By Andreas Juon
To safeguard peace in ethnically divided states, scholars and practitioners often recommend territorial autonomy, for instance in the form of federalism or decentralization. In line with this prescription, evidence shows that territorial autonomy can help prevent and terminate civil wars. However, there is increasing concern that territorial autonomy can generate new violence at the subnational level, in the form of communal clashes.
Does territorial autonomy redirect ethnic violence to the subnational level?
Recent observations from around the world highlight that territorial autonomy can provide for national-level peace, but increase the risk of subnational communal clashes. For instance, discussing Indonesia's Aceh, Barter shows that autonomy has shifted "conflict from the national level to a new, regional scale." While helping to end the 1976-2005 secessionist civil war, autonomy simultaneously enabled Acehnese militants to discriminate against Aceh's highland minorities. This has sparked protest and ethnic tensions, which repeatedly escalated into communal violence.
Judging by the experiences of other multi-ethnic states that have decentralized or federalized, the intra-regional tensions in Aceh do not appear to be an exception. Green argues that decentralization in Uganda has replaced "national" with "local-level conflict". Kendhammer finds that Nigerian federalism has exacerbated sectarian violence "at the local level". Bhattacharyya, Hausing, and Mukherjee argue that India's federal system has unleashed "bloody conflicts at the sub-state level." Finally, in a co-authored piece by Livia Rohrbach and me, we find that Ethiopia's ethnofederalism has fostered "localized forms of conflict".
Does territorial autonomy then redirect violence from the national to the subnational level, rather than reducing it altogether? As indicated by the above examples, investigating this possibility is of utmost political relevance. Communal clashes between ethnic groups are increasing around the world. In some world regions, they have even become the main threat to people's livelihoods. If territorial autonomy indeed increased communal violence, its merits as a peacebuilding tool would be in severe doubt.
The importance of combining autonomy with national-level power-sharing
In a recent article, I argue that we should take seriously the risks that territorial autonomy increases communal tensions within regions characterized by internal ethnic diversity. However, I show that whether these tensions escalate critically depends on the composition of the national government.
I depart from the observation that territorial autonomy can increase tensions in ethnically heterogeneous regions. Autonomy often disproportionately benefits specific ethnic groups who are politically dominant in a region, for instance because they form the regional demographic majority. Members of these regionally dominant groups may profit from overrepresentation in the regional government, monopolize influence over its policies, and enjoy privileged access to economic goods, including land, resource rents, and state jobs.
In contrast, territorial autonomy is less likely to improve the status of regionally non-dominant groups, often corresponding to demographic minorities. These groups often lack representation in the region's political institutions and are excluded from the economic benefits that are tied to those institutions. In some cases, they might even become victims of regional "nation-building policies", which can entail their forceful assimilation or expulsion. Through these processes, territorial autonomy can engender severe tensions between regionally dominant and non-dominant groups.
However, not all tensions between regionally dominant and non-dominant groups escalate into communal violence. I argue that much depends on the willingness and ability of the national government to mediate and intervene in intra-regional disputes. If the national government is ethnically inclusive and provides for the balanced representation of regionally dominant and non-dominant groups, it is well positioned to prevent escalation. Most fundamentally, it can address communal tensions at their root by encouraging conciliatory behavior by regionally dominant groups, for instance by pushing for regional power-sharing. Where communal relations are already tense, an inclusive national government can use its leverage over both sides to mediate compromises.
In contrast, ethnically unbalanced national governments struggle more to deescalate tensions in ethnically heterogeneous regions. If national government representation favors regionally non-dominant groups, these groups' leaders might employ violence to provoke national government intervention in their favor. Conversely, if national government representation favors the same groups that also control the regional government, regional leaders will be especially uninhibited in monopolizing regional political power and economic resources. In turn, this generates especially combustible grievances among their disadvantaged peers and increases the risk of episodic mass-driven violence.
Testing the circumstances under which territorial autonomy reduces the risk of violence
In my article, I test these arguments in a global quantitative analysis that comprises administrative regions across all multi-ethnic countries between 1989 and 2019. This enables me to compare cases where communal violence did indeed follow the provision of territorial autonomy¾as in Indonesia's Aceh, Uganda, Nigeria, India, and Ethiopia¾with other cases that never saw communal violence and with unitary states (see map for the situation in 2019).
I find that ethnic groups in highly autonomous regions are less likely to become involved in rebellions against the national government, regardless of whether they control the regional government. In contrast, groups in regions that only have limited autonomy or that lack autonomy altogether are substantially more likely to rebel. Hence, the pacifying consequences of autonomy extend even further than previously acknowledged.
However, I also find that territorial autonomy coincides with a higher risk of communal violence within an administrative region. In line with my argument, these risks are especially pronounced in countries where national government representation is unequal. In contrast, territorial autonomy is not associated with communal violence where the national government is broadly inclusive. This underlines warnings that territorial autonomy may shift violence to the subnational level. Yet, it suggests that these risks can be countered by broad-based inclusion in the national government.
Addressing the risk of subnational violence within autonomous regions
In sum, where territorial autonomy is adopted as a peace-building tool, my findings indicate that it should be combined with national-level power-sharing to alleviate tensions within ethnically diverse regions. A related remedy is subnational power-sharing, whereby diverse groups commit to sharing power and economic resources in the regional government. In multiple contexts, including Kenya and Nigeria, such arrangements have demonstrably helped to defuse communal tensions generated by territorial autonomy and reduced the risks of communal violence. Combining territorial autonomy with power-sharing at the national and regional levels promises to reinforce its already important merits as a peacebuilding tool.