Anti-Poverty Program Reduces Violence in Civil Conflicts

An anti-poverty aid program that’s been implemented in the Philippines for nearly a decade is gaining attention for the progress it has made in not only helping the poor, but also for its role in decreasing political violence and insurgency.

Joe Felter, senior research scholar at Stanford’s Center for International Security and Cooperation (CISAC) and research fellow at the Hoover Institution, presented the results of his joint research on the program before senior political figures at a conference in the Philippine capital of Manila in January.

CISAC senior research scholar Joe Felter (left) joins Philippine President Benigno Aquino (right) and the Secretary of the Department of Social Welfare and Development Corazon “Dinky” Soliman (center) onstage at a conference in Manila.Philippine President Benigno Aquino, and Secretary of the Department of Social Welfare and Development (DSWD) Corazon “Dinky” Soliman, were among the leaders in attendance at the Conference on Sustaining the Gains of the Conditional Cash Transfer Program.

“We worked for several years on this study and it was a privilege to provide these findings and results to senior officials in the Philippine government who are in a position to act on them,” said Felter. “It’s really gratifying to know that academic research can contribute to actual improvements in the conditions, livelihood and safety of those in need.”

The focus of the conference was on the conditional cash-transfer (CCT) anti-poverty aid program called Pantawid Pamilya. Administered by Soliman’s Department of Social Welfare and Development, the Philippines began deploying the program in 2007. It is similar to other CCT programs used in Brazil, Columbia, India, Indonesia and Mexico where households must meet certain income thresholds and basic health and education requirements to qualify for its benefits. CCT programs distribute cash payments to targeted poor households and are proving to be an increasingly popular tool for reducing poverty and improving livelihoods in poverty-affected areas.

The effect of aid on conflict

Felter and his colleagues conducted an analysis of the impact of aid on civil conflict that takes advantage of a randomized control trial (RCT) initiated in the Philippines by the World Bank in 2009 as part of an impact evaluation of the Pantawid Pamilya CCT program. Impact evaluations of CCT programs to date limit their findings to those areas the program was intended to address such as health, education, and employment. Published in the January 2016 Journal of Development Economics, the study estimates the effect of conditional cash transfer (CCT) programs on two other critical outcomes- civil conflict and insurgent influence.

CISAC senior research scholar Joe Felter shakes hands with Philippine President Benigno Aquino onstage at a conference in the Philippine capital of Manila.Conventional wisdom might tell you that increasing developmental aid to conflict-affected nations would uniformly help reduce the violence and stabilize these areas, but there is mixed evidence on the effect of aid on conflict. In fact, recent findings show some forms of development aid and the ways they are delivered can actually exacerbate conflict by creating opportunities for looting and incentives for strategic retaliation. That’s why the new findings by Felter and his colleagues are so important. They found the type of aid, or mechanism administered, may play a critical role in reducing conflict-related incidents.

“Considering the types of conflicts taking place around the globe, it is both timely and important to study how aid can be delivered in a manner that reduces poverty without exacerbating conflict,” said Felter. “Development aid can sometimes have the unintended effect of increasing conflict in civil wars when insurgents believe the successful implementation of government-sponsored development projects will boost support for the government and undermine their position.” 

Felter himself is no stranger to international conflict. He retired from the U.S. Army as a colonel in 2012 following a career as a Special Forces and foreign area officer that took him on missions to Central America, Southeast Asia, Afghanistan, and Iraq.  Now in academia, he uses data and rigorous quantitative methods to help those in the field better understand and more effectively and efficiently approach the challenges of stabilizing conflict areas through development aid and economic assistance. 

Despite the growing popularity of CCTs, and assessments of their effectiveness at reducing poverty and improving livelihood, there is limited evidence on how the payment programs affect the civil conflict often present in these poverty stricken areas. Felter, along with his coauthors Benjamin Crost of the University of Illinois, and Patrick Johnston of RAND Corporation, took advantage of the World Bank’s randomized experiment to identify the effect CCT programs had on conflict-related incidents and the influence of insurgent groups, even though the experiment was not originally designed to study the effect of Pantawid Pamilya on these outcomes. Their research compared these aspects of the CCT program’s impact in treatment villages to control villages in the Philippines from 2009-2011.

The Philippines is home to some of the world’s most protracted civil conflicts, including a separatist insurgency in Mindanao island with roots dating back to Spanish colonial times, and a decades long communist insurgency affecting nearly all of the country’s provinces across this archipelago.

“Studying the impact of conditional cash transfers on political violence and insurgent influence in the Philippines is especially instructive and generalizable because you have multiple, long-running insurgencies, each with distinct characteristics, and with an array of government sponsored aid programs implemented in these areas over time,” said Felter.

Two key findings resulted from the team’s analysis. First, the CCT program caused a substantial reduction in the number of conflict-related incidents in the villages where it was administered. Second, the program was effective at reducing insurgent influence in the treated villages. Significantly, their findings provide evidence that the effects of CCTs can differ from other types of aid interventions based on the type of aid provided and how it is implemented.

“That Pantawid Pamilya helped reduce the presence of rebel groups in the targeted villages is especially consequential.” Felter said. “A program that reduces violence by weakening insurgent influence is likely to have more beneficial long-term effects since insurgent influence can still undermine the rule of law and oppress citizens even without violence.”

Not all aid programs created equal

An effective aid program such as this can result in more than an economic boost for a village or community and a reduction in violence. It can also provide a psychological victory that enables the government to gain increased support from the local population – effectively “winning hearts and minds” – thus potentially enabling the government to gain better security through increased cooperation and information sharing about insurgents from the population. This is a win-win result, especially in regions where insurgents often gain support by exposing weaknesses of the government, not just through fear and coercion. Insurgents win when they are able to achieve legitimacy in the eyes of a local population whose own government is unable to provide for their basic needs.

However, a “winning hearts and minds” strategy for disbursing government aid can sometimes backfire depending on how these programs are carried out. For example, KALAHI-CIDSS, a large-scale community-driven development (CDD) infrastructure program took place in similar regions in the Philippines during the same time period as the Pantawid Pamilya experiment period. This aid program was also implemented by the DSWD, but in some cases led to different and unintended results. The CDD program was designed to empower the poorest Filipino municipalities through enhanced participation in community projects and training, but the way in which the projects were determined and the mechanisms they were delivered created incentives and opportunities for insurgents to attack the projects, resulting in increased local conflict in some cases where the program was implemented. CDD programs involve a series of public meetings and result in the implementation of widely publicized and often highly visible infrastructure projects. As a result, insurgents often attack these government “hearts and minds” initiatives that, if successful, threaten to shift popular support away from their rebel groups and towards the government.

In contrast to CDD programs, CCT programs disburse aid directly to its beneficiaries’ bank accounts, making it difficult for insurgents to anticipate when and where the transfers are occurring and inhibiting their capacity to disrupt and dismantle the program. The findings in Felter’s study provide preliminary evidence that the type of aid and mechanism in which it is delivered can be a major factor in determining its impact on civil conflict.

“The stakes are high in human and economic terms when it comes to stabilizing conflict areas and preventing a return of the deadly violence associated with civil wars and insurgency,” said Felter.

The results of this study provide rare empirical evidence that some forms of aid, and how it is implemented can reduce the intensity of civil conflict and the influence of the groups responsible for it. This evidence can help governments determine what type of aid to invest in to achieve their desired results.

“Distributing aid effectively and achieving maximum benefits from these investments is definitely a challenge and an area where more research is needed to better appreciate the many nuances and complexities of these efforts,” said Felter.

During the two-day conference in Manila, President Aquino noted how his administration had increased the CCT budget to cover close to 4.4 million poor households, up from 786,000 five years ago.

You can read Felter’s full paper in the January issue of the Journal of Development Economics.