Türkiye’s PKK Conflict: A Visual Explainer

In July 2015, a two-and-a-half year long ceasefire broke down, Türkiye's fight against militants of the Kurdistan Workers’ Party (PKK) – recognised as a terrorist organisation by Türkiye, the U.S. and the European Union – entered one of its deadliest chapters in nearly four decades.

Since that date, the conflict has progressed through several phases. Between roughly 2015-2017 the violence devastated communities in some urban centres of Türkiye’s majority-Kurdish south east and – at times – struck into the heart of the country’s largest metropolitan centres. From 2017 onward, fighting moved into rural areas of Türkiye’s south east. As the Turkish military pushed more militants out of Türkiye, by 2019 fighting shifted to northern Iraq and northern Syria.

The International Crisis Group has assembled a database of fatalities caused by this conflict and analysed data on violent incidents recorded by the Armed Conflict Location and Event Data Project (ACLED). The data visualised on this platform is based on information available in open sources, including reports from Turkish language media, the Turkish military, local Kurdish rights groups, and PKK and its affiliates' websites.

According to Crisis Group’s fatality tally, last updated on 15 May 2024:

The fatality rate in Türkiye’s PKK conflict peaked in the winter of 2015-2016. At this time, the conflict was concentrated in a number of majority-Kurdish urban districts in Türkiye’s south east. In these districts, PKK-linked youth militias had erected barricades and trenches to claim control of territory. Turkish security forces reestablished control in these urban centres around June 2016. Since then, the rate of fatalities has gradually been decreasing.

The state security force-to-PKK militant fatality ratio is a good indicator of the changing power balance in the battlefield. The rate of PKK militants killed per one state security force member has been fluctuating since July 2015.

The share of Syria-born militants killed in the conflict increased as the number of Türkiye-born militant fatalities dropped since the resumption of hostilities in July 2015.

Since July 2015, militant attacks and clashes with the PKK have killed at least 1,481 security force members, including 1047 soldiers, 304 police officers, and 130 village guards (paramilitaries comprised by ethnic Kurds armed and funded by the Turkish state).

Improvised explosive devices (IEDs) have killed 568 of all state security force members since the end of the ceasefire. An additional 46 were killed by urban snipers, while 115 were killed in attacks involving rocket propelled grenades. All other deaths occurred during militant attacks or clashes which were not described in detail in media reports or official military announcements.

The average age of members of the state security forces killed since July 2015 is 28.6.

At least 628 civilians, and 226 “individuals of unknown affiliation” were killed since July 2015.

The names of an additional 40 individuals who were allegedly killed during the urban phase of the conflict have been released by local Kurdish rights groups, but cannot be confirmed through Crisis Group's open-source methodology. These fatalities are recorded as “unconfirmed” deaths by Crisis Group.

The death of 4,639 PKK militants have been confirmed by Crisis Group since July 2015. Confirmation of militant fatalities is complicated because the PKK often reports the deaths of its members weeks, months, or even years after they are killed in fighting.

At least 4,094 members of the People's Defence Force (HPG), the PKK's primary armed wing, have been killed since July 2015. 538 militants of the Civil Protection Units (YPS), a loose network of PKK urban youth militias, and 7 members of the Kurdistan Freedom Falcons (TAK), a PKK affiliate (considered an “offshoot with no operational ties to the PKK” by some researchers) responsible for attacks in western Turkey, were killed over the same time period.

Around 25 per cent of PKK militants (including all of the above groups) killed since July 2015 are female.

Visualisation of data related to deadly conflict helps Crisis Group analysts understand its complexity. We believe it can help other advocates, too, as well as policymakers and others directly involved in mitigating the impact of conflict or resolving it.

Crisis Group has worked to identify all fatality claims made in the PKK conflict in Turkey since 2011, when it began keeping an open-source fatality database. To ensure accuracy amid a rapid escalation of violence, Crisis Group has also worked to identify the names of all fatalities since July 2015. Only fatalities that can be named are included in the post-July 2015 tally. Names are confirmed through searches of news reports or social media postings for basic biographical data, unique photos, funeral reports and by means of media interviews with the victim's relatives or friends.

In addition to its own fatality tally, Crisis Group has used data assembled by the Armed Conflict Location and Event Data Project (ACLED) to visualise trends in violent events in this conflict in Türkiye, northern Iraq and northern Syria. Crisis Group published an upgraded version of this visual explainer with visual components based on the ACLED data in June 2023.

Crisis Group values feedback and incorporates it into updates to this and other data visualisations. Please send information and inquiries to [email protected]

We do not share the complete source dataset with third parties.

Read more about data collection and verification methods for:

1. State Security Force Members

Turkey's state security force members are composed of soldiers, police officers and village guards. Crisis Group uses press reports of funerals and official announcements by the Turkish Armed Forces to verify fatalities in this category. In addition to individuals killed in direct clashes with PKK militants, security force deaths include security officials targeted as such, but slain while off duty. Crisis Group does not count traffic accidents, suicides or other non-conflict fatalities unless they are explicitly linked to security operations.

2. Civilians and Individuals of Unknown Affiliation

Civilian deaths are checked by Crisis Group via reports in Turkish-language press, as well as by reports from local human rights groups, including the Human Rights Foundation of Turkey (TIHV) which has a strong track record as dependable sources of named fatalities.

Those fatality claims are subsequently confirmed by Crisis Group, which considers an individual a civilian if they can be affirmed as non-combatants by press reports or other sources, if their age or other factors made their participation in armed hostilities unlikely, or/and if they died outside an active conflict zone.

The line between civilian and combatant has been blurred by the emergence of plainclothes PKK youth militias in urban areas throughout the south east, in particular between December 2015-June 2016. As a result, some individuals could not be positively identified as civilians, though they also remain unclaimed by the PKK or its affiliates. Crisis Group classifies  individuals between the ages of 16 and 35 killed in areas of armed clashes and not identifiable as either civilians or combatants as “individuals of unknown affiliation”.

3. PKK Militants

Militants are identified through reports from the Turkish language press or from media outlets directly affiliated with the People's Defense Force (HPG) -- the most active military wing of the PKK in Turkey; the Civil Protection Units (YPS); and the Kurdistan Freedom Falcons (TAK) -- the PKK affiliate (considered an “offshoot with no operational ties to the PKK” by some researchers), all of which release names and biographies of killed militants.

Confirmation of militant deaths is limited by the PKK's practice, the reasons for which are unknown, to sometimes announce the names of dead militants weeks, months, or even years after they are killed, adding a delay to the PKK militant fatality count.

Crisis Group only counts fatalities which can be confirmed through its open-source methodology, and its numbers should not be seen as a refutation of fatality claims made by the Turkish government.

The exact day of death cannot be established for over 100 confirmed fatalities, including civilians, “individuals of unknown affiliation” and PKK militants. These individuals overwhelmingly died in urban areas under prolonged curfews or in remote mountainous regions, and only the week or month of their deaths can be identified by Crisis Group. In the Crisis Group database, their names are added to the last day of each identified week or month.

Crisis Group classified as seasoned those PKK militants the HPG announced as “commanders” or attributed substantial battlefield roles to (ie, leading a group of militants in the battlefield, being responsible for operations in a certain location). Where available, data for seasoned militants was cross-checked with militants’ date of joining, age and whether they were classified as “senior” in Turkish-language news. Militant fatalities announced as senior by the Turkish military, but not classified as such by the HPG were not included.

4. Violent Events

ACLED tracks a wide variety of political events, including protests, armed clashes, looting and non-violent transfer of territory. For purposes of this analysis, “violent incidents” are defined as those in the following ACLED sub-event types: “air/drone strike”, “armed clash”, “grenade”, “remote explosive/landmine/IED”, “shelling/artillery/missile attack”, “suicide bomb” and “attack”. Event locations are placed randomly within 20km radius of ACLED coding to show their density.  

Data for Syria in 2016 are taken from the Integrated Crisis Early Warning System (ICEWS) database, using event codes approximating the aforementioned ACLED categories (codes 169-205), because ACLED coverage begins in 2017. ICEWS events constitute 0.2 percent of events represented in relevant figures and are consistent with Crisis Group analysts’ understanding of YPGPKK–Turkish state interactions in Syria during 2016.