Coding decisions on drug-related violence
Published On: 26 April 2024 | Last Updated: 1 April 2018
President Rodrigo Duterte assumed office in the Philippines in June 2016, and declared a ‘war on drugs’ shortly thereafter. This ‘war’ has resulted in the direct and indirect deaths of thousands of Filipinos, mostly urban poor, carried out by the Philippine National Police, as well as by anti-drug vigilantes. These killings, initiated and encouraged by Duterte and other senior officials could amount to crimes against humanity. Human rights organizations, both local and international, have found evidence of police falsifying evidence to justify the unlawful killings.
Given the highly politicized nature of the drug-related violence in the Philippines, this is categorized as political violence and is recorded by ACLED. This type of violence manifests in a number of ways. ACLED includes drug violence in the Philippines – the killing of drug suspects by both government security forces (police/military) and by “vigilantes”, as well as inter-gang-related violence. Given the nature of these reports and their suspected veracity, methodological decisions are outlined below denoting the assumptions ACLED has made in efforts to accurately capture the patterns and trends of violence in the Philippines.
Local police reports often note shoot-outs between drug suspects and state forces, which are then picked up and reported in local media. While drug suspects are almost always reported as having been armed, this may not have always been the case. Suspicions arise when journalists investigate the veracity of these police reports, and when reporters speak to relatives of victims. These worries were reiterated in discussions with a number of Filipino human rights organizations consulted by ACLED.
In cases where state forces (police or military) are killed or injured, it is assumed that the drug suspect(s) in question was indeed armed; such cases are hence coded as “Battle-No change of territory” within the ACLED dataset with the relevant state force(s) as one primary actor and “armed drug suspects” as the other. In cases where state forces suffer casualties as a result of the drug war, local organizations consulted noted that the casualties tend to be highlighted in reports to point to the gravity of the crisis and to help justify the actions of state forces.
However, in cases where there is no report of injuries or death to state forces and only on the side of drug suspects, ACLED does not assume that the drug suspect(s) in question was armed. Often such reporting tends to rely on the same ‘canned responses’, suggesting a need for closer scrutiny. Photojournalists noted that victims were often found with multiple bullet wounds, or having been shot in the back or in their hands (suggesting that they were trying to block a bullet and/or displaying they were unarmed), or exhibiting signs of torture. Local organizations also noted that there is often reason to believe that weapons found next to victims’ bodies may have been planted there after the fact to create staged scenes. A recent report by Amnesty International highlights the ‘economy of murder’ that has been established as a result of “cash reward[s] paid out for every dead body incentiviz[ing] police to kill individuals who were mostly poor and often defenseless”—with no payment given for arrests.
Reporting bias across different contexts covered by ACLED suggests that the number of casualties is often one of the most difficult facets of violence to accurately obtain and is oft disputed information in reports. However, in the Filipino context, journalists suggest that state forces like to highlight instances in which one of their own are injured or killed as propaganda. As such, cases where “drug users fired” yet no one on the side of state forces was reported hurt or injured raise concern given that if someone had been hurt or killed, it would have been highlighted.
In this vein, such cases are coded as “Violence against civilians” within the ACLED dataset with the relevant state force(s) as one primary actor and civilians as the other, with “drug suspect” as the associated civilian actor.
Sometimes such cases are carried out by suspected ‘vigilantes’ as opposed to formal state forces. These armed agents are often described as riding on motorcycles in tandem, shooting victims before driving away—rendering victims vulnerable, regardless of whether they were armed or not, through not allowing for any sort of reciprocation by victims. Such vigilantes might also at times leave cardboard notes next to victims, identifying them as drug suspects. These agents behave in ways that more closely resemble political militias (coded with the ‘Inter’ code ‘Political militia’) rather than ‘vigilante militias’ (coded with the ‘Inter’ code ‘Identity Militia’). Unlike ‘vigilante militias’ that may appear elsewhere in the ACLED dataset, these agents are not taking the law into their own hands. Rather, they are more akin to a pro-government militia, with indirect links to the state, especially given Duterte’s inciting of such violence, with reports that some of these vigilantes are actually police in masks and/or rely on police to secure the perimeter in the lead up to such attacks. Such cases are also coded as “Violence against civilians” within the ACLED dataset with ‘anti-drug vigilantes’ as one primary actor and civilians as the other, with “drug suspect” as the associated civilian actor.
In cases where armed drug gangs clash amongst one another, they are coded as “Battle-No change of territory” within the ACLED dataset with both primary actors coded as “armed drug suspects”.
