By Andrew Garfield
In the final weeks before the second round of the Afghan presidential elections, my company undertook two polls of likely voters to determine which of the two candidates was in the lead – Dr. Abdullah Abdullah or Dr. Ashraf Ghani. Our primary motive for doing this was to provide transparency in the election process and to try and deter fraud at the polling stations and in the counting process.
Many in the Western media and policy community assumed that Abdullah was a “shoo-in” for the second round because the official results from the first ballot had Abdullah at 45 percent and Ghani at 31.5 percent and because many of the candidates eliminated in the first round had endorsed Abdullah, the assumption being that voters would follow their chosen candidates’ advice. I had seen no evidence, however, that Pashtun and Uzbek voters would follow such endorsements and vote for Abdullah. In a straight two-way contest, I believed that the demographic numbers still favored Ghani over Abdullah.
The two polls we undertook – one by telephone and the other face-to-face – had Ghani gaining significant momentum since the first round with a lead amongst likely voters of 48-49 percent to Abdullah’s 42-45 percent. These results showed that Ghani had secured the support of most Pashtun and Uzbeks voters and that he had retained the limited support he had secured with the other ethnic groups. Given that Abdullah was still around the 45 percent level, this indicated that he had just about maxed out his support base and could only increase his share of the vote if he could secure additional Pashtun support, which I considered unlikely
My team then carried out two exit polls – again one face-to-face conducted at polling stations, and the other a telephone survey. We asked Afghan voters, who confirmed that they had just voted, whom they had voted for, and a majority (53/54 percent) said they had chosen Ghani over Abdullah (47/46 percent). These results were entirely consistent with the polling since March, showing Abdullah stuck at around 45 percent and with Ghani winning by 4-6 points in a reasonably free and transparent election.
POST-SECOND ROUND CONTROVERSY
It seems likely that significant election fraud has taken place during and after the second round of these elections. There was certainly a significant level of fraud in the first round, with hundreds of examples of the more serious types of fraud including ballot box stuffing – “stuffing the sheep.” Very soon after the end of second round voting, Abdullah and his supporters began making very public accusations about rampant fraud and within days he had withdrawn from the counting process. According to the BBC, Abdullah said his decision to stop co-operating with the election authorities had not been intended to disrupt the process, but to prevent a fraudulent election result and to protect people’s votes. The BBC has also noted that Ghani has also made accusations of fraud but it has been Abdullah’s accusations that have received the most coverage in Western media, perhaps because most Western journalists were persuaded – wrongly, as I have shown – that Abdullah would easily win the second round.
Another reason for the differential treatment of the candidates’ accusations of fraud is the pervasive myth that Abdullah was robbed of the presidency 5 years ago in the 2009 election. Indeed, a BBC report concluded: “Both presidential contenders have lodged complaints about the conduct of the elections, and for Mr. Abdullah – who felt he was robbed of the presidency back in 2009 – there is a sense that history is repeating itself, our correspondent
A LOOK BACK AT THE 2009 ELECTION
In the summer of 2009, my company was commissioned by the U.S. Department of State to run two polls prior to the 2009 Afghan Presidential Election. Our first poll was fielded between 8th and 17th of July 2009, with the second running from July 15th to July 23rd. The first wave was conducted amongst 3,556 Afghans age 18+ and our poll had an overall margin of error of +1.64% in 19 out of 20 cases. Data shown among the 2,823 registered voters interviewed had a margin of error of +1.84% in 19 out of 20 cases. The surveys were conducted in person across all provinces, using a multi-staged stratified sampling procedure in all of Afghanistan’s provinces.
These slides clearly show President Karzai well ahead of his nearest rival Abdullah with all of those interviewed and with likely voters. With both groups Karzai was ahead of Abdullah by around 15 percent. Our team therefore correctly assessed that in a fair and transparent election, President Karzai would fail to secure the 50.1 percent majority he would need to win the election outright and that there would have to be a runoff with Abdullah. However, we had Abdullah well behind in that first vote probably by as much as 15 points. The final official results of the election had President Karzai at around 49 percent and Abdullah at around 31 percent, although as we know for some time prior to the official results finally being released, President Karzai’s supporters claimed that he had won outright – a result that could only have been achieved with systematic nationwide election fraud.
As quickly became apparent, there was certainly ample evidence of widespread fraud in President Karzai’s favor and it was only under heavy U.S. pressure that the President agreed to a runoff. However Abdullah then refused to participate in the second round saying “I will not participate in the November 7 election,” because his demands for changes in the electoral commission had not been met, and a “transparent election is not possible.” Abdullah also said the Afghan people should not accept results of an election from the current election commission, and stated that Karzai’s government had not been legitimate since its mandate expired in May 2009. However, it is my contention that this was a shrewd political move on the part of Abdullah designed to create a narrative of the wronged and robbed candidate, when in reality he knew he could not win the second round. Following the advice of Sun Tzu, he did not fight the battle he could not win. And in doing so the myth of his being robbed was born.
Using the evidence of our two polls, we projected as early as August 2009 that President Karzai would secure an overwhelming victory in a second round, defeating Abdullah by approximately 60 percent to 40 percent of the vote. We came to that conclusion for two primary reasons – ethnic support and the overall popularity gap between the two leading candidates.
This ethnic breakdown of support for the main candidates in our first poll showed that President Karzai had a huge lead amongst Pashtun voters and that he had secured a sizable minority of the Tajik vote, which was Abdullah’s base. Karzai was also slightly more popular with Hazara and significantly more popular with Uzbeks. Ergo in a second round he was far more likely to pick up votes from the eliminated candidates from these key minority ethnic groups. We also felt that in all probability he would secure most of the Pashtun vote, once Pashtun candidates like Ashraf Ghani had been eliminated.
Data from the first poll also showed that President Karzai was far more popular with likely voters than Abdullah. This is another strong indicator that in a second round a majority of voters would choose the President over his rival. Indeed President Karzai remained popular across all ethnic groups, which bode well for an easy victory in a second round.
Glevum was not able to publish the findings of our second poll, which was completed after a media ban imposed by the Department of State in the final weeks of the election. However, this poll also showed Karzai with a commanding lead over Abdullah (33 percent to 21 percent) and a similar split of the ethnic vote.
Both polls demonstrate that while President Karzai could not win the first round outright he was well on course to win the second round handsomely – a narrative that is now overshadowed by the perception of Abdullah as a candidate that was robbed. Notwithstanding the accusations of fraud, I still think that voters would have favored Karzai over Abdullah had the second round been held. Indeed, Karzai retained positive popularity numbers even after the second round was cancelled.
There is no doubt that the second round of the 2014 President elections has been marred by fraud, as was the first round. However, at this point the perception in the Western media at least is that Abdullah is the candidate that has suffered the most and this is reinforced by his own narrative that he was robbed in 2009, which seems to have become accepted wisdom even with the BBC. I contend that this narrative is a total myth created by Abdullah himself and that the evidence simply does not support that perception. President Karzai would have won a second round in 2009 by a wide margin, as Abdullah had maxed out his core vote and could not erode the core vote of the President.
Notwithstanding the accusations of fraud from both sides this time around, the four Glevum polls undertaken just before and during the second election also indicate that he has maxed out his vote this time, too, peaking at around 45 percent and Ghani, by dint of simply turning out his larger core vote, has won the 2014 election. Whatever the final result of the election, the future prosperity of Afghanistan is not helped at such a sensitive time by Western media and officials validating Abdullah’s assertions about the fraud at this election by suggesting he was robbed in 2009. That simply was never the case.
About the author:
Andrew Garfield is a Senior Fellow in FPRI’s Program on National Security. A U.S citizen since 2010, served as a British military then senior civilian intelligence officer, finishing his U.K. government service as a policy advisor in the UK Ministry of Defense (MOD). Since emigrating to the U.S. in 2004, he has worked exclusively for US clients including the U.S. Department of Defense, U.S. Army, and more recently the Department of State. In late 2006, Mr. Garfield founded Glevum Associates LLC, a company that specializes in conducting Face-to-Face Research and Analysis in conflict and post-conflict societies.
This article was published at FPRI and may be accessed here.