Research

Can the provision of minority rights effectively resolve conflicts? This paper challenges the prevailing literature on ethnic conflict, which predominantly focuses on large ethnic groups that can afford their own army. Instead, I investigate the consequences of collective rights granted to small indigenous peoples in the Philippines. I argue that privileges bestowed upon minorities without sufficient self- or state-enforcement mechanisms can trigger aggression from local majority elites who have economic interests in the ethnic territory. This, in turn, can lead to the militarization of both development-oriented local elites and indigenous communities striving to protect their tribal lands. Using a Difference-in-Differences framework, I provide evidence that the introduction of a progressive law aimed at safeguarding indigenous peoples’ rights actually attracted insurgencies into indigenous areas, particularly those with mining operations. Moreover, even among indigenous peoples, those who applied for these privileges were more likely to experience conflict.

Where does the counterinsurgency successfully suppress rebel activities? The literature in conflict emphasizes the role of civilians as the source of intelligence. However, the relationship between the local civilians and the military is often mediated by a local agent: local governments and police. Formulating the role and incentives of local executives who supervise local police, this study argues that the party alignment between the center and local executives constitutes an important condition for the suppression of rebel activities in the context of conflict in democratic countries. Capitalizing on the case of nation-wide communist insurgency and the closely run mayoral elections in the Philippines, I provide causal evidence demonstrating that electing a mayor who has the same party affiliation with the President decreases the violent clashes between the military and rebels.

How does a peace agreement influence public attitudes in the post-conflict society? In spite of the important role of public opinion as a domestic enforcer of peace deals, their micro-foundations have never been examined. This study uses an endorsement survey experiment fielded in post-conflict Mindanao in the Philippines, to measure the effect of elite consensus on public support for post-conflict policies. The sheer fact that conflict parties agree to the policies does not increase the support. However, when informed that the agreement is a formal peace treaty, civilians give an additional support to it.

Recognizing the importance of grassroots and public opinions in peace implementations, peacemakers are increasingly engaging with non-elites. Joint mediation efforts by domestic and international actors are gaining attention as potential guarantors of enduring peace, encouraging popular participation and support for peace processes. However, the premise that mediators positively influence public opinions is yet to be tested. To explore this, a survey experiment was conducted in Muslim Mindanao, Philippines, where a noble hybrid mediation system is in place. Contrary to expectations, the survey found that elite-outsider mediation negatively affects public attitudes toward the peace process, while insider and hybrid mediation showed no strong effects. These findings push future work to interrogate more deeply the extent to which the public perceives insider mediators in civil wars and how they affect mediation effectiveness.

Digital connectivity has become a critical input to modern warfare, but its causal effects on battlefield outcomes remain speculative to date. We exploit a natural experiment arising from SpaceX’s September 2022 decision to enforce geofencing restrictions on Starlink satellite internet in Ukraine, which reduced high-bandwidth connectivity for a subset of frontline locations while leaving comparable nearby areas unaffected. Drawing on high-frequency geospatial data, we compare outcomes across the coverage boundary before and after the policy shift. Ukrainian forces in frontline locations that lost Starlink access retained 30–40% less territory. This gap opens within weeks, widens over time, and never closes, implying a cumulative loss of 516–853 km2 relative to the connected counterfactual. Strikingly, combat frequency and composition were unaffected: firefights, shelling, and drone strikes occurred at similar rates on both sides of the boundary. Forces fought just as often, but they accomplished less. This dissociation identifies a channel through which communications infrastructure shapes conflict: it affects not the quantity of violence, but its conversion into durable military gains. Our findings reinforce how private actors’ decisions can move a front line, and with it the distribution of territory that anchors bargaining over war termination.

This paper introduces the South China Sea Data Initiative (SCSDI), an original dataset on incidents of conflict and coercion in the South China Sea from 2012-2025. Overall reported conflict increases from fewer than 20 events in 2012 to a peak of more than 300 in 2024, corresponding with an increase in China’s foreign policy and military muscularity under President Xi Jinping. The vast majority of the incidents are of the lowest intensity grade; there is no significant change over time in higher intensity events, what we refer to as “physical coercion” or “kinetic use of force.” Similarly, the occurrence of injuries or fatalities are extremely low throughout. Examining two major explanations for contestation in the South China Sea – oil and gas resources, or access to fisheries – we find no evidence of greater contestation in areas with oil/gas discoveries or production, or in fishery exclusion zones. We find limited evidence that leadership transitions in SE Asia produce changes in conflict trends, most prominently in the Philippines. However, these effects are again restricted to the lowest grade clashes. Overall, the data are consistent with a dynamic of primarily performative contestation in the South China Sea, rather than aggression at risk of escalating to war.

How can governments encourage individuals from under-represented groups to run for office? While most of the literature has focused on changes to electoral rules (e.g quotas and reserved seats), we argue that non-electoral policies can also affect political selection, either by increasing the salience of government policy and the returns of being elected to office, by reducing financial constraints faced by potential candidates or by creating opportunities to build and showcase lead-ership skills. In this paper, we explore those issues and test whether a large-scale social protection program in the Philippines affected the decision to run for village elections. Using data on all village elections between 2002 and 2013 and employing both a DiD design using the nationwide rollout and an RCT covering 300 villages, we find that the program increased the number of candidates, especially of program-eligible women, in village elections. Eligible women are also more likely to be elected village captain and councilors. Further analysis indicates that our results are driven by the leadership opportunities that the program created among beneficiary households.