What If the Feature Is the Bug?

Election monitoring, new power of social media and old power of structural power

Zeynep Tufekci

Turkey’s election night was a strange one. Many questions swirled around the process, coalescing into an uproar — and then a citizen-led, digital-media led recount.

But I write not in praise of new media, but to ask a question: What if the new power that social media gives ordinary citizens is also part of the new problem? What if the very abilities of social media —ad-hoc but powerful capacity for organizing and logistics— lead to shortcomings which then hobble movements, at least in the short to medium term?

What if the positive capacity we celebrate is actually the problem we are lamenting?

What if, as the technologists themselves say, the feature is the bug?

It’s not that these weren't tumultuous times in Turkey. They were. These were local elections only, but they came at the end of a tense period for the Turkish government. First the Gezi protests of last June, facilitated by social media, broke the image of internal harmony under Prime Minister Tayyip Erdogan’s Justice & Development Party (AKP). Then a corruption scandal challenged the AKP’s image as morally principled servants of the people’s interest. Finally, a string of audio recordings allegedly of the prime minister and his inner circle, implicating them in massive corruption and shocking abuses of power, were released on social media. In the days just before the election, the government responded first with a ban on Twitter, the key platform in which corruption allegations were discussed, and then, after the release of a national-security sensitive leak, of YouTube, which had hosted many of the ostensibly damning recordings.

Given this reality, the elections were going to be tense. But the night was strange in many more ways.

Anadolu Ajansi, Turkey’s official news agency, and Cihan, an opposition news organization, reported wildly different early results. Anadolu is headed by Erdogan’s ex-press advisor, who seems to have not noticed that he has a new job. Cihan also does not inspire confidence, as it is associated with a group that was the government’s closest ally until mere months ago. There was no other source of cumulative reporting on the totals, and no exit polls. Many polling companies have been caught with thumbs on the scale, anyway. Most television stations stuck by the official agency, reporting results that AKP was leading by 10-15% in hotly contested big cities, and ignoring Cihan’s reports showing a closer race. Others were reduced to announcing both, and shaking their heads.

And then there were the power failures in dozens of cities, which the minister of energy attributed to a stray cat loose in a power substation. The banned but still heavily populated Turkish Twitter had big fun with that one, as “kedi lobisi”—the cat lobby— was added to the “lobbies” that the prime minister blames for problems in Turkey, such as the “robot lobby,” which he alleges lives on Twitter.

Turkish dissident group @kollektifler maps out the “kedi lobisi” showing the places where unprecedented power failures occurred on election night.

In the midst of all this, hours before the results were fully in from either source, the prime minister gave a “balcony speech” announcing his victory. In hotly-contested Ankara, the opposition candidate had appeared to be leading by a slim margin when the official news agency stopped updating for many hours. Then, all at once, and much later than in previous elections, it suddenly gave a sizable lead to the AKP’s incumbent mayor.

Official news agency uncharacteristically stopped updating Ankara totals for hours before coming back up to give decisive lead to government candidate

In Turkey, all elections are done via paper ballots, which are counted with representatives of multiple parties present (if they are in attendance), and then entered into an online system available to the political party. This is a fairly solid system when it works and when it is properly monitored. In this election, growing distrust in institutions, coupled with the rise of cameras and social media, meant that a large number of people waited for the final tabulations in their polling stations and then took pictures of the official paper with the results signed by representatives. And people then went online to compare their photo of the tabulation with government database, made accessible by the opposition party. This gave people a chance to check for discrepancies. And they found them.

Quickly, pictures of discrepancies spread, as people retweeted ballot tabulation after ballot tabulation unmatched by the online result. With nerves already frayed among the Turkish public (especially dissidents), reports of voting irregularities spread like wildfire on social media.

A Twitter user compares the official paper tabulation with the online result, which differs.

It’s not that these new digital platforms aren't empowering. They are. This was amply demonstrated by what happened next: a crowd-sourced recount.

A small scrappy citizen journalism outlet, @140journos, had already spent the day archiving and patiently tweeting ballot box tabulations, one after the other, hash-tagged by city, district and ballot box number. Similar initiatives popped up as citizens gathered and tabulated their pictures of the paper ballots counts.

One might expect the main opposition party, the CHP, to hold a copy of each tabulation. However, in line with the complaints I heard from hundreds of the Gezi protesters that I interviewed, the opposition’s incompetence was part of the problem, and many reports on social media indicated that they had not sent representatives to all polling stations. Lacking a full database of the paper tabulations weakened their ability to challenge any manipulation of the vote.

The plucky CHP candidate in Ankara, who had recently switched parties as part of a semi-official coalition, announced that he’d be contesting the elections—and he was somewhat atypical among CHP candidates and no doubt partly because he had not been steeped in CHP’s culture of resignation to loss. So did a few others in Istanbul, Antalya and few other closely contested cities.

Many citizens were upset about the process, and even if they believed their candidate may well have lost, they wanted the numbers to be corrected. They wanted a recount even if it wasn’t going to make a difference. It was the principle and the ideal as much as the result. The rule of law, and confidence in institutions, may seem boring subjects in 6th grade social science books, but when they diminished, we no longer have the luxury of rolling our eyes at them.

It all sounds like classic story of social media to the rescue right? It does, but it won’t end this way. Indeed, watching social media enable citizen power can hardly fail to impress. Using the hashtag #tutanakNO, Ankara residents gathered the missing ballot box tabulations, one by one, from citizens who had taken pictures, from representatives of other parties who agreed to cooperate, and from online searches seeking anyone who had uploaded a copy. In smaller districts, a similar pattern followed. At some points, the search for the few remaining missing ballot box tabulations was so intense that the related hashtags (#tutanakno) were trending on Twitter.

Slowly, but surely, the ballots were gathered. The Ankara candidate appealed (and lost most of the rounds) and the legal battle isn’t yet over. Meanwhile, @140journos also has developed a multi-layer map of citizen-reported tabulations and official results, which is now being combed for patterns of irregularities.

It wasn’t the first time Twitter had been used so effectively in an election monitoring effort. I had previously documented a similar effort in Egyptian presidential elections when what sure looked like a regime considering whether it could get away with stealing the election was blocked from doing so by crowd-sourced (but organized by a political party, the Muslim Brotherhood) election monitoring. (Egypt did not end well either—a military coup a year after that election followed by a blood bath effectively ended the democracy movement in that country—for the moment). Although interrupted by military coups, Turkey has been having elections for more than 60 years. The process has many safeguards, at least on paper, if properly implemented and monitored. However, Turkey’s political polarization and collapse of checks and balances has certainly eroded confidence in the process, and created an environment of distrust.

So far, this may sound like another new media success story. Regardless of whether the contestation is successful, new media capabilities empower citizens to remain engaged in the electoral process, and provide a check on increasingly less accountable institutions.

But… remember that little note about the incompetence of the opposition party? And their failure to have a copy of each ballot box tabulation? And to have organized an aggressive election monitoring system as is possible under the law?

The one that keeps losing election after election to the well-organized, competent electoral organization of the AKP?

As you can imagine, election monitoring is not their only point of failure. By almost all accounts, CHP’s organizational weaknesses is part of the problem of an increasingly unbalanced political system in Turkey, dominated by a single party. I don’t say this to discount any criticisms of the governing party, which I’ve reported upon many times. The question remains: from where does electoral change come?

The impressive ad hoc capacity that can be focused via digital tools– and with the aid of trending topics and other social media affordances — allows citizens to carry out actions for which they would previously have needed to build powerful and robust social institutions. Such institutions could then do other things besides the specific actions of the moment for which the citizen-capacity came together.

The governing party in Turkey has done that, it hasn’t just built ad hoc coalitions, it has built an electoral machine and placed itself at all levers of power and governance. Ironically, this is exactly what led to the current political crisis—existing political parties remain incompetent, under-organized and under-powered. The impressive, large-scale and energetic citizen movement which organized mainly through social media and expressed themselves through street protests, online campaigns and neighborhood forums has not changed that. AKP has just weathered its sixth electoral challenge in more than a decade, and again emerged as dominant.

A hint, hint moment if there ever was one.

Citizens may well be able to find all the missing ballots, enter them into databases, collate the results, run sophisticated statistical analyses that raise eyebrows, and may even overturn an election or two. All those are major wins, with major impacts both short and long term.

But who, if anyone, is going to fix the failing institutions that continue to have an enormous impact on the everyday lives of everyone, regardless of their opinions about how corrupt, unpopular or even illegitimate these institutions are? Especially given how tedious, tiring, unexciting organization-building can be, compared with humorous, energized, adrenaline-filled efforts such as protests and occupations?

New media certainly enables even the coalitions of ordinary citizens to realize impressive logistics. I am amazed at the energy and creativity I’ve witnessed in country after country, as citizens organize everything from election monitoring to disaster aid. They do not lack in numbers, talent or creativity, and they have an impressive array of new tools. It’s not that the protesters are shying away from sacrifice or hard work, and it’s not that they are preferring online over offline—the Gezi Park protest thoroughly mixed online and offline, as had Occupy Wall Street. And yet, these new movements keep failing to mount a successful — or even credible — electoral challenge, and they have not yet found a way to impact institutions which hold great sway and influence over our lives.

What if the feature is the bug?

The Message

A Pandaemonium Revolver Collection. Season 2 stars @anildash @alanalevinson @ftrain @hipstercrite @itsthebrandi @jamielaurenkeiles @vijithassar @yungrama @zeynep. Season 1 available on DVD shortly.

Thanks to Robin Sloan, Craig Mod, and Quinn Norton.

    Zeynep Tufekci

    Written by

    Thinking about our tools, ourselves. Assistant prof at UNC iSchool. Princeton CITP fellow, Harvard Berkman faculty associate, Sociology.

    The Message

    A Pandaemonium Revolver Collection. Season 2 stars @anildash @alanalevinson @ftrain @hipstercrite @itsthebrandi @jamielaurenkeiles @vijithassar @yungrama @zeynep. Season 1 available on DVD shortly.