Speaker
Description
Introduction
We investigate whether agricultural commodity booms can lead to social conflict. The economic literature suggests that commodity booms reduce conflicts by creating employment, raising incomes, and hence increasing the opportunity cost of violence. Yet, simultaneously, stronger incentives to contest the distribution of this additional income may lead to a rising potential for conflict. This "rapacity effect" has been observed in booms in capital-intensive commodities like oil and minerals, but less so in agricultural windfalls. Our paper introduces a unique third mechanism triggered when a commodity boom causes land-use transformation—this process itself can become a primary cause of social conflict when rising incentives for land acquisition and agricultural expansion clash with ambiguous legal frameworks and non-inclusive economic institutions. In such a setting, inequalities in land rent benefits or favoritism towards certain groups can lead to social unrest and conflict, as marginalized groups resort to violence or disruption to express their grievances or claim the share of resources they feel entitled to. Our study fills a gap in the empirical literature in economics that has overlooked the role of social conflict stemming from land-use changes.
We conduct our analysis in the context of Indonesia’s oil palm boom, where from the 1990s onwards palm oil production expanded rapidly in response to rising global demand. While this boom contributed to a substantial improvement in rural livelihoods, it has also increasingly been associated with social conflicts, particularly violent land disputes. We argue that these conflicts are rooted partly in Indonesia’s institutional land governance framework. In the Indonesian context, where political representatives face strong rent-seeking incentives and the effective exploitation rights for land are biased, economic incentives for land conversion can lead to conflicts related to the distribution of economic benefits and to violent grievances over local representation.
Method
To study this issue, we link highly detailed data on local conflict outbreaks between 2005 and 2014 sourced from newspaper reports by Indonesia’s National Violence Monitoring System (NVMS) to incentives for oil palm expansion at the level of 2,755 rural sub-districts in Indonesia. Our empirical models control both for sub-district-fixed effects capturing unobserved local heterogeneities, and for common time shocks through year fixed effects. Our main explanatory variable captures national-level variation in incentives for oil palm expansion, which we bring to the local level via a suitability-based exposure measure that reflects the relative size and potential profitability of the agricultural area that is not yet used for oil palm. Our identification strategy thus relies on a one-dimensional shift-share measure. We further control for differential trends in remoteness, to account for potential reporting biases in the conflict data. We further supplement our analysis with a wide variety of secondary data sourced from satellite data and from Indonesia’s PODES village census.
Findings
We find that stronger economic pressure to expand oil palm plantations increases the incidence of violent conflicts at the local level, especially in the form of resource conflicts and conflicts involving popular justice, as well as violence around local elections. Importantly, this mechanism of conflict induced by land-use change is also distinct from income shocks in producing areas. While we find that social conflict increases with negative income shocks (negative price shocks or localized drought shocks), these variables do not change the effects of land-use change when estimated in a joint framework.
At the local level, the adverse impacts of expansion incentives are very heterogeneous. We document three major sets of results that are in line with theoretical predictions and qualitative evidence: First, social conflict related to land expansion pressure is closely related to scarcity. Adverse impacts are stronger where less land is available and where other sources of income are lacking, suggesting that competition over increasingly scarce resources could be an important driver of violence in this setting. We interpret this finding as suggestive evidence that alternative income sources in rural areas could mitigate competition for agricultural land by reducing its relative importance and hence perceived feelings of scarcity.
Second, we find that initial land distribution plays an important role. Locations with a higher share of communally-owned land experience a stronger increase in conflicts due to expansion pressure. This finding is in line with qualitative and descriptive research arguing that local elite-control over village lands had very adverse effects on social cohesion during the oil palm boom, and that resource conflicts over communal lands are frequent.
Finally, we find that expansion pressure is closely linked to election conflicts. Whenever incentives for land conversion are stronger, local elections and official appointments are surrounded by more outbreaks of violence. This is driven by elections at the village and district level, where most decisions about land-use are made in Indonesia.
Overall, our findings align with assessments by various qualitative and anecdotal sources stating that the rise in reported land disputes in Indonesia is primarily driven by imbalanced ownership and disputes over the economic potential of these lands, rather than cultural, social, or environmental factors. We also test some of these competing explanations and do not find evidence for environmental, ethnic, or migration-related disputes in our context.
Conclusion
We document that the oil palm boom increased conflicts in Indonesia despite its positive effects on income and employment. We provide evidence that this effect is driven by economic motives, distributional grievances, and an institutional setting that has historically favored local and political elites. We believe that our findings are not limited to Indonesia and might also apply in other contexts where rapid land conversion occurs. They are especially relevant in light of the fact that palm oil production in other regions of the world, such as tropical Africa and Latin America, is expected to increase significantly in the next few decades. This could lead to similar problems to those identified in the Indonesian context. Overall, our findings thus underline the importance of Indonesia’s ongoing land reform efforts and the necessity of rural land transformation to go hand in hand with conflict mitigation strategies.
