The memory that power-sharing makes

Power-sharing is usually sold as a sequence. First stop the fighting, then share the institutions, and in time the former enemies will come to share a story about the war. Bosnia and Herzegovina, thirty years after Dayton, is the place to test that sequence, and it does not hold.

Bosnian Croats and Bosniaks fought on the same side for much of the war and agree on a great deal about it, above all that the Serb forces were the aggressors. One might expect that shared narrative to deepen inside shared institutions. The opposite has happened. Competition within the consociational system, over positions, vetoes and the legitimacy of the Federation itself, has pulled the two communities’ accounts of the war apart. Historical grievances are mobilised not against the old enemy but against the partner, to contest the fairness of the arrangements and to shore up ethnonational boundaries.

The lesson is uncomfortable for peacebuilders. Institutions designed to accommodate difference can, without anyone intending it, give groups a standing reason to insist on their distinct victimhood. Memory does not simply follow institutions; it is produced by them.

That does not make power-sharing wrong. It makes the sequencing wrong. Where recognition of the past is left to arrive on its own once institutions are in place, it tends not to arrive at all. Memory work has to be built into the settlement, not expected as its dividend.