Tag Archives: Abdullah Abdullah

Lessons learned: a very brief post-mortem of the 2014 elections

Inside the voting booth in Afghanistan

Inside the voting booth in Afghanistan

Regular readers of this blog realize that I almost completely refrained from election-related analysis here. This was partly because it is very difficult to predict Afghan elections, especially a field as wide open as this one. But after the announcement of the full preliminary results of the first round of the presidential election, we have enough information to do a post-mortem of some of the assumptions and the facts held to be common knowledge during the campaign period.

We can also add to the things we learned and list the new trends we observed this cycle. We still don’t have the full data about the election – turnout and voting behavior by various demographic cross-sections, for example – but we have enough to draw some lessons. Some of these lessons are classics in electoral politics, but in Afghanistan’s short-lived exercise with democracy, we are just learning them.

This list is not exhaustive, of course, so feel free to add your thoughts in the comments.

  • Marital status and the spouse’s religion don’t matter. Certain media outlets predicted that voters would be turned off by Zalmai Rassoul’s bachelor status or by Ashraf Ghani’s Christian wife of Lebanese origin. Pre-election polls never revealed these to be voters’ top priorities, which means they were largely made into campaign issues by the media and the punditocracy. Post-election results show no demonstrable impact of these factors on the candidates’ performance – Ghani and Rassoul came in second and third, respectively.
  • Money can buy votes, but only for the right candidate. IEC’s records show that the top-place finisher Abdullah Abdullah spent AFN10.02 million, which was almost double the second-place finisher Ashraf Ghani’s AFN5.6 million. This bought Abdullah 13 points…which is what we know to be true from Obama’s epic $760 million to McCain’s $358 million in 2008. But Zalmai Rassoul found that his AFN9.8 million didn’t win him as many votes as he would have liked, putting him 20 points behind Ghani – which is what Megan Whitman ($177 million) learned running against Jerry Brown ($36 million) for California governor in 2010. Of course, there’s no comparison between the campaign finance regulatory enforcement between the two countries and as the IEC admits a lot went unreported, partly because of an unrealistically low campaign spending cap. But in the 2009 election when there was no cap, Karzai spent a whopping AFN98 million, besting Abdullah’s AFN23 million.* The results speak for themselves.
  • Opinion polls have predictive value. The amount of doubt raised about opinion polls was almost equal to the coverage they received, which was ad nauseam. The campaigns that performed badly in the polls said they were inaccurate, the President banned opinion polling alleging they unduly influenced the electorate, and analysts said polls didn’t account for the expected widespread fraud. But most polls accurately predicted the order of the top three finishers – though the polls lacked the hair-width accuracy people have come to expect from polls in established democracies. The final opinion poll, released just days before election day, showed a difference of less than three percentage points between Abdullah and Ghani, but actual results show a whopping 13-point difference. Still, polls have a margin to improve their accuracy and a lot of potential for use by future campaigns, political parties and the citizenry.
Voters on election day

Voters on election day

  • We know how heavy the heavy hitters can really hit. Sayyaf and Ismail Khan, the Karzai family and Hizb-e-Islami all performed less well than expected. The Sayyaf-Ismail Khan duo only got 7% of the votes, with Khan failing to win more than 15% in any of the southwest provinces  expected to turn out for him. Karzai’s supposedly favorite candidate only won one province, and the Hizb-e-Islami candidate won none. Dostum only delivered Jawzjan for Ghani. On the other hand, Mohaqeq bagged the Hazara votes for Abdullah. He delivered Bamiyan and Daikundi and was behind Abdullah’s strong showing in Ghazni, Kabul, Ghor, Samangan, Wardak and Balkh.
  • Electoral politics have setbacks, and setbacks are recoverable. Lest we mistake Sayyaf’s 7% votes for his real influence, we should remember that Abdullah was the runner-up and Ghani placed fourth in 2009. The turn in Abdullah’s and Ghani’s fortunes shows that the electorate is open to persuasion, or somewhat more optimistically, the electoral institutions are capable of reflecting the will of the voters, or even more optimistically, that Afghanistan’s democracy actually kind of works despite the criticism.
  • Money mattered, but did Big Money? Abdullah outspent Ghani and ultimately outperformed him. One would imagine that with a campaign spending cap as low as AFN10 million, there wouldn’t be much space for giant donors, and this is borne out in the campaign contribution reports. My quick glance through all candidates’ reported incoming contributions showed that corporate donations were roughly around the half-million range (see example). In fact, the only million afghani-plus contribution to a campaign was AFN1.1 million by Helal to his own campaign (Ghani spent AFN2.3 million of his own money, but it wasn’t reported as a contribution to the campaign). All of this raises the question: where were the Fortune 500 of the Afghan corporate sector that bankrolled Karzai’s 2009 campaign and invested in Abdullah’s? It is not possible that all of them chose to sit on the sidelines and not contribute this year. A plausible answer is that the unrealistically low campaign spending cap just meant that 1. campaigns under-reported their spending to keep them within margins, so they didn’t report major contributions, and 2. the lion’s share of campaign-related spending didn’t occur through campaigns themselves. For example, the IEC admits that several candidates didn’t include expenses incurred in staging huge rallies in multiple provinces because those rallies were paid for directly by supporters and not through the campaign.** This unregulated and potentially unlimited money sloshing about the campaign trail is reminiscent of super PACs in American politics, with the major difference being that the Afghan super PACs are not formally recognized and can coordinate with campaigns on spending and messaging.
  • The “battleground” provinces didn’t decide the election. There were only four provinces where the top two candidates were within a five-point margin. They are Farah (Abdullah 35, Ghani 31), Nimroz (Ghani 30, Rassoul 33), Uruzgan (Ghani 27, Abdullah 23) and Helmand (Ghani 32, Rassoul 27). The vote in other provinces was so decisively in favor of one candidate or the other – exclusively between Ghani and Abdullah, except Kandahar – that if we take these neck-and-neck races out, the overall results would remain almost exactly the same.
  • The campaigning wasn’t ethnic but the voting was. None of the major candidates openly presented themselves as representatives of their respective ethnic base on national media. This was good in a deeply divided country. But it didn’t matter how they presented themselves, it mattered how voters saw them – which is why Ghani and Rassoul won in Pashtun provinces, Abdullah won in Tajik provinces, Dostum delivered Jawzjan and Mohaqeq delivered Bamiyan and Daikundi. If we take out Jawzjan, the vote was very clearly along north-south lines, just as it was in 2009.

—-

References:

* Data pulled from table 6 of this IFES paper on the 2009 elections.

** The IEC website has a pretty useful report, in Dari, about the reporting of campaign contribution and spending, and the challenges and recommendations.

Note: Any views expressed in this post are all strictly personal and do not represent the views of my employers, current or former.

The electorate is confused and the field is wide open

Image

The election season’s first public opinion poll results were released yesterday by TOLO News/ATR, with the headline that ‘initial election polling reveals three favorites.’ Those favorites are Dr. Abdullah, Ashraf Ghani and Qayum Karzai, respectively.  This is fascinating insight for Afghanistan, because popular punditry also had Zalmai Rasoul, Sayyaf, Gen. Wardak and Gul Agha Sherzai as among the top favorites, but they barely even registered (Sayyaf, for example, was the highest at 3.4%).

But the headline buried the lead, because 50% of the respondents to the survey said they do not have any favorites or have not decided whom to vote for, suggesting that the field is wide open. This means that even Gen. Wardak’s 0.8% could theoretically go up to 50.8%, making him president.

There are many reasons to take these results with a pinch of salt,* but the poll results show that the electorate is open to being wooed, that ethnicity will matter less and platforms will matter more. Unlike previous elections when ethnic blocks were represented by one premier ticket, this time those ethnic blocks are fractured — the Tajiks, Hazaras and Pashtuns all have more than one prominent person running.

In the absence of an optimal ethnic ticket, the so-called vote banks cannot vote based on their habitual reflexes but have to choose between several tickets. If voters were shoppers, and their trusted brand of toothpaste is on longer available, they are now searching for a good new alternative. A significant portion of the electorate is open to being influenced, and the ticket that most appeals to the voters and reflects what they want has a greater chance of winning.

This does not mean that ethnicity does not matter anymore, but it does mean that it matters less than in previous elections. A shrewd ticket would capitalize on that and instead of focusing on getting the best combination of warlords on the ticket, it would work on developing a platform and an effective campaign message.

The question then becomes about what the voters want. We have a fairly good idea of what they want for the country, which is security, economic development, reduced corruption, electricity and education, justice, etc. But there is a dearth of hard data on what the electorate wants in a candidate. We know that Afghans have always wanted peace but they have reliably voted for warlords; they have always wanted economic development but they have been swayed by candidates with religious credentials; they have wanted education but they have voted for conservative mujahids.

But people with technocratic reputations – Ashraf Ghani, Rasoul, Qayum Karzai, Abdullah and Wardak – dominate this field, and ethnic tickets are multipolar. The voters have to develop new calculations about what they want in a vote-worthy ticket, what factors matter more than others, and what attributes could replace or proxy for traditional ones such as ethnicity, religious and jihadi credentials, etc.

That could perhaps be the subject of the next TOLO News/ATR poll.

—-

* Reasons to take the poll with a pinch of salt:

  • Opinion polls are untested in Afghanistan, so we don’t know how well polls reflect actual voting patterns
  • It is still too early: tickets will change, alliances will shift, people will get disqualified
  • Opinion polls may not reflect who will vote and who will get to vote –  the Ministry of Interior says more than half of the polling stations are in insecure areas, and the Taliban say they will target anyone who becomes a candidate or votes
  • Opinion polls cannot factor in the amount of fraud/irregularity that is likely to occur
  • Two-thirds of respondents are in urban areas, which does not reflect Afghanistan’s actual population spread
  • Et cetera.

Note: I have expressed a lot of views in this post. The views are all strictly personal and do not represent the views of my employers, current or former.