Rising crop prices usually quiet conflict, as labourers choose farming over fighting. In Myanmar, rising rice prices instead fuelled state-led violence against civilians. Our findings challenge narratives that frame such atrocities as a reaction to insurgencies, showing how economic incentives can systematically shape persecution when power is heavily skewed toward the state.
Hundreds of thousands of Rohingya civilians were forced to flee Myanmar in 2017. When interviewed in refugee camps in Cox’s Bazar in southern Bangladesh, they described indiscriminate killings, sexual violence, and the burning of entire villages. The International Court of Justice described the violence as “genocidal in character”. Myanmar government officials, including Nobel Peace Prize laureate Aung San Suu Kyi, have denied systematic wrongdoing and repeatedly portrayed this violence as chaotic, localised, or a response to insurgency.
We assemble surveys of displaced Rohingya in Bangladesh, a database of conflict events across Myanmar, satellite images tracking crop suitability, and fluctuations in rice prices in international markets to analyse the economic drivers of conflict and persecution of civilians, both in Rohingya areas and across Myanmar as a whole (Davis, López-Peña, Mobarak, and Wen 2025). While the Rohingya, a Muslim minority stripped of citizenship since 1962, have been particularly targeted by the state, they are not the only affected group. Myanmar has a long, well-documented history of state violence against civilians that spans decades and affects multiple ethnic groups. Human rights organisations have documented state violence against Karen, Shan, and other ethnic minorities.
The global evidence on crop prices and conflict
For labour-intensive crops like rice or coffee, higher prices typically reduce two-sided conflict, e.g. fighting between armed groups. Higher agricultural wages raise the opportunity cost of fighting, pulling potential combatants into farming instead (Dube and Vargas 2013, Dal Bó and Dal Bó 2011). Recent work by McGuirk and Burke (2020) refines this point, showing rising food prices pacify violence in food-producing areas while sparking unrest in food-consuming regions.
State violence against civilians in Myanmar is a fundamentally different type of conflict: it is more accurately characterised as one-sided persecution of civilians by the state. We find that this distinction matters profoundly for how price shocks translate into conflict.
Violence and rice in Myanmar
Rice is essential as a source of food and livelihoods in Myanmar; the crop covers over 60% of the country’s cultivated area. But suitability for growing rice varies substantially across the country’s 330 townships, and domestic prices closely track international markets. Thus, the economic incentive to expropriate rice varies across space and over time.
Myanmar’s government claimed the 2017 violence was chaotic, unpredictable, and isolated. We tested this claim by asking whether violence against civilians followed a systematic pattern, both in Rohingya areas and in the country as a whole.
Identifying cause and effect
Establishing a causal link between economic conditions and violence is challenging. Violence itself can affect prices, for example, by disrupting agricultural production. Omitted factors like political developments could simultaneously drive changes in prices and violence. Observing that prices and violence move together is not enough to demonstrate that one causes the other.
Our approach addresses these challenges. The economic incentives created by changing rice prices apply most directly to places where rice is widely grown. We compared conflict rates between periods of high and low rice prices across townships with high versus low suitability for growing rice. Since conflict itself can change domestic prices, we used international prices instead as these are likely to be insulated from events in Myanmar, which accounts for less than 5% of global supply.
Higher prices led to more violence in rice-growing areas
When rice prices increased from the 25th to the 75th percentile, the likelihood of violence against civilians increased by 55% in townships with average rice suitability. State-perpetrated violence rose by 41%. Looting and property destruction surged by 196%, suggesting economic motives at play.
The economic logic was especially powerful in Rakhine State, where Rohingya are concentrated. Violence against civilians there responded more strongly to rice prices than in the rest of Myanmar. Violence targeting Rohingya civilians responded sharply to rice prices, while violence perpetrated by Rohingya militias showed no such pattern. This asymmetry is consistent with our interpretation: those with more power appropriate resources from vulnerable civilian populations.
Why one-sided violence reverses the standard pattern
Our findings point to an important gap in the evidence base on commodity prices and conflict: the balance of power between adversaries shapes how price shocks translate into violence. Existing work focuses on two-sided conflicts where both sides choose between economic activity and fighting. But in one-sided violence, the aggressor does not face a meaningful trade-off between farming and fighting; they can simply take what others produce. Higher crop values therefore raise the returns to expropriation much more than the opportunity cost of violence. This is consistent with Mitra and Ray’s (2014) theory, which predicts that when a much stronger group faces a weaker one, increases in the weaker group’s resources can place them at greater risk.
This interpretation complements ethnic and political explanations. Historical and cultural factors, including the stripping of Rohingya citizenship and systematic disenfranchisement, were prerequisites that made civilians vulnerable to state forces. But economic forces help explain when and where violence intensified.
What refugees lost
If expropriation helped motivate violence, then people who fled should have lost agricultural assets. We conducted a representative survey of Rohingya refugee and host community households in Cox’s Bazar, Bangladesh, where over 900,000 refugees now live. The survey, called the Cox’s Bazar Panel Survey, included 5,020 households in total, roughly half from the refugee community and half from the host community (Davis, López-Peña, Mobarak, and Wen 2024).
Among refugee households displaced in 2017, the losses were severe: 79% lost agricultural land, 60% lost draft animals, 56% lost agricultural equipment, and 34% reported losing stored crop inventory, direct evidence of stolen agricultural assets.
Myanmar’s government openly confirmed expropriation of Rohingya agricultural resources. State media reported harvesting over 70,000 acres of “abandoned” Rohingya farmland, generating US$5.3 million in revenue “deposited in a bank account as a part of the national budget” (Global New Light of Myanmar 2017). Government ministries sent tractors and combines to harvest. Land was ultimately claimed by the military and settled by non-Rohingya civilians.
This pattern extends beyond 2017 and beyond the Rohingya. Human rights groups document decades of military looting targeting ethnic Karen, Shan, and other minorities. One survey found 49 rice looting instances by state forces in just three years across seven districts (Hudson-Rodd et al. 2004).
Implications for policy
Our findings challenge how policymakers think about food price shocks and civilian safety. Rising food prices will not always pacify conflict in producer regions; the balance of power between armed forces and civilian populations matters critically. In the existing evidence base, price increases for capital-intensive resources like oil and minerals tend to increase conflict because they raise the spoils of war (Collier and Hoeffler 2004, Ross 2004). Our results suggest that power asymmetries can make a labour-intensive crop like rice behave the same way, because the aggressor appropriates the crop but may not face the opportunity cost of labour.
The risk of price-driven violence against civilians is likely to be highest where multiple risk factors overlap, including weak property rights, state forces that face few constraints on the use of violence, and pre-existing marginalisation of specific groups. These are not the only relevant factors, but their combination with food price volatility exacerbates the risk of violent expropriation.
Policymakers and international organisations working in these contexts should consider that rising food prices may signal increased danger for at-risk civilian populations, particularly where marginalised groups reap valuable agricultural land. Economic monitoring should be integrated with political risk assessment. And breaking the cycle ultimately requires strengthening property rights for vulnerable groups, international monitoring of resource appropriation during conflicts, and accountability for expropriation.
As climate change and geopolitical instability fuel food price volatility, it is essential to recognise that the same shock that pacifies one conflict could inflame another.
References
Collier, P, and A Hoeffler (2004), “Greed and grievance in civil war,” Oxford Economic Papers, 56(4): 563–595.
Dal Bó, E, and P Dal Bó (2011), “Workers, warriors, and criminals: Social conflict in general equilibrium,” Journal of the European Economic Association, 9(4): 646–677.
Davis, C A, P López-Peña, A M Mobarak, and J Y Wen (2024), “Refugees are hosted in highly vulnerable communities,” AEA Papers and Proceedings, 114: 75–79.
Davis, C A, P López-Peña, A M Mobarak, and J Y Wen (2026), “The economic drivers of state violence against civilians: Evidence from Myanmar,” Economic Journal, ueaf098.
Dube, O, and J Vargas (2013), “Commodity price shocks and civil conflict: Evidence from Colombia,” Review of Economic Studies, 80(4): 1384–1421.
Global New Light of Myanmar (2017), “Sales from Maungtaw paddy kept as national budget,” Global New Light of Myanmar.
Hudson-Rodd, N, M Nyunt, S Thamain Tun, and S Htay (2004), “State induced violence and poverty in Burma,” Federation of Trade Unions-Burma.
McGuirk, E, and M Burke (2020), “The economic origins of conflict in Africa,” Journal of Political Economy, 128(10): 3940–3997.
Mitra, A, and D Ray (2014), “Implications of an economic theory of conflict: Hindu-Muslim violence in India,” Journal of Political Economy, 122(4): 719–765.
Ross, M L (2004), “How do natural resources influence civil war? Evidence from thirteen cases,” International Organization, 58(1): 35–67.