Making Sense of the Complexity of Decentralised Governance Comment on "The Effects of Health Sector Fiscal Decentralisation on Availability, Accessibility, and Utilisation of Healthcare Services: A Panel Data Analysis".

The article by Rotulo and colleagues suggests that health sector fiscal decentralisation has been bad for Italy. But given the complexity of fiscal decentralisation, this interpretation is not necessarily so. Their analysis was based on assumptions about causality that are better suited for simple interventions. Assumptions of simplicity show up as misleading artefacts in the conclusion of evaluations of complex interventions. Complex interventions work by triggering mechanisms - eg, reasoning and learning processes - that manifest differently across the units of a decentralised system, contingent on context, evolving over time. Evaluation findings can only be partial and provisional; neither summarily good nor bad. The goal of evaluating a complex intervention - such as decentralised governance - should be to understand how, under what circumstances and for whom they are good or bad - at a point in time.