Land-Grabbing in Civil Conflicts: Non-State Actors and Property in Colombia

• In Event: Causes and Consequences of Conflict: Evidence from Colombia

Abstract

Land-Grabbing in Civil Conflicts:
Non-State Actors and Property in Colombia

In Colombia and other countries mired in complex civil conflicts, land theft has played an important role in exacerbating the impacts of these conflicts on civilian populations and extending the lengths of conflicts since armed actors can use abandoned and stolen land to settle supporters and enrich their cadres and allies. Despite our rapidly growing knowledge about civil conflict, we know relatively little about the conflict dynamics that drive land theft. Through an analysis of armed actor operations and territorial control in Colombia this paper seeks to understand what types of actors, including paramilitaries, guerrillas, and criminal organizations, engage in land grabbing and under what economic and conflict dynamics these groups are most likely to engage in land theft. Our preliminary results show that paramilitaries are most likely engage in land-grabbing and that land-grabbing emerges in regions that concentrate particular types of agricultural production.
Land has long been at the heart of Colombia’s 70 year-long civil conflict. In the 1930s Colombia pursued a broad social-inclusion policy known as the Revolución en Marcha. This effort led to the extension of social rights to urban workers but foundered on efforts to provide property to peasants. The fear of Jorge Eliecer Gaitain assassination in 1948, a pro-reform politician, started a 10-year civil conflict known as La Violencia (The Violence) that left 200,000 dead. A power sharing agreement ended this phase of the conflict but some rebellious Liberal-leaning guerrillas refused to lay down their weapons and officially reorganized as the Fuerzas Armadas Revolucionarias de Colombia (FARC) in 1964. Property rights have suffered as land was stolen when internally displaced populations migrated to save their lives. Colombians have labeled this process land-grabbing.
Land-grabbing in Colombia follows two models. The first is open land theft, referred to as despojo, where families holding particular parcels of land are targeted by armed actors for removal in order to take their property. The second is desplazamiento (displacement) where a family leaves their land because of the fear of conflict. These abandoned lands may be later distributed by armed actors.
Over this time guerrillas and other groups, including powerful landowners, businesses, and right-wing paramilitary groups have pursued land-grabbing for slightly different reasons. The guerrillas do so partly for ideological reasons seeking to promote land reform in the areas they control and also to help settle ideologically-allied peasants displaced from other regions. The paramilitaries seek to displace leftist peasants and seek to take control of the land for themselves and their allies, large landowners and business people that have provided them with economic support. Criminals may also pursue land-grabbing to promote coca growing.
We use two types of data in this paper. The first is an original data set of armed actor control across Colombia constructed through a computerized identification of actors and locations using Eventus ID building on the Noche y Niebla database which has geo-coded information about municipal-level armed actor presence over the past 30 years. Second, we use secondary longitudinal data sources developed in Colombia. Information about land-grabbing comes from the institutional dataset of the Land Restitution Agency. Data on the fluctuation of palm oil plantations comes from the Ministry of Agriculture’s Statistical Yearbook. We gathered international palm oil data from the International Monetary Fund and data for other control variables from the Municipal Panel Data set constructed by the Center of Economic Development Studies (CEDE) at Universidad de Los Andes.
Our preliminary findings show a relationship between armed actor presence and land seizure. Our data shows that both guerrilla and paramilitary presence are correlated with land-grabbing though paramilitary activity is more predictive of land-grabbing. The presence of purely criminal organizations is not associated with land-grabbing. Our analysis also shows that land-grabbing is associated with particular types of economic activities. Both coffee production and oil production appear to be associated with lower rates of land-grabbing while palm oil production is associated with higher rates of land grabbing. This would statistically confirm the numerous descriptive accounts of individual cases where palm oil cultivation was motivated by the actions of illegal armed groups and acted as a driver for illegal land seizure in the country. Coca cultivation did not have significant effects.

Authors

• Camilo Pardo, George Mason University

• Enrique Desmond Arias, Baruch College, CUNY