Examining the Effects of Drug-Related Killings on Philippine Conditional Cash Transfer Beneficiaries in Metro Manila, 2016–2017

Is the Philippine War on Drugs a ‘War on the Poor’? Focusing on beneficiaries of the Philippine Conditional Cash Transfer (CCT) or Pantawid Pamilyang Pilipino Program as the most legible cohort of poor, we examine the effects of the anti-narcotics campaign on impoverished families in Metro Manila from April 2016 to December 2017. From field validation and interviews with families affected by drug-related killings (DRKs), we find that at least 333 victims out of 1,827 identifiable DRK cases in Metro Manila during the study period were CCT beneficiaries. These are extremely conservative figures since field validation did not saturate all cities in Metro Manila and does not include deaths after December 2017 or poor families who are not CCT beneficiaries. The findings illustrate that DRKs negatively affect CCT beneficiaries and their families. Most victims were breadwinners, leading to a decrease in household income. The reduced available income and the social stigma of having a drug-related death in the family often cause children beneficiaries of the CCT program to drop out of school. Widowed parents often find new partners, leaving the children with paternal grandmothers. DRKs are often bookended by other hazards such as flooding, fires, and home demolitions. The direct effects of these DRKs, compounded with disasters and other socio-economic shocks, traumatizes CCT families, erodes social cohesion, and pushes them further into poverty. We conclude with recommendations for the design of support packages to mitigate untoward effects on families, particularly single parent households.


Introduction
'Pantawid, ime-maintain ko 'yan. Yung pang-ulam dagdagan mo (I will maintain the Pantawid program. Increase their funds for food).' -Rodrigo Duterte, 2 February 2016. 1 'If you know of any addicts, go ahead and kill them yourself as getting their parents to do it would be too painful.' -Rodrigo Duterte, 1 July 2016. 2 Three years since President Duterte launched the Philippine War on Drugs, much has been written regarding the degree of violence deployed by state and quasi-state mechanisms to rid the Philippines of an alleged 3.7 million drug users. 3 Approximately 29,000 4 deaths attributable to this policy have been recorded as of July 2019. 5 More recently, the Philippine Drug Enforcement Agency (PDEA) reports 5,601 drug personalities killed in anti-drug operations as of January 2020 (Realnumbers.ph 2020). A 2019 analysis of 23 different datasets from official and supplementary sources observed that official figures are grossly understated and that a more accurate tally is 2.94 times greater than what is reported by police (Ball et al. 2019). Human Rights Watch (2017) has condemned the punitive treatment of suspected drug users and dealers. The concentration of victims in urban poor communities has led international and local groups such as Amnesty International (2017) and PhilRights (2018) to report that the war on drugs is a war on the poor. This is supported by the Ateneo School of Government's (2018) preliminary analysis of 5,021 DRK cases, which found that 47% of those killed were low-level drug suspects and 40% of the killings were concentrated in poor communities in Metro Manila.
Due to the limitations of poverty data in the Philippines, we focus on beneficiaries of the Philippine Conditional Cash Transfer (known as the Pantawid Pamilyang Pilipino Program or 4Ps, and herein referred to as CCT) to study the direct and indirect effects of drug-related killings (henceforth referred to as DRKs) on the lives of impoverished families in Metro Manila during the first phase of the Duterte administration's anti-narcotics policy. While the CCT beneficiaries are only a subset of the universe of poor households in the Philippines, they are still the most 'legible' to policy interventions since household-level data are encoded in official government databases. The Philippine government generates poverty data annually, but this is based on an official national poverty line whose definition is periodically revised and only provides an estimate of the magnitude of poor in the country. The database of Listahanan, the government's official list of the poor, 6 and that of the CCT program contain data from actual families that were assessed to be poor.
The CCT provides monthly cash transfers to poor households in return for compliance to conditionalities to break intergenerational poverty by 'keeping children in school and keeping them healthy.' Currently, the CCT covers 4.4 million households 7 nationwide. As of March 2018, approximately 223,000 households were covered by the CCT in Metro Manila.
To examine the direct and indirect effects of DRKs in CCT households, we analyze patterns across communities in Metro Manila, focusing on four of its sixteen component cities: Manila, Mandaluyong, Marikina, and Quezon City. 8 We seek to answer the following questions: • How many CCT beneficiaries were victims of DRKs in Metro Manila from 2016 to 2017? • Where are these CCT beneficiaries who were victims of DRKs located? • How does the DRK of a family member affect CCT families and their communities?
The CCT is the flagship social protection program of the Philippines and was used to anchor government's other socioeconomic interventions. With the Duterte administration's pivot to intensified anti-narcotics operations, interrogating the effects of DRKs on CCT beneficiaries allows for policy assessment at household and community levels. While we focus on CCT beneficiaries because they are the most legible cohort of the country's poor, this does not mean that other victims of DRKs are not poor.
This study stems from more than two years of fieldwork in Metro Manila. We draw our findings from three bodies of data: (i) a list of DRKs; (ii) a validated list of CCT beneficiaries that are confirmed to be DRK victims (henceforth referred to as the CCT-DRK list); and (iii) a set of key informant interviews (KIIs) with affected families.

Poverty in the Philippines and the CCT
Poverty in the Philippines is defined by the Philippine Statistics Authority (PSA) in terms of families living below the official national poverty line and therefore unable to sustainably afford basic needs such as food, education, health, and other essentials for living. While there are unresolved debates on how to define and measure poverty in the country (i.e. via income and consumption vs. self-perception), the Philippine Government targets social protection and poverty reduction programs using a database of poor households called Listahanan, which combines both income and non-income indicators to identify a poor household. The CCT beneficiaries are poor families with children aged 0 to 18 years old as identified through the government's Listahanan.
As one of the most vulnerable countries in the world, ranking second in terms of internal displacement (Global Climate Risk Index 2018), the Philippines' exposure to various risks has an impact on poverty. Almost 17% percent of Filipinos still live below the national poverty line, lagging behind other countries in Southeast Asia (Philippine Statistics Authority [PSA] 2019; ASEAN 2020). Poverty levels are also linked to educational attainment, with two-thirds of poor households headed by people with only elementary school education (Asian Development Bank 2009), while 13.54% of children in secondary education are already working to help their families (Maligalig et al. 2010).
Cash transfers contribute to short-term and long-term welfare outcomes through income augmentation and human capital investment in education and health. In the Philippines, the CCT began implementation in 2007 and was passed into law in 2019. As of 2019, government investment in the CCT totals approximately 500 billion pesos (10 billion USD using October 2019 conversion rates) including loans from international financial institutions to benefit the 4.4 million households currently enrolled in the CCT program. The Philippine Government spends an average of 16,000 pesos per year per CCT family based on compliance to conditionalities. 9 Acosta and Velarde's (2015) benefit incidence analysis of the CCT demonstrated that the program reduced short-term poverty by adding to the income of CCT households. The CCT reduced the national poverty rate by up to 1.5 percentage points, lifting 1.3 million households out of poverty (World Bank 2018a). Education expenditure among CCT households is 82% higher than non-CCT children based on annual expenditure per school-aged child (Orbeta et al. 2014). In terms of health services, 70% of CCT beneficiary mothers delivered their babies in health facilities, compared to 56% of non-CCT beneficiaries. The 2014 impact evaluation also showed that the program kept older children (12-15 years old) in school. The Philippine CCT also anchors other social protection programs such as the Sustainable Livelihood Program (SLP) 10 and the National Health Insurance Program (NHIP).

Poverty and drug wars
During the 2016 elections, then-Mayor Duterte was elected on a platform of eradicating crime and illegal drugs within six months, while trumpeting the principles of courage, compassion, and change (tapang, malasakit, at pagbabago) -a stark contrast to his opponents who campaigned under promises of economic development and poverty reduction. Duterte's campaign promises were blunt. These promises included the massacre of criminals and drug users and dissolving of congress, all of which appealed to 16 million voters who eventually propelled him to the presidency (BBC 2016). On its first day in office, the Duterte administration operationalized the Drug War through projects such as Oplan Tokhang (Command Memorandum Circular 2016-16) of the PNP and the Department of Interior and Local Government (DILG)'s MASA MASID, 11 which serves as the linchpin of community-based drug rehabilitation initiatives of the government.
The administration was quick to publicize its accomplishments, with the Philippine National Police (PNP) reporting that the nationwide crime rate from July 2016 to June 2018 dropped by 21.48% compared to the same period from 2014 to 2016. However as in the case of Metro Manila, the murder rate increased by 112% with 3,444 cases from July 2016 to June 2018 compared to 1,621 murder cases from July 2014 to June 2016 (Macapagal 2018). The sudden increase in murder cases can be attributed to the increase in DRKs in Metro Manila during this period, including the deaths of children who were directly or indirectly targeted by anti-drug operations.
UNICEF (2018) estimates that at least 33 children have been killed as ' collateral damage,' while the Children's Legal Rights and Development Center (CLDRC) estimates that 74 children have been killed as of December 2017 (Child Rights Network 2018). In 2017, the death of 17-year-old Kian Lloyd de los Santos triggered mass condemnation and the only recorded murder conviction of three policemen under the Duterte administration. David and Mendoza (2018) state that anywhere from 18,000 to 32,000 children may have been orphaned due to the Drug War, a number that excludes children arrested or detained for drug-related charges. This is supported by the statement made by former DSWD Official Hope Hervilla (cited in See 2016) where an estimated 18,000 children were affected by the Drug War in 2016.
Despite the high death toll, the Philippine Drug War has not been able to curb the country's illegal drug supply. A 2019 report released by Vice President Leni Robredo after her short-lived appointment as Chair of the Inter-Agency Committee on Anti-Illegal Drugs (ICAD) questioned how PDEA was able to seize a mere 1,344 kilos 12 of shabu (methamphetamine) out of the projected 3,000 kilos of shabu in circulation every week (Cepeda 2020). This has led to criticisms about the outsize focus on low-level drug users and slumdwellers instead of arresting so-called 'high value targets' or drug lords. Nevertheless, Duterte's approval ratings remain high particularly amongst the upper class.
The Philippines' early experience of surges in DRKs is similar to the experience of other countries where 'drug wars' become prone to abuse and increased violence among the poor (Vitale 2017; Sandvik and Hoelscher 2016). Aside from the tendency to strengthen police and military power in the guise of order and public safety (Correia and Wall 2018), ' drug wars' are also built on the concept of the criminal as the social enemy who can be systematically dehumanized and stripped of rights.
Violence that comes with state-sanctioned drug wars, as in the case of the Philippines, is a development concern. The dynamic nature of poverty makes poor and near-poor households move up and down the poverty line depending on their exposure to socio-economic shocks. These shocks include various forms of violence and conflict plaguing the Philippines, which range from petty crime to multiple ideology-based armed conflicts. Global studies show that a loss of a household head amongst poor families often causes a reduction in investments in children's human capital (Gertler et al. 2006;Case et al. 2004). Intergenerational trauma is another direct consequence of sustained violence, as in that deployed by anti-illegal drug campaigns. Intergenerational trauma occurs when children who witness abuses or violence possess a higher tendency to perpetrate violence later in life (World Bank 2011: 60).

Building the DRK and CCT-DRK databases
There is no publicly-available, disaggregated, and complete database for DRKs in the Philippines. The PNP releases official data for deaths in police operations and deaths under investigation (DUIs) only as aggregates. While multiple databases have been managed by different organizations since the Philippine Drug War started in 2016, these are subsets of an unknown universe of DRKs. The main challenge in doing research on DRKs is collecting and cross-referencing data, especially as the killings still continue to date. Given the uneven distribution of partners on the ground, we were able to collect more data from areas where partner organizations have a strong presence, particularly Manila, Quezon City, Marikina, Mandaluyong, Caloocan, Malabon, and Navotas. From April 2016 to December 2017, we collected a total of 2,267 DRK cases from community sources including civil society organizations, community organizers, one local police station, and media reports in Metro Manila. The process of data collection and validation employed in this study is described in Figure 1.

Database matching and descriptive statistics
Cross-referencing with other data sources was necessary to remove duplicates. Of the 2,267 entries, 1,827 DRK cases were identifiable with first and last names, which makes them eligible for cross-validation with other databases.
A first list of 604 DRK cases identified as CCT beneficiaries was generated from community reports and initial name-matching done by the DSWD-NCR. Due to time and resource limitations, only 333 of these 604 cases were validated 13 through house-to-house visits and through a second desk-based name-matching process. More than a hundred cases from the second name matching of DSWD-NCR were also in the list of 333 CCT-DRK validated cases. Basic descriptive statistics were generated based on available variables such as age and sex, type of DRK operation, time of death, and location of death.
13 Validated CCT-DRK households are those who were able to provide their CCT-household ID number and a copy of their ID.

Key Informant Interviews (KII)
The respondents were CCT beneficiaries who lost at least one family member to DRKs. The interviews followed a life-story format and were carried out in four areas: the district of Tondo in Manila, Quezon City, 14 Mandaluyong, and Marikina. These areas were selected based on the recommendation of partner organizations. 15 Respondents were selected using purposive sampling based on three factors: i) they are validated CCT beneficiaries; ii) their willingness to take part in the study as respondents; and iii) their level of personal security. The safety of the respondents and field validators was prioritized given the sensitive information collected by the study. Consent forms were obtained and aliases were used to protect respondents' identities. Table 1 shows that the four KIIs areas in Metro Manila have different socio-economic, political, and hazard vulnerability profiles. Since the KII areas are predominantly informal communities, respondents are often vulnerable to numerous hazards, which include displacement, fire, and flooding. It also shows the distribution of CCT households where Manila and Quezon City have the greatest number of beneficiaries.
30 out of 31 respondents were female, often mothers and spouses of DRK victims. Table 2 shows that 13 out of 31 respondents were mothers of victims aged between 50-85 years old; 12 were spouses aged between 25-50 years old; and the remaining six are extended family members. Three respondents are from Tondo, Manila, eight are from Marikina, 14 are from Mandaluyong, four from Quezon City, and two requested their location to be withheld for security reasons. There were twice as many respondents in Mandaluyong 14 Due to security threats while doing interviews in Tondo and Quezon City, we decided to discontinue KIIs in these areas in the course of the study. 15 Manila and Quezon City were difficult areas to get KII respondents because of the fear that respondents' lives will be put at risk.
Meanwhile, the magnitude of killings in Marikina and Mandaluyong is difficult to establish, given that these areas are not always covered by media reports. Due to the perception that DRKs in these two cities were underreported, our field partners recommended that these be included as case study areas.  because of the proximity of the respondents' homes to each other. Field interviews had to be cut short in Tondo, Manila, and Quezon City for the safety and security of both respondents and field partners. The average household size for the respondents is seven. The largest household had 13 members while the smallest had three members. The respondents have an average of five children per household, with two children eligible for the CCT. Among the KII respondents, two experienced multiple deaths in the family due to DRK. Three cases experienced ' collateral damage' during the operation, where an additional family member was hurt but not killed.

Validation and matching of CCT-DRK cases
Of the 1,827 identifiable DRK cases, 604 cases were reported and found to be probable CCT-DRK cases. The researchers sought to validate all 604 cases, but due to limited resources and security constraints, only 333 were physically validated. Based on average household size for poor families in the Philippines, the list of 333 validated CCT-DRK households in Metro Manila is equivalent to a range of 1,365 to 1,865 individuals directly affected by a DRK in the family. 16 Thus, available data indicate that conservatively, one out of five DRK victims in Metro Manila during the period 2016-2017 may be a CCT beneficiary.     Figure 4) but spiked again in June 2017, two months after Duterte lifted the suspension of Oplan Tokhang. Nevertheless, these numbers must be considered highly conservative due to the likelihood of unreported CCT-DRK cases.

Most victims were male breadwinners, leaving behind female headed households
Majority of the 333 validated cases were male (92.5%) and a handful were female (Annex 1). For cases where age is known, most victims were between 30-39 years old, followed by victims aged 40-44 years old. Most of the CCT-DRK victims were within the working and employable age range (Annex 2). This is consistent with the KII findings, as majority of the respondents (28) said that their deceased relatives were male while three lost a female relative. Regardless of sex, most victims were their families' main providers. Each victim was supporting at least two child-beneficiaries of the CCT.
Orphaned children are often left in the care of the grandparents. More than 50% of the respondents were grandmothers who had to assume the responsibility of taking care of at least two grandchildren after their children were killed in the Drug War. After the death of a husband or partner, widows often move out of the house to remarry and leave their children with the paternal grandparents. The children, often aged between five to 18 years old, are left in the care of grandmothers aged between 55 to 85 years old. The grandmothers no longer have stable sources of income and are usually suffering from various illnesses. The grandmothers are forced to work for additional income aside from the cash grants received from the CCT. Dina, 61 (interview 5 September 2018), suffered from depression after both her husband and son were killed in separate DRKs. Despite diabetes and hypertension, Dina works as a laundry woman so that she can provide for her grandchildren and children. Ditas, 17 85, (interview August 2018) can no longer work and has resorted to begging on the streets so that she and her grandchild can buy food and pay for basic services. 17 Ditas passed away a few weeks after the interview, leaving her grandchild a total orphan. These women now carry the multiple burdens of providing for their family and taking care of young children. One respondent expressed that women still do not get equal access to earnings as compared to men and must resort to part-time work. Meanwhile, many of the widows shared that they have a hard time sustaining rent and food needs. Trisha, 34 (interview 4 September 2018), said she cannot afford to send all five children to school or feed them all. As such, her children tend to stop schooling. Unable to raise rent, families like Ditas' are also forced to live with their parents or siblings as a sub-household.

The CCT-DRK killings happen at home, at varying times of day, and sometimes affect multiple members of a family
Despite limited available data, certain patterns emerge. The location of death was not available for almost all (88.29%) of the 333 validated CCT-DRK cases. However, for the 39 cases with available location data, it is striking that 22 CCT-DRK entries were killed at home, while ten were killed on the street and five were killed in public (Annex 3). For the 155 CCT-DRK victims with known time of death, most of them (78) were killed in the evening (Annex 4). Many of the CCT-DRK victims were also killed by unidentified assailants while 31.2% were killed by state actors (Annex 5).
From the 31 interviews, 18 many of the victims were also killed in their homes. Katrina's 31-year-old daughter, Cherry, was asleep on the second floor of their shanty when 32 individuals or 16 pairs of armed ridingin-tandem 19 assailants barged in. Katrina (interview 19 September 2018) recalls: 'Cherry was shot four times in the head. She was three months pregnant [with] her third child. [She] was in the barangay [drug] watchlist [along] with her partner, Toto.' Seven KII respondents said that their relatives were killed in public spaces, often along the street near their homes. Only 20 out of 31 respondents were able to indicate the time when their family member was killed. Eight respondents said that the DRK happened during the day, while 12 said their relatives were killed either early morning or in the evening.
From the 333 CCT-DRK cases, there were 12 cases where multiple members of the family were killed. Eight out of these 12 cases involved 18 individuals who belong to parent-and-children pairs. Two interviewed respondents admitted to multiple DRKs in the family. Dina, 61 (interview 5 September 2018) shared that she lost her husband and her son within the span of six months.
'We just finished [eating] dinner … around 10 PM. My husband was closing the store when 14 people, seven riding-in-tandems, barged into our house looking for my son […] They shot at my husband, despite my grandchild begging them not to shoot him. Three months after, our house was 18 To protect the respondents' identities, all names used in this study are not their real names. 19 'Riding-in-tandem' is the local term used to describe pairs of hired killers often seen riding pillion aboard motorcycles during operations.  'I earn 150 pesos a day, not even enough to sustain our daily meals. I saved up money so that I can buy their uniforms but they [my daughters] still dropped out. They are both in high school… no matter how I convince them that high school is fun; they dropped out because they were being bullied after what happened to their father […] Right now they are under ALS (alternative learning system). They are supposed to get the CCT grants but now, we're not getting anything.'

Exposure to multiple vulnerabilities
All interviewed households were from informal settlements, occupying houses made of light materials with no security of tenure, and had experienced both natural and man-made disasters such as flooding, fire, and demolition in the last two years. Many of these families are single income households, with the (often male) head of household serving as the sole breadwinner. Their poor living conditions also trigger various health issues, with at least half of the respondents experiencing socio-economic shocks brought about by having a sick member of the family.
Lorna, 58 (interview 17 August 2018), lost her son, Kulit, when he was shot by unknown assailants in September 2016. Prior to Kulit's death, they lost their house to a fire that almost wiped out their whole community in Mandaluyong. Kulit left behind three children under Lorna's care. She is suffering from diabetes and needs a cataract surgery. They have yet to rebuild their house because without the income of Kulit, her whole family is dependent on her husband's income as a construction worker. Lorna expressed the challenge of providing for their day-to-day food needs. Aside from caring for Kulit's three orphaned children, Lorna is also taking care of six other grandchildren.
Similarly, Mel (interview 20 August 2018) and her two children had to evacuate their home a few days before the interview because of severe and prolonged flooding in their area in Marikina. Since her husband's death, Mel and her children have been staying with her sick father, whose house is also made of light materials and the constant exposure to the elements makes living on the second floor a risk in itself.
For widows like Toni, 28 (interview 17 August 2018), her husband's death was the biggest tragedy. She recounts: '[A] Few days after Jerry was killed, our house was one of the hundreds of houses [that] got… burned down…I was not at home, so we were not able to save anything, including Jerry's ashes, which was on our small altar…[I] tried to save it but it's gone. It's worse than the fire because not [having] his ashes meant that my young children do not have anything to remember their father by.'

DRKs can cause trauma among families and communities and is negatively affecting social cohesion
Due to the stigma tied to DRKs, neighbors and relatives are afraid to associate with bereaved families and are unable to condole at wakes. This also leads to weaker support systems for affected families. Children are bullied and drop out of school. KII respondents noted that some informant ' assets' and assassins are also members of the community, eroding trust among neighbors. One orphaned grandchild of a respondent is saving up money to buy a gun so that he can avenge his father's death. Lorna, 58 (interview 17 August 2018) described her exchange with her 18-year-old grandson: 'That's what he wants, to be a policeman when he grows up. I told him he should study hard so he can achieve his dream. He says he's a big boy now, that he's saving up his school money, bit by bit, so he can buy a gun. That's what he said, ma'am. So we can have something to fight back. To go headto-head with them. Our neighbor…he really does tokhang. That's his only job. That's why he's got a big house now. When he killed our other neighbor, he gave "charity." He treated the neighborhood to beer. Bought ten cases for everyone.' Although most female respondents described turning to religion for support-with at least two respondents converting to different faiths, from Catholicism to born-again Christianity and Islam in particular-the interviews pointed to emotions of anger, sadness, and despair that remain unaddressed by both adults and children.
Traumatic effects on surviving family members and children are also observed. Ging (interview 11 August 2019) described how her youngest son remains quiet and withdrawn after being shot in the leg while witnessing the death of his father almost two years ago. The son was only able to regain the ability to walk after extensive surgery.

DRKs negate the educational gains of the CCT for CCT-DRK families
The DRKs negate the gains of the program for CCT-DRK families, because instead of keeping children in school, the loss of family members due to DRK make CCT children drop out of school.
With an average of two children enrolled in the CCT, the 604 reported CCT-DRK cases and the validated 333 CCT-DRK cases means that for Metro Manila alone, a conservative estimate of 600 to 1,200 children beneficiaries were directly affected by DRK. Children belonging to these affected households are at risk of dropping out of school and not receiving the cash grants. This is consistent with related literature that finds the death of a parent reduces a poor family's spending on education and human capital (Gertler et al. 2006;Case et al. 2004).
These families also face the risk of losing the CCT benefits because of failure to comply with health and education conditionalities. Based on the 31 KIIs, 19 out of 50 or 38% of children eligible for the CCT were not in school. This can be attributed to observance of traditional grieving periods, compounded with the trauma and the stigma of being children of 'Tokhang victims,' deter these children from returning to school. This is exacerbated by the trauma endured by the children for losing a family member in violent circumstances--or even witnessing the actual DRK. Since they now have to work, mothers and grandmothers often fail to meet health conditionalities when they miss the monthly family development sessions.
The initial effect of DRKs on CCT beneficiaries is a manifestation of what Albert and Vizmanos (2018) described as capability deprivation. Children and other members of the households may be deprived of the opportunity to improve human capital because they had to stop school and work in order to fill in the income gap brought about by the family member's killing. Combined with national policies reallocating investments to security initiatives and infrastructure rather than social protection programs, the Philippines is lagging behind according to global standards 20 (ADB, 2013cited in Diokno-Sicat & Mariano 2018. The failure to maximize investments in human capital of poor families may lead to chronic poverty.

The rise of female-headed households
The direct consequence of the deaths of household heads is that women, either widows or grandmothers, are left with the burden of providing for their families and raising the orphaned children. Chant (2014) speaks of the feminization of responsibility, wherein women now assume greater liability for dealing with poverty. Women now carry the responsibility of productive and domestic labor. With few skills, limited education, and low job prospects, many DRK widows are forced to find new husbands to ensure a source of income.
As widows remarry or find new partners, orphaned children are often left with the grandparents. In the case of CCT-DRK households, the grandmother takes over the responsibility of being the primary grantee of the program. This scenario is encapsulated in the Filipino phrase repeatedly mentioned in the course of data collection, 'Si tatay sumakabilang buhay, si nanay sumakabilang bahay, si lola ang naiwan (Dad crossed over to the afterlife, mom moved to another house-only grandma is left).' The DRK is an economic shock that highlights the limited social protection programs available to the poor, especially for women. Since 92.5% of CCT-DRK victims are male household heads, each killing leads to a loss of income of approximately 4,000-10,000 pesos per household per month, based on KII findings. DRK widows have limited choices and often resort to low-income domestic labor, with anecdotal evidence of turning to sex work.

Legibility: A community disadvantage
Identifying eligible beneficiaries is critical in implementing social protection programs. For Scott (1998), this is tied to legibility, or the exercise of the state of its power to document and control its population. Programs like the CCT make the poor legible because they are represented in various state systems and they benefit from the state's programs and services. CCT beneficiary lists are available to the barangay and the local government unit. While it is beneficial for the poor to be legible so that they may access the various social services offered by government, legibility is a disadvantage when existing state systems become prone to abuse.
The validation process shows that CCT-DRK victims were killed in the first six months of the Drug War (July to December 2016). There is a strong indication that available lists of CCT beneficiaries were used as an early source of information in identifying the location of suspected drug users and dealers in urban poor communities in Metro Manila while institutional structures to operationalize the Philippine Drug War were being developed.
Being on the side of the 'informal' also makes it easy for the state to generalize informal settler families as 'undesirables' because of their proximity to crime and violence. Respondents experienced difficulties in accessing government services such as death or burial assistance or benefits for new single parent households due the stigma of their kin being victims of DRK. The few families who have the wherewithal to flee often do not update their new addresses-making the monitoring of their compliance to CCT conditionalities difficult for the DSWD to track.

Conclusions and Recommendations
So, is the Philippine War on Drugs a 'War on the Poor'? The available evidence shows that drug-related killings in 2016-2017 negatively affected the poor in Metro Manila, as illustrated by the experiences of CCT beneficiaries. Cases collected for the purpose of this study and by other datasets show that DRKs for the period were concentrated in poor urban neighborhoods that are affected by multiple vulnerabilities. Factoring in data gaps, the available qualitative and quantitative information from CCT households suggests that one out of five of the DRK victims belong to the CCT and are classified poor. By killing mostly male household heads, the effect and magnitude of the Drug War-compounded by various socio-economic shocks including damage to homes due to fires and flooding-reduces income and makes it difficult for orphaned children to stay in school, pushing already-deprived and vulnerable families further into poverty.
The death of CCT-DRK beneficiaries undermine efforts for human capital formation especially when beneficiary children drop out of school due to trauma or lack of financial support. Despite the CCT's role as the flagship government program for poverty reduction, early evidence suggests that the decade long investment made on CCT households in Metro Manila is being undermined by the DRK. Killing household heads reduces a family's income for food, clothing, shelter, and health. CCT brings communities together, but the Drug War has eroded respondents' trust between neighbors and in the State.
To mitigate these effects, we recommend that further killings must be prevented. Support should be provided to those left behind, particularly packages to address challenges to income, psychological and overall health, and the education of children, as well as harm reduction programs to address community drug use. The needs of sub-households, solo parents, and other female-headed households must be considered.
Specific monitoring tools must be developed for orphaned children due to DRK to ensure that they will return to school and that issues such as bullying are addressed. Given that children and youth are being exposed to the killing of family members and neighbors on a regular basis, the effects of transgenerational trauma and how it contributes to intergenerational poverty will need to be factored in the design of programs for addressing the fallout in the years to come.
Issues with both poverty and drug war data demand extensive work moving forward, as the 333 validated CCT-DRK households in Metro Manila are likely to be a small subset of the true affected population. Government and other relevant institutions should take steps to identify other affected households not only in Metro Manila but in other regions nationwide. However, the privacy of 'legible' families must be protected, particularly those included in government databases such as Listahanan or lists of drug surrenderees or former combatants and their families.
The results also raise questions for future research and the need for longer and more expansive dataset to allow studies on phenomena such as 'palit-ulo' and 'palit-puri ', 21 which highlights the ease with which punitive policies can be abused.
Ultimately, the relative costs and benefits of the current Philippine drug policy should be reviewed, along with safety nets and support programs to ensure that the most vulnerable in society are unharmed.
Annexes 21 Palit-ulo ('head swap') is a practice linked to the drug war wherein another victim is killed, often a family member, in exchange for the initial target. Meanwhile, palit-puri ('sex swap') is linked to reports involving police officers who sexually abuse and rape drug personalities, wives, or female children of suspected drug personalities in exchange for dropping the drug-related charges.

Competing Interests
Maria Karla Abigail Pangilinan and Nastassja Quijano worked in DSWD from July 2010 to March 2016 and May 2012 to July 2016, respectively. This research was conducted independently using an objective methodology and was self-funded. The views presented in this study was a product of a rigorous research process and does not represent the views of these institutions. All other authors have no competing interests.