This article reconciles the conflicting evidence on the relationship between governance and maritime piracy at the macro (state) and micro (sub-state) level using original data (2000-2011) and a study of piracy’s emergence in and around the Horn of Africa. It finds a consistent, positive relationship between state weakness and increased piracy production cross-nationally. However, Somali piracy belies the "anarchy as opportunity" mechanism implied by the large-N study. The pirate industry there received substantial support from local authorities, more consistent with a quadratic, "stability-enabled" argument regarding governance. I conclude that each is half-right. At the macro level, anarchy presents permissive conditions for simple banditry, but limits the potential for robust organized crime unless corruptible sub-state authorities exist and can be co-opted. Policy-wise, efficient pirate deterrence programs should concentrate on regions within failed states with this mix of characteristics. Methodologically, the results point to the limitations of existing metrics of state fragility and governance, and recommend that more nuanced measures be used.