Over at MICROCON, a Micro Analysis of Violent Conflict working group, Eleonora Nillesen and Philip Verwimp have put out a new working paper on Rebel Recruitment in Coffee Exporting Economy. Verwimp and Nillesen use Burundi as a case study and explore violence in 1993 and 1994 by analyzing new community-level data sets. Interestingly, the paper sheds light on ‘export crop dependent, war prone economies’:
We find that large-scale indiscriminate violence in 1993 was counter effective; instead of repression rebellion it led to a large increase in recruitment the year after. Following 1993, the war got more selective in character. We test to what extent coffee producer prices have a strong effect on recruitment the year after, which is likely to result from lower opportunity costs due to depressed prices. The interpretation is confirmed in our robustness analysis where we found that province-year deviations of long-term rainfall affect recruitment.
If you dig economics speak and the debate on the role of greed and grievance in conflict, this paper is for you.
Mo'dernity, Mo'problems 
A little rationality lifts the quality of the debate here. Thanks for coittiburnng!