Zachariah Mampilly's Rebel Rulers is a highly original and carefully crafted addition to the literature on internal war. While most work in this literature focuses on explaining the length, duration, recurrence or causes of conflict, Mampilly studies how insurgents govern the civilian populations in the territories they control. He reminds us that rebels have been rulers in a broad range of conflicts across time and across continents but that we have given scarce systematic attention to how and why the nature of rebel governance varies so dramatically. Mampilly's work helps fill the void by asking three intriguing questions of lasting relevance. First, why do some rebel rulers provide civilians with extensive public goods while others provide few if any goods at all? Second, how do rebels make decisions about the sort of governance they undertake? And finally, why are rebels accorded legitimacy in some cases but resisted and even fought violently in others?

Mampilly answers these questions using an impressive range of original material from three well-chosen case studies including, the LTTE in Sri Lanka, the SPML/A in Sudan and the RCD in D.R. Congo. He shows us that the nature and extent of rebel rule is partially an outcome of rebel preferences (as one would guess) but that rebel relations with social and political actors at the local level and in the international arena are key as well. His multi-level, nuanced argument explains the variation in his cases – with the LTTE representing extensive and legitimating rule, the RCD representing wholly ineffective and almost exclusively coercive rule and the SPML/A representing something in between.

The local level factors that shaped rebel governance in these cases are multiple and wide ranging. Mampilly does not seek parsimony and instead embraces the complexity of each case. He does, however leave us with seven testable hypotheses extracted from his rich comparisons. Each of these is sustained within the stories he tells and worth examination elsewhere. Summarised, these hypotheses posit the following associations: that governance provision is likely to be more extensive in areas where state penetration raised citizen expectations prior to the outbreak of conflict; that effective governance is more likely to develop if insurgents aspire to secession or to ethno-nationalist autonomy; that effective governance is more likely if insurgencies adopt Maoist organisational structures; that neglected internal divisions weaken governance capacity; that stalemate or ceasefires boost the likelihood of more effective governance; that effective governance is more likely if rebels can co-opt humanitarian organisations and finally, that competition from local and transnational organisations can increase incentives for effective governance if, for ideological or other reasons, the rebels choose to deal with competition non-coercively.

This is not a simple book but it is, beyond doubt, a profound and provocative book that anyone interested in internal war should take seriously. Mampilly took huge personal risks to witness rebel rule first hand and to interview both the rulers and the ruled in all three of his cases. His risks provide rich rewards for scholars and policymakers alike.