Revisiting the Contested Role of Natural Resources in Violent Conflict Risk through Machine Learning

Marie K. Schellens and Salim Belyazid
Additional contact information
Marie K. Schellens: Department of Physical Geography, Stockholm University, 114 19 Stockholm, Sweden
Salim Belyazid: Department of Physical Geography, Stockholm University, 114 19 Stockholm, Sweden

Sustainability, 2020, vol. 12, issue 16, 1-29

Abstract: The integrated character of the sustainable development goals in Agenda 2030, as well as research in environmental security, flag that sustainable peace requires sustainable and conflict-sensitive natural resource use. The precise relationship between the risk for violent conflict and natural resources remains contested because of the interplay with socio-economic variables. This paper aims to improve the understanding of natural resources’ role in the risk of violent conflicts by accounting for complex interactions with socio-economic conditions. Conflict data was analysed with machine learning techniques, which can account for complex patterns, such as variable interactions. More commonly used logistic regression models are compared with neural network models and random forest models. The results indicate that a country’s natural resource features are important predictors of its risk for violent conflict and that they interact with socio-economic conditions. Based on these empirical results and the existing literature, we interpret that natural resources can be root causes of violent intrastate conflict, and that signals from natural resources leading to conflict risk are reflected in and influenced by interacting socio-economic conditions. More specifically, the results show that variables such as access to water and food security are important predictors of conflict, while resource rents and oil and ore exports are relatively less important than other natural resource variables, contrasting what prior research has suggested. Given the potential of natural resource features to act as an early warning for violent conflict, we argue that natural resources should be included in conflict risk models for conflict prevention.

Keywords: conflict prediction; environmental conflict; logistic regression; machine learning; natural resources; natural resource conflict; neural network; random forest; sustainable peace (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2020
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.mdpi.com/2071-1050/12/16/6574/pdf (application/pdf)
https://www.mdpi.com/2071-1050/12/16/6574/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:12:y:2020:i:16:p:6574-:d:398836

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().