Skip to main content

Climate Variability and Security Threats

Periodic Reporting for period 3 - CLIMSEC (Climate Variability and Security Threats)

Reporting period: 2018-09-01 to 2020-02-29

Recent uprisings across the world have accentuated claims that food insecurity is an important trigger of political violence. Is the Arab Spring and more recent comments on the environmental origins of the Syrian civil war representative of a general climate-conflict pattern, where severe droughts and other climate anomalies are a key driving force? Research to date has failed to conclude on a robust relationship but several notable theoretical and methodological shortcomings limit inference. CLIMSEC tackles these research challenges head on. It asks: How does climate variability affect dynamics of political violence?

The overarching research question will be addressed through the accomplishment of four key objectives: (1) Investigate how food security impacts of climate variability affect political violence; (2) Investigate how economic impacts of climate variability affect political violence; (3) Conduct short-term forecasts of political violence in response to food and economic shocks; and (4) Develop a comprehensive, testable theoretical model of security implications of climate variability.

To achieve these objectives, CLIMSEC advances the research frontier on theoretical as well as analytical accounts. Central in this endeavor is conceptual and empirical disaggregation. Instead of assuming states and calendar years as unitary and fixed entities, the project seeks to identify and evaluate plausible causal processes that act at multiple temporal and spatial scales, involve various types of actors, and potentially lead to very different forms of outcomes depending on the context. The empirical component makes innovative use of new geo-referenced data and methods; focuses on a broad range of insecurity outcomes, including non-violent resistance; and combines rigorous statistical models with out-of-sample simulations and qualitative case studies for theorizing and validation of key findings. Based at the Peace Research Institute Oslo (PRIO), the project is led by Research Professor Halvard Buhaug, a leading scholar on climate change and security.
Work since the commencement of the project can be summarized as follows:
* A PhD project (candidate: Elisabeth Rosvold) investigating the conditions under which climatological disasters and extreme weather events affect the risk and dynamics of political instability, and the extent to which relief and development aid can strengthen local coping capacity and minimize the risk and impact of armed conflict. Several working papers are in progress and are expected to be published in prominent international peer-reviewed journals within the next year. This PhD project involves an ambitious plan to geocode more than 2,000 disaster events in the EM-Dat database (a global database of natural disasters and emergencies managed by the Centre for Research on the Epidemiology of Disasters (CRED) in Belgium). This data collection effort opens up new opportunities for more detailed and local quantitative assessments of disaster-security dynamics, so we decided to expand the original plan and grant the PhD candidate more time to make the end product considerably greater in spatial and temporal scope.

* A major update and expansion of the PRIO-GRID spatial data framework, including a user-friendly data portal and visualization tool. To this end, CLIMSEC hired a Research Assistant (Karim Bahgat) with expert knowledge on spatial data management and geographic databases. The new PRIO-GRID v. 2.0 was released in November 2015 and has since been maintained with financial support from this project. This data framework is a major achievement and an important contribution to the community of scholars conducting advanced quantitative analyses of armed conflict with spatial data. Alongside the gridded data, which come as yearly files as well as a static representation, users may also download scripts to regenerate the grid structure underlying the database. In association with the new v.2.0 we also developed and launched a new dynamic web portal ( where users may visually inspect and access all variables contained in the database, as well as all background documentation, and programming scripts to replicate the grid structure.

* A major update and expansion of the Urban Social Disorder (USD) dataset. The new USD v. 2.0 was released in June 2017. This constitutes the only available catalogue of event-specific information on political disturbance events (strikes, protests, riots, terrorist attacks) in national capitals and other major cities, now expanded from Africa and Asia to also cover the Middle East and Central and South America for all years, 1960-2016. We are now looking into options to develop a future v.3 that additionally will cover major cities in Eastern and Western Europe, thus making it a truly global database of urban disorder events.

* During the first reporting period, the project has resulted in five peer-reviewed articles in leading international journals as well as three chapters in anthologies and three reports. This work has shown that climatic parameters exert a weak and inconsistent influence on local conflict risk although there is evidence to suggest that adverse climatic changes affect conflict dynamics, notably by making ongoing conflicts more challenging to end. Moreover, early publications from the project have demonstrated that the climate effect is not universal but likely conditional on various configurations of political and socioeconomic contexts. For example, the study by von Uexkull et al. (2016), which represents the first subnational comparative assessment of livelihood-conflict dynamics, shows that drought and resultant income loss from agriculture affects conflict incidence primarily among agriculturally dependent and politically marginalized groups of society.

* Several project-related activities geared toward policy have taken place. The PI is an invited participant in the European Commission / Joint Research Centre’s effort to improve their Global Conflict Risk Index and early warning system. The PI also organized an internal seminar at the UN Food and Agriculture Organization (FAO) in Rome in November 2016 to present and discuss food security – conflict linkages, and a similar dialogue has been established between the project team and the UN World Food Programme (WFP). We anticipate future meetings and possible collaboration with FAO and WFP as the project progresses. Moreover, the PI was recently selected to contribute to the development of the UN Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6) as Lead Author for Working Group II; Chapter 16. This is a remarkable recognition of his competence and academic standing in the field of climate security. Buhaug is the only conflict scholar in the IPCC team of authors.
* We have also set up a specific blog for popular dissemination of findings and knowledge on climate and conflict: In addition to regular publications by project members, the blog is also open to outsiders. Among the list of prominent guest contributors so far, we have David Beasley, the executive director of the UN’s World Food Programme:
By definition, the new datasets released as distinct deliverables of the CLIMSEC project, the PRIO-GRID 2.0 framework and the USD 2.0 dataset, provide new opportunities for research and facilitate the combination and analysis of geo-referenced data in a rigorous and space/time-consistent fashion. These products clearly redefined the state-of-the-art in spatial data framework and systematic data on urban political violence.

The Hegre et al. study (Environmental Research Letters 2016) presented the first attempt to simulate security implications of alternative future climate change and socioeconomic changes. This is an important contribution to a field that thus far has made little progress in quantifying direct social implications of different climate change scenarios. We are now working on follow-up studies that seek to address obvious endogenous links between socioeconomic development, climate change adaptation and mitigation, and conflict risk, which so far have been ignored.

Also, the study by von Uexkull et al. (PNAS 2016) represented a clear step beyond the research frontier by presenting the first group-specific spatial analysis of how growing-season drought in each group’s agricultural areas influence conflict risk and occurrence, explicitly modeling moderating roles of level of agricultural dependence and political context. We are now working on a follow-up study that both expands the data material in space (now global) and time, and, importantly, presents a more elaborate theoretical model of how climate-induced income shocks may alter incentives for social conflict under various conditions.

Moving forward, we plan to extent the scientific scope to invest even more in disentangling complex indirect associations between climatic changes and conflict dynamics. This work is likely to include investigations into how violent conflict increases environmental vulnerability as well as material and human costs of subsequent climatic extremes. Moreover, as outlined in the original project description, we will develop and assess the merit of incorporating climatic parameters and potentially sensitive processes (e.g. food price fluctuations) in conflict early warning models. The PI is in dialogue with multiple potential user groups (EC Joint Research Centre; FAO; WFP; USAID) who all have expressed a clear need for better early warning models as well as evidence-based guidance on what works in different contexts. Likewise, the PI’s role in the development of the IPCC AR6 will ensure that relevant research on security implications of climate change will inform policy and the general public.