On Thursday, August 20, Afghanistan had its second-ever presidential election. It’s a big deal for several reasons, not least of which is that the entire U.S. mission there is predicated on the attempt to create a self-sustaining government, one capable of holding meaningful elections. The Afghan-centric reasons are even more important, however. Any old country can hold elections once— but they start becoming genuinely meaningful when the country can hold them a second time.
But does the August 20 election really count as one? The Afghan government estimated it fell 11 percent short of the minimum number of polling stations necessary to call the vote a success. Depending on the source one consults, between twenty-six and seventy-eight people died in 135 attacks on polling stations across the country. Even ignoring the pre-election reports of opposition harassment and the police murdering likely voters, the election didn’t meet the basic minimum requirements of a free, fair, representative event.
That doesn’t necessarily make Thursday’s election a failure, however. To borrow a phrase from Donald Rumsfeld, you go to Afghanistan with the election you have, not the one you want. So how can we assess what actually happened there? Western journalists rarely travel outside of Kabul without a military escort; Afghan journalists faced government threats if they reported acts of violence during the election. How do you tell what’s going on?
For more and more people who try to follow these things, the answer is Twitter. Most famously during the riots in Iran earlier this year, Twitter has been an efficient workaround to state-enforced media bans, and a wonderfully democratizing influence on how news is reported in difficult-to-reach areas.
On a normal workday, I get my information from two basic sources: my RSS feed, and my Twitter feed. I’m much more likely to pore through Twitter than my news RSS feeds for one simple reason: the people whose feeds I follow are mostly other analysts and researchers, and they act as an information filter. Pre-filtered information, at least from people you trust, is much more efficient than doing the filtering yourself.
During the election on Thursday, though, this efficient filtering system broke down. While the AP wrote a breathless story about Afghans using Twitter to get the word out, there were really only three sources: Alex Strick van Linschoten (@strickvl on Twitter), a researcher and freelance journalist living in Kandahar; the Pakhwok Afghan News agency (@Pajhwok), a truly remarkable independent news organization operating across Afghanistan; and maybe one or two others (like Felix Kuehn, one of van Linschoten’s colleagues).
As a result, my Twitter feeds during Election Day had an incredible noise-to-signal ratio—by an informal calculation, 90 percent of the tweets filling my computer screen were just retweets of something I had already seen (that is, other users were reposting, with credit, someone else’s tweets) from those same three sources. It got so bad that, by mid-morning, I had to stop reading Twitter—everything became a blur of identical tweets from the same place (mostly Pajhwok, which did excellent work). I thought I knew how to “tune” Twitter so that I could hyper-efficiently barrel through inhuman amounts of information. What I learned on Thursday was that Twitter only becomes a game-changing tool when there are a lot of people contributing to it.
That’s not to say Twitter is an invalid tool for research, or for monitoring elections. But it is just that, a tool, and it still has a long way to go for it to be a reliable source of useful news and information. If there were lots of people contributing tweets from around the country, especially in a dangerous area like Helmand, Twitter could have provided a revolutionary way of gauging the election’s prospects. Similarly, even in safe areas there weren’t many original, local tweeters—almost all of the Afghan election tweets came from Pajhwok, not from individuals.