Responses to JPR Climate and Conflict Special Issue: John O’Loughlin, Andrew M. Linke, Frank Witmer (University of Colorado, Boulder)

Guest Contributor

Responses to JPR Climate and Conflict Special Issue: John O’Loughlin, Andrew M. Linke, Frank Witmer (University of Colorado, Boulder)

April 12, 2012 By John O’Loughlin, Andrew Linke & Frank Witmer

The Journal of Peace Research recently devoted a special issue to the work of researchers studying the linkages between climate change and conflict. Special guest editor Nils Petter Gleditsch introduces the issue here.

Complexity, in terms of economic, cultural, institutional, and ecological characteristics, weighs heavily on contemporary attempts to unravel the climate change/variability and conflict nexus. The view that local-level complexity can be “controlled away” by technical fixes or adding variables in quantitative analysis does not sit well with many geographers (though some do try to adopt a middle ground position).

Geographers are typically hesitant to make broad claims and sweeping generalizations across varied temporal and spatial spans; we stress the role of context in shaping particular outcomes. In an effort that most human and political geographers should certainly appreciate, many of the studies featured in this special issue of the Journal of Peace Research illustrate a determination to move toward finer resolution analysis than the nation-state scale. Further, when an association is uncovered, it is accompanied by silent caveats. These by no means nullify the findings of the respective articles, instead they suggest caution of simplistic interpretations in any direction (positive, negative, or null).

Recognizing the effects that varying definitions and regional foci have on conclusions that might be drawn from the cumulative body of research is important.

For example, conflict data definitions vary within the special issue – relationships are found for “social conflict” (Saleyhan and Hendrix) and when a distinction is made between “rebel” and “communal” conflicts (Raleigh and Kniveton). Some of the research includes analyses of civil war events, which is not necessarily the type of conflict that one would expect to see in the wake of ecological change.

Additionally, regardless of a study’s individual contributions, there will be concerns about external validity of single-country case studies (e.g. Kenya or Mali).

It is also a little surprising that not a single climate-conflict analysis in the special issue is based on survey data. The extent to which individuals are involved in conflicts is a function of personal- and local-level circumstances. Survey analysis is used increasingly in post- and ongoing-war zones to probe conflict experiences, reconciliation attitudes, and adaptations (e.g. in the Russian North Caucasus).

Area data analysis – whether grid cells, towns or villages, or sub-national administrative units – is limited, but it is often the best option for aggregating geographic information systems data from multiple sources (media reported and crowd-sourced conflict event data, climate and census data, or remotely-sensed environmental imagery).

In this broader spectrum of climate and conflict studies, there is much leverage to be gained from data collection and innovative estimation models using surveys of individuals and households across a variety of geographic contexts.

The issue marks a sort of mid-term stock-taking of the current state of research on possible climate-conflict links but the field has a long way to progress in terms of regional coverage, disciplinary approaches, nuanced and detailed climate measures, public outreach, and policy recommendations.

Further responses from the environment and security community on the special issue can be found here. If you’re interested in weighing in please feel free in the comments below.

John O’Loughlin is a professor of geography at the Institute of Behavioral Science, University of Colorado, Boulder. Andrew M. Linke is a PhD student in geography and a research assistant and Frank Witmer is a post-doctoral research associate.

Topics: climate change, conflict, environment, environmental security, foreign policy, Guest Contributor, JPR Special Issue, Kenya, Mali, Russia, security
- Julien Katchinoff
  
  I'm always in support of research that doesn't try to avoid complexity, that calls for recognition of nuance and context, and — naturally — includes John O'. Thanks for this post! If possible, it would be great to linkt to an example of innovative area data analysis called for in the piece.
  
  Go Buffs!
- Tom Deligiannis
  
  In my view, one of the most remarkable statements made in this special issue came in Gleditsch's introduction, where he acknowledged the seeming inability of quantitative studies to deal with complex, interactive social-ecological systems. Commenting on the lessons that the large-N community could learn from the case study environment-conflict researchers Gleditsch wrote: Kofi Annan (2006: 9–10) argued in one of his last reports as UN Secretary-General, that ‘pollution, population growth and climate change are . . . occurring now and hitting the poorest and most vulnerable hardest. Environmental degradation has the potential to destabilize already conflict-prone regions, especially when compounded by inequitable access or politicization of access to scarce resources.’ Here, he is invoking an interaction effect of climate change with no less than three other variables. Unfortunately, it seems unlikely that case study researchers or large-N scholars will launch a systematic investigation of such complicated interaction patterns any time soon."
  
  If the large-N researchers are incapable of "systematic investigation" of climate change-security research with four variables, then this is a bankrupt research agenda. For case study researchers, even four variables seems remarkably restricted. Future studies have to address this issue more systematically. What are the barriers to designing more complex research studies with multiple variables? If climate modelers are using dozens of variables in their models (albeit, with variables of a different level of precision), then why can't it be done in social-ecological research? Also, is there not some way to bridge the gap between large-N studies and case study approaches – something that has been a constant theme in this field for the past decade? The findings of these large-N studies need to be ground-truthed, and case study approaches can help to do this, using survey and/or ethnographic approaches. A recent study in the right direction was done by USAID and International Alert – Marisa Goulden and Roger Few, Climate Change, Water and Conflict in the Niger River Basin.We should all applaud the efforts by large-N researchers to lower their analytical focus to the sub-national level. However, doing so also means accepting the sensible conclusions of the commentators above – that complexity can't be ignored. In fact, it's probably true that the lower the level of analysis, the greater the requirements that studies are sensitive to complex interactions across scales and between variables.
- Johno, Andrew, and Frank
  
  Tom,
  
  Thanks very much for your comment. We wanted to point you to Theisen et al.'s 2012 article in International Security titled "Climate
  Wars?: Assessing the Claim That Drought Breeds Conflict". In this article, the authors test dozens of interaction effects between drought/precipitation and ethnicity, population, poverty, and IMR.
  
  With institutional access, you should be able to view the article here:
  
  http://muse.jhu.edu/journals/international_security/toc/ins.36.3.html
  
  Note in particular figure 4:
  
  http://muse.jhu.edu/journals/international_security/v036/36.3.theisen_fig04.html
- Nils Petter Gleditsch
  
  Of course, statistical studies of the relationship between climate change and conflict do not hesitate to invoke four or more variables in their explanatory models. And as Johno, Andrew, and Frank point out, research on interactions between climate variables and other factors is well under way. However, a full study of all possible interaction effects (two-way. three-way …) between climate variables and three explanatory factors, would be difficult indeed. Studying complexity is important, but it won't help us much if models are constructed to be so complex that they are untestable.
  
  Nils Petter Gleditsch

New Security Beat

NewSecurityBeat

Guest Contributor

Responses to JPR Climate and Conflict Special Issue: John O’Loughlin, Andrew M. Linke, Frank Witmer (University of Colorado, Boulder)

Join the Conversation

Featured Media

New Security Beat

NewSecurityBeat

Guest Contributor

Responses to JPR Climate and Conflict Special Issue: John O’Loughlin, Andrew M. Linke, Frank Witmer (University of Colorado, Boulder)

Join the Conversation

Trending Stories

Featured Media

Related Stories