This study seeks to empirically determine to what extent continual protest by regionalist parties may generate revenue for their regions. To this end, we perform an econometric estimation using the collaboration agreements between Spanish governments and the autonomous communities as the dependent variable (first-level political and administrative divisions, CCAA in their Spanish initials). We test our hypothesis by analogously applying the economic specifications employed in studies of "pork barrel politics", including control variables regarding per capita income, regional financing systems, political variables such as support for regional governments from the same political party or the existence of pivot parties. The results support the theoretical conclusions reached by Treisman (1999), namely that non-sovereignist regionalism generates revenue while sovereignist nationalism or regionalism leads governments to react by applying unfavourable treatment. Similarly, the fact that a regionalist party plays a key role in the investiture of the national president brings with it even greater revenue to the region in question, concurring with the results predicted by Brancati (2008).

Key words: regionalism, asymmetrical federalism, Hirschman, pork barrel.
JEL classification: H77, D74, D72, D78.