A fire burns as a left-wing militant throws a molotov cocktail toward a water cannon during a demonstration in Istanbul’s Gazi district in October 12. (Ozan Kose/AFP/Getty Images)

In Turkey’s June 2015 election, the ruling Justice and Development Party (AKP) lost its majority in the Turkish Grand National Assembly, and the pro-Kurdish People’s Democratic Party (HDP) won seats as an independent political party for the first time. This victory initially generated enormous hope for change in Turkish politics. Instead, the AKP refused to form a coalition government with other political parties, forcing a snap election in November, which the AKP won decisively. Since then, the broad crackdown on the media, civil society and academia has become a major threat to Turkish democracy itself.

The Kurdistan Workers Party (PKK) relaunched its insurgency in between the two elections. The AKP’s refusal to bring Kurdish demands into the Grand Assembly for legislation convinced the PKK that the AKP government was not seriously interested in peace negotiations. The HDP’s new political direction also failed to match up with insurgent objectives. In the wake of growing insurgent activity, the Turkish security forces retaliated with extreme force, launching counterinsurgency operations that have led to hundreds of civilian deaths and displaced thousands. Anti-HDP discourse has resurfaced amid the violence.

Many Turkey analysts claimed that the resurgence of insurgent violence propelled the AKP to victory. After all, between the June and November elections, the AKP increased its vote share by 9 percent while support for the HDP eroded across the country. But, as our new research demonstrates, this isn’t exactly right. The HDP did not directly lose support as a consequence of the violence. Meanwhile, the incumbent AKP enjoyed the growing support of civilians who didn’t want to see violence return to their home towns. Put differently, while civilians’ exposure to sustained violence benefited the pro-insurgent party at the ballot box, the unrealized threats of violence by the PKK consolidated incumbent loyalties.

A new dataset, available from authors, allows us to test the effect of this violence on voting patterns. It includes all insurgent attacks between the two elections in 113 districts of 13 provinces in southeastern Anatolia, which were previously under emergency rule (1987 to 2002). These districts experienced 397 attacks, which correspond to 86 percent of all incidents in the country. The region’s residents are overwhelmingly Kurdish, co-ethnics of the insurgents, but have divided loyalties between pro-insurgent and incumbent parties. We compiled the data set from the online version of the leading national newspaper in Turkey, Hürriyet. Election results for urban areas are drawn from the websites of Turkish Institute of Statistics and Supreme Electoral Board (YSK).

Our findings suggest that sustained violence did shape civilians’ political preferences in the November election. Without taking violence into account, the AKP victory seems straightforward: The incumbent increased its urban vote share by 9.4 percent on average, whereas its pro-insurgent rival lost almost 7 percent of its support from June to November election. Inserting violence into the equation by looking at the precise location of attacks and their targets tells a different story. As civilians’ exposure to sustained violence grows, the AKP gains disappear and the HDP makes up most of its losses. The best illustration of this trend comes from districts that received urban attacks with security casualties. In these settings, the change in electoral support for both parties were extremely close, falling within the 2 percent range.


The lack of violence helped the AKP to increase its vote share. Peaceful districts, most of which were already AKP strongholds, delivered major dividends for the incumbent party. The AKP received 5 percent more votes (12.8 percent in total) from districts that received no attacks compared to those that did. The other contributing factor was swing voters in settings characterized by fluid political loyalties. Contested districts, where the AKP or HDP received 25 to 49 percent of the vote, defected from the HDP in large numbers and flocked to the incumbent.


The HDP losses were 3.3 percent lower in districts that experienced multiple PKK attacks compared to ones that received none. This figure is striking given the fact that the overall decrease in HDP vote was only 6.9 percent from the previous election. Furthermore, violence became an important difference maker in HDP strongholds, where the pro-insurgent party received at least 50 percent of the vote in June election. Districts experiencing multiple attacks recorded the fewest losses for the HDP, demonstrating the strong and positive impact of violence on political preferences.

The political effects of violence were mediated through two mechanisms. Civilians’ access to information about violent events was instrumental in delivering major benefits to the pro-insurgent HDP. In particular, urban voters who lived in close-knit neighborhoods, had cross-cutting ties, and followed media outlets closely reacted strongly and positively to violence in their districts. Meanwhile, the increase in turnout worked to the benefit of the incumbent. The AKP’s rhetoric of insurgent fear further mobilized voters in incumbent strongholds despite the fact that these districts rarely experienced violence between the two elections. In sum, PKK violence shaped the November election in ways favorable to the AKP without necessarily hurting the HDP at the ballot box.

Aysegul Aydin is an associate professor of political science at the University of Colorado at Boulder. Cem Emrence is a sociologist and author of “Remapping the Ottoman Middle East” (I.B. Tauris, 2012). Aydin and Emrence are the co-authors of “Zones of Rebellion: Kurdish Insurgents and the Turkish State” (Cornell University Press, 2015).