Philippines “war on drugs”

Coding decisions on drug-related violence in the Philippines

Background

Former President Rodrigo Duterte assumed office in June 2016 and declared a “war on drugs” shortly thereafter. This campaign has resulted in the deaths of thousands of Filipinos, most of whom were from poor, urban communities.¹ The killings were carried out by the Philippine National Police, the Philippine Drug Enforcement Agency, and anti-drug vigilantes and have been widely criticized as unlawful. Human rights organizations have documented evidence of falsified police accounts,² and these actions have been characterized by some international bodies as possible crimes against humanity.³ 

In 2022, Duterte’s ally Ferdinand “Bongbong” Marcos Jr. assumed the presidency, pledging a less punitive approach to the drug war. However, rights groups note that violence linked to anti-drug operations has persisted under his administration.⁴

While ACLED does not normally include criminal violence in its dataset, given the highly politicized nature of drug-related violence in the Philippines, ACLED considers such events to be political violence and thereby includes them in its Philippines dataset.⁵ ACLED includes drug-related violence in the Philippines in the following situations:

• Killings of drug suspects by government security forces (police/military).
• Clashes between armed drug suspects and government security forces (police/military).
• The killing of drug suspects by vigilantes.
• Violence between drug suspects, including clashes among gangs.

There are also a small number of events in which drug suspects are attacked by unidentified armed groups. In these rare cases, reporting is too limited to make reliable assumptions about the perpetrator. 

ACLED employs an intentionally broad definition of “drug suspect,” which encompasses any actor reported to have a link to drugs — whether credible, fabricated, or mistaken — in order to fully capture the scope of the drug war.

Given these challenges, the following sections outline the methodology ACLED applies in order to consistently capture patterns and trends of drug-related violence in the Philippines. 

The killing of drug suspects by government security forces

Local police reports frequently describe alleged shoot-outs between drug suspects and state forces. These reports are subsequently picked up by local media. While drug suspects are almost always reported as being armed in local media, journalists and human rights groups often dispute these claims.⁶ These worries were reiterated in discussions with a number of Filipino human rights organizations consulted by ACLED.

When reports indicate that only drug suspects were killed, with no casualties among state forces, ACLED does not assume that the suspects were armed. These events are therefore coded as “Violence against civilians,” State forces are coded as a primary actor, and Drug Suspect as the civilian actor. 

Unverified reports by state forces that suggest or infer that the suspects were armed are treated with significant caution and scrutiny, as the armed responses of the alleged drug suspects are frequently staged or fabricated, according to some accounts.⁷ These accounts include police narratives in local media, accusations of planted weapons, and incentives that encourage extrajudicial executions of unarmed suspects.⁸ In this vein, such cases are coded as “Violence against civilians” within the ACLED dataset. The relevant state force(s) are coded as Actor 1, civilians as Actor 2, and the drug suspect is coded as Associated Actor 2. 

Clashes between armed drug suspects and government security forces (police/military) 

In cases where state forces (police or military) are killed or injured, ACLED assumes that the drug suspects involved were armed. These events are coded as “Battles,” and the relevant state forces and “Armed Drug Suspects” as Actor 1 and Actor 2, respectively. 

The killing of drug suspects by vigilantes

Some killings are carried out by suspected vigilantes. These actors are often described as riding in tandem on motorcycles, and in some cases, leaving notes that identify victims as drug suspects. Reports and patterns suggest that these groups often function as pro-government militias,⁹ and may maintain indirect links to state forces. There are reported instances of suspected police involvement with these groups.¹⁰

Such cases are coded under the “Violence against civilians” event type, with “Anti-Drug Vigilantes” and “Drug Suspects” as the respective actors. 

When victims of these attacks are members of state forces themselves, ACLED will code the event type depending on the nature of the operation: One-sided attacks are coded as “Violence against civilians,” while exchanges involving casualties on both sides are coded as “Battles.” 

Violence between drug suspects

Clashes between armed drug gangs are coded under the “Battles” event type, with both Actor 1 and Actor 2 coded as “Armed Drug Suspects.” When armed drug suspects attack other drug suspects without evidence of a return of fire, these events are coded under “Violence against civilians.” 

Tracking events related to the Philippines’ war on drugs in the ACLED dataset OR How to identify these events in the ACLED dataset?

In order to systematically monitor and identify emerging trends in drug-related violence and incidents in the Philippines, ACLED employs a tag to track all the above-mentioned violent events and additional non-violent Strategic development events such as Arrests, Disrupted weapons use, Security measures and, Looting/property destruction (see more on Strategic development events in the ACLED codebook). Events related to the war on drugs in the Philippines can be tracked by filtering for the “war-on-drugs-related event” tag coded in the Tags column in the ACLED dataset. The tag captures events related to the war on drugs in the Philippines back to 2016, the start date of ACLED’s coverage of the Philippines.