How weak property rights cause land conflict and deforestation in Brazil

Agricultural - Natural Resource Economics
Conflict, Crime and Violence

The combination of weak institutions and natural resource abundance are a recipe for civil conflict and the overexploitation of natural resources. However, because institutions often come as a package, what we know about the type of institutional change required to resolve these problems is limited. This column presents evidence from Brazil indicating that insecure property rights are a major cause of land related conflict and a contributing factor in Amazon deforestation.

There is a prevailing view that natural resource abundance paired with weak institutions breeds civil conflict (Collier and Hoeffler, 2004). Further, weak institutions may foster the unchecked exploitation and erosion of natural resources (Mehlum et al., 2006). Yet, there is a limited understanding of what type of institutional change is required to reduce civil conflict and unsustainable natural resource exploitation.

Institutions can distinctly affect conflict by directly and indirectly shaping the underlying incentives. Most existing empirical work has focused on the important role played by the value of agents’ outside option relative to the contested prize, and the institutions and policies that ameliorate these effects (e.g. Dube and Vargas, 2013, Fetzer, 2013).

In a recent paper (Fetzer and Marden, 2016), we show that the assignment of secure property rights can dramatically reduce civil conflict, even in the absence of changes in enforcement. Indeed, at the local level, we cannot rule out that substantively all violent land related conflict is a consequence of Brazil’s failure to securely assign property rights over land.

Our study explores land related civil conflict in the Brazil’s Amazon states. As in many developing countries, Brazil has high levels of land related conflict, with at least 280 murders,  and many my more lesser events, occurring between 1997 and 2010 (see figure 1). Also as elsewhere,  the historically unequal distribution of land ownership in Brazil has lead to the adoption of policies and institutions that favour redistribution and land reform.  As such contestability of property rights over land is enshrined into the Brazilian constitution. Land not in ‘productive use’ is vulnerable to invasion by squatters, who can develop the land and appeal to the government for title.


We obtain variation in the share of municipal level that is contestable land by exploiting the dramatic expansion in the share of land under ecological or indigenous protection between 1997 and 2010 (see figure 2). Protected land automatically fulfils productive use requirement by virtue of being protected, and is thus no longer contestable even in the absence of economic activity. Consequently, an increase in the municipal share of land under protection reduces the share of land in that municipality that is contestable.


Using this variation in the municipal share of land with contestable title in the Brazilian Amazon, we show that (in)secure property rights are strongly associated with land related conflict. We obtain these results using classical panel data regression methods, but also employ a novel quasi-event study approach. In the latter,  the identifying variation comes only from comparing changes in violence in neighbouring municipalities that are treated by the same protected area. This relaxes the identification assumption and allows us to test for, and confirm, the presence of parallel trends in conflict prior  to treatment. Regardless of specification, our results indicate that weak property rights, and the resulting contestability of land title, is a primary cause of this conflict. Indeed, we cannot reject the hypothesis that, at the municipal level, substantively all land related violence is due to insecure titling. However, our results also indicate that securing property rights in one location in part reduces land conflict by diverting it elsewhere.

No evidence suggesting improved enforcement

While land related conflict is strongly predicted by the availability of contestable land, there is no detectable relationship with non-land related violence (proxied by homicides), economic activity or environmental enforcement. The changes in land conflict do not appear to have been a byproduct of other changes that might have accompanied the expansion of protected areas.

Deforestation as primary mechanism to contest title

Deforestation, and particularly permanent deforestation and agricultural conversion, is a key input into land conflict as land clearing is necessary to signal productive use. This threatens the Amazon as the last remaining ecosystems not subject to anthropogenic changes.

Using high resolution land cover data, we verify that protection is associated with a decrease in deforestation.  However, not all types of deforestation are affected equally.  While “permanent deforestation” of the type necessary to establishing a title decreases substantially, short run deforestation of the kind associated with illegal logging and temporary pasture actually increases.  The overall net effect suggests a reduction in deforestation, which is in line with previous studies that have found the bulk of deforestation being due to agricultural conversion (Morton et al., 2004). These results indicate that protection may have been successful in reducing deforestation, despite weak enforcement, by removing a path to title.


Our paper highlights the importance of establishing secure property rights over land, and more generally the role of law in order in limiting the set of contestable assets and goods, in preventing civil conflict.  The presence of spillovers suggest that there may be negative externalities associated with local efforts to improve property rights, and so national or regional efforts to improve property rights are likely to be more beneficial than more local approaches. These findings complement those of an existing literature which has emphasised the economic costs of property rights (e.g. Besley and Ghatak, 2010) and vividly illustrate what “guard labour” can really entail.


Besley, T., & Ghatak, M. (2010). Property Rights and Economic Development. In D. Rodrik & M. Rosenzweig (Eds.), Handbook of Development Economics (Vol. V).

Collier, P., & Hoeffler, A. (2004). Greed and grievance in civil war. Oxford Economic Papers, 56(4), 563–595.

Dube, O., & Vargas, J. F. (2013). Commodity Price Shocks and Civil Conflict: Evidence from Colombia. The Review of Economic Studies, 80(4), 1384–1421.

Espírito-Santo, F. D. B., Gloor, M., Keller, M., Malhi, Y., Saatchi, S., Nelson, B., … Phillips, O. L. (2014). Size and frequency of natural forest disturbances and the Amazon forest carbon balance. Nature Communications, 5, 3434.

Fetzer, T. R. (2014). Social Insurance and Conflict: Evidence from India. Mimeo.

Fetzer, T. R., & Marden, S. (2016). Take what you can: property rights, contestability and conflict. Forthcoming in Economic Journal.

Morton, D. C., DeFries, R. S., Shimabukuro, Y. E., Anderson, L. O., Arai, E., del Bon Espirito-Santo, F., … Morisette, J. (2006). Cropland expansion changes deforestation dynamics in the southern Brazilian Amazon. Proceedings of the National Academy of Sciences of the United States of America, 103(39), 14637–14641.

Share this