political Landscape

Elections 2014 (32): A second round surge in turnout in Loya Paktia?

Kate Clark Pakteen Ibrahimi 13 min

The three provinces of Loya Paktia – Paktia, Khost and Paktika – have seen some of the worst fraud in Afghanistan’s previous elections, particularly the mass proxy use of women’s ballots, so it was understandable that reports of a high turnout in the second round there would be met with suspicion. However, those on the ground did report a surge in voting compared with the first round, including of actual women voters. They attributed this to tribally organised support for Ashraf Ghani, as well as mobilisation by the clergy and the raising of arbaki to protect the vote from Taleban and other threats. As Pakteen Ibrahimi who was in Paktia for the election for AAN, and Kate Clark report, local deals with Taleban not to attack polling centres may also have encouraged voting in some areas.

Village in Wazi Dzadran district. Photo: Thomas Ruttig (2006)

In the first round of the elections, the three provinces of Loya Paktia showed a strong showing for Ashraf Ghani (between 63 and 74 per cent), a weak showing for Abdullah (between 4 and 11 per cent) and a poor turnout (71 per cent of ballots were used in Paktia, 44 per cent in Paktika and only 33 per cent in Khost). (1) In other words, Ghani may have had ‘landslide’ victories in Loya Paktia in the first round, but on the national scale of things, it was not that important as absolute turnout figures were comparatively low in Loya Paktia. However, these numbers start to get significant if there was a surge in second round voting.

Particular focus, this week, has been on Khost province. Dr Abdullah’s team have complained about a turnout figure (which originated with them, although they say they do not know its source) of more than 400,000, up from – a confirmed – 130,000 in the first round; they say the figure is simply unbelievable and evidence of mass fraud. (2) They also allege mass ballot stuffing in Paktia and Paktika, as well as a number of other provinces. (3)

The IEC has yet to give any provincial breakdown for turnout, so any discussion of actual statistics, which are notoriously tricky, unknown or contradictory (AAN’s Thomas Ruttig makes some sense of them here) is impossible. However, an impression of the election can be built up from how it appeared on the ground.

For AAN, indications that there might have been a very different sort of second round, with a high turn out in a number of places, including of actual women voters, came on polling day itself and the following day. Given the history of fraudulent voting in the region, however, we had already decided to cross-check with more sources before publishing anything (in other words, before the Abdullah campaign made this such a particularly more sensitive issue). So, as well as our researcher, Pakteen Ibrahimi, who was in Paktia for the elections, we have also spoken to journalists and other observers in Khost and Paktia to try and get a fuller picture of how the second E-Day went. Their reports do corroborate each other and point, not only to a surge in turnout, but also the ways this was achieved. This is not to say there was not fraud as well – indeed, we also have some reports of this, and many districts had had no or only a thin presence of observers who might also have been pressured to report or not report certain details. Our suspicion at the moment is that there was both fraud and higher turnout.

How to mobilise in the southeast

Loya Paktia is overwhelmingly Pashtun, with some Tajik and Hazara communities in more urban places like Gardez and Urgun (Paktika). It is one of the three main larger Pashtun regions – the others being the east (mashreq) and the south (sometimes called Loy Kandahar) – and has a distinct character. The tribes here are smaller, as many of them live in the mountains and this has resulted in a stronger tribal organisation and institutions. This was seen, in 2001, for example, when Loya Paktia, through tribal gatherings, was the only region of Afghanistan where the Taleban were overthrown by the people, as one of the authors witnessed. Even though post-2001 developments have weakened these institutions, not least by their co-option by both the US military and the Kabul government, there remains a strong capacity for popular mobilisation here, once a consensus on action is reached.

Before the second round of voting, AAN reported on the reaching of just such a consensus between many of the major tribes in the region, in concert with the Ashraf Ghani campaign. We described a tribal agreement, binding on all members (tarun), that women family members should go out to vote. This was an astonishing development in this generally highly conservative region (even given the number of districts in Paktia and Khost which are known as more ‘progressive’ where such a decision would be easier to implement). Since the election, more details have come in of just how widespread the Ghani/tribal mobilisation was, with an agreement not just on women voters, but also making it mandatory for all members to vote for Ashraf Ghani and not commit fraud for Abdullah – with a penalty of being beaten or fined. There were also actually a range of agreements, at the regional, provincial and tribal level.

The taruns were key to the mobilisation, people said, but they also cited the ulama as critical in persuading people to vote, and to vote for Ghani. Ghani team members had campaigned strongly among the ulema, AAN was told. “Thousands of mullahs gave a pro-Ghani Friday sermon, the day before the vote,” said one journalist. “They played a visible role in mobilising people.” Another reporter described how, in Khost, “one influential cleric, Mawlawi Qudus, drove from village to village to appeal directly to the people, telling them it was wajib (compulsory according to religion) to vote and wajib to vote for Ghani.” In Paktia province, it was only in Zurmat, one of the most important Taleban centres of influence nationally, that mullahs did not preach for Ghani. There, it would have been too dangerous to risk being labelled a ‘PRT mullah’.

The taruns had also bound the tribes to raise the traditional (for the southeast), temporary, tribal defence mechanism known as arbaki (4) to protect polling centres. One journalist said the jirga had wanted 200 men per district, but the Ministry of Interior had agreed to a lower figure – 1000 across the three provinces. These were then distributed to where it was thought threats might come from. The reason for wanting the arbaki, reporters said, was not just the very real fear of the Taleban disrupting the vote. Elders were suspicious, they said, that General Muhammad Sharif Yaftali, commander of 203rd Thunder Military Corps (his takhalus indicates he is from the village in Badakhshan of former president, Burhanuddin Rabbani), as one reporter put it, “make an area insecure deliberately and use the votes for Dr Abdullah,” He said the tribes felt they needed “a backup”. Some local press reports also alleged the Afghan National Army was planning to fix the ballot).

On election day, some arbaki were in action against the Taleban, as will be seen. However, reports have also been coming in of very local, pre-election agreements with Taleban not to attack the voters. In Paktia, AAN spoke to people who said there had been similar deals in the districts of Gerda Tserai (where the Haqqanis are from), Dandi Patan and, on a smaller scale, Wazi Dzadran. (5) Everyone emphasised that the agreements were made at a very local level (below district commander level), were not official and had gone against Emirate policy (there may also have been a desire for ‘plausible deniability’).

According to local sources, following the drubbing the Taleban received in the media for failing to disrupt the first round, the leadership had urged their commanders to disrupt the second round and had wanted them to distribute night letters and threaten people, via the mosques, not to vote because ‘the US chooses the president’, and to fire at polling stations on the day itself. However, the sources said, the tarun, where it was agreed could not be ignored. An example was given of a particular village in Wazi Dzadran district, Paktia, which had a polling centre and was also where the local commander was from. If he and his men had disrupted the election there, AAN was told, and the centre had received fewer votes, the people would have said to him, “We have a tarun, you disturbed our election and you are responsible for this.” In other words his tribal reputation would have been tarnished. Another example given of how deals worked would be a local commander saying, “Use that polling centre, not this one, and you’ll be OK.”

In the three districts of Khost where the Dzadran tribe live, Spera, Nader Shah Kot and Dwa Manda, AAN had reports of tribal elders and clergy sitting down together and deciding to speak to the local Taleban. Commanders are reported to have said they could not officially agree to refrain from attacking, but would give informal assurances that there would be no violence against the election. In two of these districts, there were no Taleban attacks.

These deals are, at least, plausible. Where there is strong popular consensus, it is very difficult for Taleban to go against it, especially where there are good communications between elders and commanders. As there have never been strong Taleban bases or controlled territory in the southeast, they mainly operate cross-border and need, at least, elders’ tacit consent for their activities. Moreover, in the face of a high turnout, real efforts to disrupt the vote would have meant high casualties which, for local commanders, might well have been unthinkable. (AAN is hoping to publish a dispatch on how the Taleban ‘dealt’ with the election nationally, including a look at deals in other areas.)

Why vote for Ghani?

As to why the Ghani team managed to persuade Paktiawals to vote for him, they can probably be summed up in a trope heard by the authors several times:

The deeds of the house are moving from Kandahar to Loya Paktia! If we do not vote for Ghani, we will be the slaves of the Panjshiris for another 10 (or 50) years.

Ghani, a technocrat, who spent many years in the west and did not participate in the jihad, has not often really spoken of as a ‘son of the soil’ in Loya Paktia before. (6) However, in this election, he has come to be seen as one of the region’s own, an Ahmadzai from neighbouring Logar, which is also seen as part of Loya Paktia by many. Moreover, the ‘slave of the Panjshiris’ fear carries some weight here. In 2001, following the Northern Alliance capture of Kabul, Panjshiris from the Shura-ye Nazar network of Jamiat-e Islami filled all the security ministries and some others. People in the Loya Paktia region – with only one minister in the cabinet – felt unjustly excluded from the new Afghanistan. Moreover, the then defence minister, General Fahim, appointed one of his Panjshiri commanders, Gul Haidar, as security chief for the zone, and several local Tajik commanders from Gardez became local security chiefs; they were seen as the main source of crime in the region and it took many years to get rid of them (for more details, see AAN reporting here). Two of the local commanders, Abdullah Mujahed and ‘General’ Ziauddin, were campaigning for Dr Abdullah, although people said this was not a major reason why they did not support Abdullah. Since 2001, people have continued to feel under-represented in a polity dominated, as they see it, by a Kandahari-Panjshiri alliance. Ghani then, represented a possibility for change and Abdullah the fear of continuity or worse.

People did cite other reasons for supporting Ghani. As in other parts of the country, among Pashtuns and non-Pashtuns alike, some spoke about his professional background, especially in economics, and perceived ability to run the country well. There were also less obvious reasons. Some people said they had been angry that Ghani’s election banners had been torn down and mud put on his face and ‘on the Quran’ (his election symbol). Others had heard rumours that the chairman of the High Peace Council, Salahuddin Rabbani, seen as an ally of Dr Abdullah, had signed an agreement to recognise the Durand Line as Afghanistan’s border. However, these last two reasons sound like pretexts. The real passion and zeal in this election clearly stemmed from the desire to get a man seen as one of their own into the presidential palace.

Curiously, but not without precedent, the model of the Hazaras was also cited. A long-marginalised community, they have used mass, organised, political participation, including of women, to move into the mainstream.

E-Day – turnout, women, security

Journalists and other observers on the ground reported higher turnout across the region, even in normally insecure districts, such as Sabari and Terezai in Khost and Sayed Karam, Janikheil, Mirzaka in Paktia. AAN had reports of ballots running out in Shwak (at about 2 pm) and of contingency ballots requested, but not always able to be delivered to Sabari, Terizai, Mandozai, Lakan and Khost city. We have not been able to verify this with the IEC or get a complete picture of where reported shortages were. Moreover, as we reported in the first round, ballot shortages can be evidence of stuffing, as well as unexpectedly high turnout – or both. One man from Tanai district, Khost, Jawaz Khan, whose village had decided to vote, said he delayed announcing the death of his daughter, fearing others in the village would come to epxress their condolences, rather than going out to vote.

One journalist said he had seen queues of women waiting to vote in two districts of Paktia (Tsamkani/Chamkani and Ahmad Aba, the core Ahmadzai district) and had received reports of queues from the third (Dzadzi/Jaji). He explained this by saying these are districts where women are a little more educated. Elsewhere in Paktia, he said women’s turnout was not so high, but still higher than last time. AAN’s researcher saw older women voting in the Zurmat district centre, although he said the usual practice of men voting for other female family members (accepted by IEC officials in this region in earlier elections) had also taken place. Pictures of women voting – which are only snap-shots, of course and prove very little generally – can be seen here and here.

In Khost, a reporter said the women’s vote was also up and said Ghani’s team had spent a lot of money transporting women from their villages to polling centres, across the province, with buses going from 6am to 4pm; he had had reports of this, he said, “in almost all districts”.

The tarun was cited as responsible for getting the vote and also, why, as several sources told us, in Sayed Karam district of Paktia, pro-Abdullah senator, Abdul Hanan Haqwayun, who allegedly, with armed men tried to take ballot boxes home to be stuffed, was stopped and beaten.

Arbaki were reported to have been useful in several districts in protecting polling stations, including in Chamkani, Janikheil, Dzadzi Aryub and Dandi Patan. In the latter district, in the area of the Muqbel tribe – which had raised 42 arbaki to protect the polling stations – there was fighting with Taleban for a few hours on election day. In Sayed Karam, there were minor clashes, AAN was told, before it was brought under control by the arbaki (tribes in the district had provided five men per centre).

Generally, in Paktia, there was less violence than the first round, the major exception being Zurmat. Here, there was no discussion with local Taleban and no deals and there were attacks against the election, including on polling centres on E-Day and two ambushes on the convoy transporting the boxes to Gardez, a small one near Astugan and a more serious one in Zowu when two ANA were killed. Even so, even in Zurmat, some people did turn out to vote and act as agents for both candidates. The picture was mixed, said AAN’s reporter: some polling centres saw a large turnout, others saw few voters, and the Taleban played hide and seek with the security forces all day.

Did the vote get out? 

“I, as a citizen,” Dr Abdullah’s spokesman, Mujib al-Rahman Rahimi, told AAN, “would highly appreciate that [if there had been mass voting in Loya Paktia], but the people [Dr Abdullah’s agents] representing each district and in the centre said something else: no election has taken place in the majority of these areas. They said there was no election at all.”

Rahimi was scathing about the possibility that there had been an actual large turnout:

The tribes can mobilise? I doubt it. That is a myth. Women can’t come out of their houses there. How can they go out to vote for Ghani? 

Ashraf Ghani’s spokesman and deputy leader of Jombesh, Faizullah Zaki, responded, equally strongly:

In Panjshir, 97 per cent of people voted for Dr Abdullah in the first round. That is natural, that is normal. We didn’t question it. How can they not understand that, after a long period, someone from Loya Paktia is coming to the presidency? Why shouldn’t the tribes support him, the ulema go to mosques and encourage people to vote? Why shouldn’t the most orthodox mullahs in the country go to mosques and say: let your women go and vote? It is normal and logical.

Zaki accepted some fraud will also have taken place in the Loya Paktia, as, he said, it had been done in pro-Abdullah areas.

It seems likely that only when we get the reports in detail of independent observers (FEFA, TEFA and others) of where they were and what they saw – and with the hope that their people on the ground were impartial and felt free to report – can we hope to have a more comprehensive assessment. Our researcher was also clear about the difficulties of reporting in Loya Paktia in this particular election: “No-one was independent,” he said. In a region with such apparently high support for one candidate, it is not surprising that reports of pro-Ghani fraud were few, although, in defence of this, one elder told AAN, “There was no need for fraud. Everyone wanted to vote for Ghani anyway.”

However, what does stand out from the region – and this does not seem to be fake – is the sheer delight and pride of people who (say they) have voted en mass. Several elders in Dandi Patan district, for example, told AAN just how happy people were with the elections. “We were competing with other districts to get our vote out,” said one. “The tribal unification and tarun played a major role in getting a higher turnout,” said another. In Sayed Karam district, Haji Rozi Khan, leader of the Tutakhel tribe said:

We voted and proved our pashtunwali. First, my wife went to vote and later on other women started to vote too. People are very happy. People were dancing with drums [dhools] on election day and this way they were motivating people to come out from their houses and to participate in the election. 

In Janikheil, people told AAN, drums and song also accompanied the vote, including a song famous for rallying fighters to join the Faqir of Ipi’s fight against the British in Waziristan (in the 1930-40s):

“De paktia de zhano saba lashkar di” (Tomorrow the lads of Paktia are forming an army to fight)

د پکتیا د ژنړوصبا لښکر دی

Actual statistics on turnout and fraud in Loya Paktia, as elsewhere, will come later. For now, at the very least, what can be said is that a new narrative is being told, of mass popular participation which people hope will drive political change.



(1) First round results were as follows:


Ghani 63.32%

Abdullah 5.37%

Others 31.31% (Sayyaf, Helal and Rassul all polled more than Abdullah)


Ghani 74.01

Abdullah 3.57

Others 22.42 (Helal and Rassul both polled more than Abdullah)


Ghani 65.21

Abdullah 10.55

Others 24.24 (Rassul polled more votes than Abdullah)

(2) A figure of 400,000 turnout for Khost and the comparison with the Central Statistics Office numbers for population (which would have meant more than 100 per cent voting) originated with the Abdullah team and was cited in a number of newspaper articles after he gave reporters an off-the-record briefing (which several shared with AAN) on 16 June. They included The Wall Street Journal, Arman-e Milli (Kabul) and others, here and here. When AAN asked Abdullah’s spokesman, Mujib al-Rahman Rahimi, about the source of the figure, he said: “The [turnout] figure was based on rumours… We need some time to figure out the exact figures.” And when asked again, he said, “We don’t know about the source, but rumours are flying all over the city.”  He continued to cite the 400,000 turnout figure for Khost during the interview, however. It is also worth pointing out that both teams had agreed to 420,000 ballots being sent to Khost (for more discussion on this, see this AAN dispatch).

(3) Abdullah’s spokesman told AAN they were also concerned that high turnout in Kandahar, some parts of Nangrahahr, Laghman and Logar were indicative of mass fraud.

(4) The term arbaki is indigenous to the south east where it describes a temporary defensive armed contingent raised by the tribe through a jirga. The word has, in recent years, been exported to other parts of Afghanistan as a pejorative term to describe the Afghan Local Police and unofficial, pro-government militias. The plural of tarun in Pashto would be tarununa.

(5) The Dzadran inhabit three districts each of Paktia, Paktika and Khost, forming the so-called Dzadran Arch.

(6) His name did not come up as ‘representing’ the region when he became a cabinet minister in 2002, for example, when one of the authors reporting on unhappiness in Loya Paktia, after the Emergency Loya Jirga in June of that year, over what people felt was their unjust exclusion from government.



2014 Elections Ashraf Ghani Loya Paktia tarun


Pakteen Ibrahimi

More from this author