Research

Research interests

Substantive: Inclusion and Exclusion in Peace Processes, Civil War and Peacebuilding, Rebel Governance, Power-sharing, Climate and Conflict, Spatial and Network Diffusion

Methods: Computational Social Science, Social Network Analysis, Quantitative Text Analysis, Geospatial Data, Quasi-Experimental Designs

Publications

Consociational Democracy in Post-War Lebanon: Maintaining Negative Peace and Stability Amidst Sectarian Divisions

Somboonkittichai, K. (2023). Asian Review, 36(2), 119–140.

DOI

Working papers

Spreading Inclusivity: Network Channels and Inclusive Peace Negotiations in Civil Wars

Near submission

Presented at the VIP Workshop (2026), the JQCon Workshop (2026) and the UCD Centre for Peace and Conflict Research (invited talk, 2026)

Why is the inclusion of non-warring parties in peace negotiations more likely in some civil wars than others? This paper argues that inclusion spreads across conflicts through consequences-focused learning: warring-party elites deliberate under uncertainty about inclusion’s costs, reasoning over which precedents are representative rather than weighing every case exhaustively. Cross-border ties function as information channels, but they differ in what they carry. Social ties show whether inclusion proved survivable for those who chose it, leading elites to update positively from successful precedents and negatively from failed ones; the same ties carry advocacy models and strategic knowledge to the non-warring parties pressing for a seat. Economic ties expose warring parties to the consequences of a partner’s peace process: where countries depend on one another for economic welfare, a partner’s instability is a material cost, so warring parties observe that partner’s negotiations closely and update on whether they are followed by stability or renewed disruption. To capture these channels, I construct two original composite indices of the inter-state network: a social tie index and an economic tie index. I test the argument with Stochastic Actor-Oriented Models, using the Peace Observatory dataset on non-warring actor inclusion, linked to UCDP conflict data and to the bilateral source datasets underpinning each index, across all intrastate armed conflicts, 1989–2023. I find that government–rebel dyads whose socially tied conflicts already negotiate inclusively are substantially more likely to adopt inclusive negotiation themselves. This effect holds across alternative specifications of controls, rival explanations, weighting, functional form and sample, and it is not reproduced when the exposure network is rebuilt on geographic proximity or shared mediators, which are the channels existing work most often assumes. Economic ties also tend to increase the likelihood of inclusive negotiations, although this effect is less consistent across specifications. The findings carry implications for diffusion theory, the determinants of inclusive peace processes, and the design of mediation strategy in a densely connected international system.

Who Governs After Disaster? Political Exclusion and Ethnic Rebel Service Provision

with Elisa D’Amico

Near submission

Why do natural disasters increase rebel governance provision in some conflicts but not others? This paper argues that political exclusion of ethnic groups creates both the motivation for rebellion and strategic advantages for service provision, and that disasters function as legitimacy tests that reveal state discrimination. When states respond poorly to shocks affecting excluded groups, rebels drawn from those groups gain a window to demonstrate superior responsiveness to their constituencies. A formal model yields two hypotheses: excluded ethnic rebels provide more services than non-excluded ethnic rebels (H1), and disasters amplify this difference (H2). We test these predictions using rebel-year data (1982 to 2011) combining EPR political status measures, ACD2EPR rebel linkages, UCDP conflict data, EMDAT disaster records, and RQSI indicators of rebel social service provision. With actor and year fixed effects and two-way clustered standard errors, political exclusion alone shows a positive but only marginally significant effect. Its interaction with major floods is robust: excluded ethnic groups increase social aid provision by 7.5 percentage points during floods, a result stable across control sets, clustering choices, and alternative disaster measures. Non-excluded ethnic groups show no amplification. Discriminatory disaster response thus inadvertently strengthens rebel legitimacy, making inclusive governance strategically as well as normatively important.

Looking Next Door: Spatial Diffusion and the Thematic Content of Civil War Peace Agreements

Work in progress

Presented at the PSAI Annual Conference (2026)

Recent scholarship demonstrates that peace agreement content is likely to influence post-conflict institutions, the durability of peace and the likelihood of war recurrence. However, far less attention has been paid to explaining why the agreements differ in the substantive priorities they place on specific policy areas. I address this gap by advancing a policy diffusion and cognitive heuristic theory across intrastate conflicts. I argue that, when uncertain about which topics to prioritise, warring parties use recent agreements in geographically proximate countries as reference points for their proposals and drafting choices, because such agreements are more accessible and, through the representativeness heuristic, appear more relevant to their own circumstances. Nearby precedents should carry more weight for institutional topics, which more often offer identifiable arrangements that negotiators perceive as adaptable to their own setting. Using 421 framework-substantive peace agreements from PA-X covering intrastate conflicts between 1990 and 2025, I measure topic prevalence with a structural topic model fitted to chunks of about 200 words from the corpus, and capture spatial diffusion through a time-decay-weighted spatial lag with a 500 km buffer and a five-year half-life. OLS estimates show that the spatial lag is positively associated with topic prevalence for four of the 18 substantive topics, three of which are institutional, and a random-effects meta-regression is consistent with a stronger association for institutional than for non-institutional topics, at the 10 per cent level. All four associations remain positive across several robustness checks. The evidence for a stronger institutional association is sensitive to how topics are classified, how neighbours are defined and how the comparison is estimated. The findings suggest that peace agreement content is linked to proximate precedent as well as to domestic conditions, and they carry implications for the study of peace agreement design, policy diffusion theory and how peace practitioners target capacity-building and drafting support.

The Power Puzzle: Cultural Autonomy, Economic Shocks and Ethnic Power-Sharing

MA dissertation, International Relations, Seoul National University (2025).

Why do some countries share power amongst their ethnic groups whilst others do not? This study proposes two theoretical mechanisms to explain variation in ethnic power-sharing outcomes. First, cultural autonomy arguably creates a pathway to broader political power-sharing through a four-stage process: institutional protection, priority evolution, organisational transformation and political mobilisation. When excluded ethnic groups secure formal protections for their cultural practices, they establish a foundation of cultural security that enables a shift in priorities from basic cultural survival to broader political autonomy. The cultivation of cultural autonomy is expected to enable them to build upon their existing capacity by transforming into ethnic-based political parties, thereby enhancing their institutional and financial capacity for sustaining and maximising their political mobilisations. This increases the costs of non-accommodation for ruling groups and thus making power-sharing more likely. Second, economic shocks are likely to influence ethnic power-sharing by impacting ruling groups’ cost-benefit calculations and excluded groups’ mobilisation capacities. Positive economic shocks theoretically make power-sharing more likely as the increased absolute resources from these shocks are expected to allow the ruling groups to accommodate power-sharing with lower opportunity costs. Conversely, negative economic shocks decrease the likelihood of ethnic power-sharing, as ruling groups may need to consolidate over their diminishing absolute resources to maintain private rewards for their winning coalition at previous levels, and to preserve their capacity to suppress regime challenges through strategic concessions or repression.

Using data drawn from the Ethnic Power Relations dataset and logistic regression, I find that cultural autonomy associates positively with ethnic power-sharing, and economic shocks demonstrate a strong correlation with ethnic power-sharing, with positive shocks increasing the likelihood and negative shocks decreasing such likelihood. This thesis introduces cultural autonomy as a potential theoretical pathway to ethnic political power-sharing, addressing a gap in scholarship focused primarily on political dimensions. This thesis also provides novel theoretical frameworks which advance our understanding of how economic shocks influence power-sharing dynamics through ruling groups’ cost-benefit calculations and excluded groups’ mobilisation capabilities. Furthermore, this thesis demonstrates that incremental ethnic power-sharing (beginning with cultural autonomy and progressing toward political inclusion) is more effective than immediate full power-sharing. Economic stability plays a critical role in this process: negative shocks reduce the likelihood of power-sharing by 34.5%, whilst positive shocks increase it by 45.6%. These findings are particularly relevant in light of the recent $60 billion in U.S. aid cuts, which risk undermining prospects for sustainable ethnic power-sharing.

Presentations

2027

International Studies Association (ISA) Annual Convention (accepted)
2027

Development Studies Association Ireland (DSAI) Annual Conference (accepted)
2026

Political Studies Association of Ireland (PSAI) Annual Conference

Looking Next Door: The Spatial Diffusion of Inclusion Provisions in Civil War Peace Agreements
2026

Development Studies Association (DSA) Annual Conference
2026

Violence, Instability, and Peace (VIP) Workshop, online, 7 October

Spreading Inclusivity: Network Channels and Inclusive Peace Negotiations in Civil Wars; discussant: Sloan Lansdale (University of Maryland)
2026

Junior Scholars in Quantitative Conflict (JQCon) Workshop, online, 2 October

Spreading Inclusivity: Network Channels and Inclusive Peace Negotiations in Civil Wars; discussant: J Andrés Gannon (Vanderbilt University)
2026

SPIRe PhD Workshop, University College Dublin, Ireland, 28 April

discussant: Samuel Brazys (University College Dublin)
2026

UCD Centre for Peace and Conflict Research (CPCR) Talk, Invited Speaker, University College Dublin, Ireland, 14 April

Spreading Inclusivity: Network Channels and Inclusive Peace Negotiations in Civil Wars