This study revisits the drivers of income inequality with political institutions at the core. We take a multidimensional institutional approach by defining political institutions in terms of governance, political freedom, political fragmentation and political scale. The main contributions of our study are (i) presenting the most extensive analysis up-to-date of the role of political institutions by decomposing it into distinct elements and (ii) considering the difficulty and the lack of consensus and clarity regarding the choice of instruments in the literature we define an alternative methodology to deal with the potential endogeneity of political institutions. We combine an analysis of club convergence, a clustering mechanism according to the long term income trajectories of the countries under investigation with Bayesian Model Averaging (BMA) within each cluster to determine the most important variables that affect inequality out of a large set of potentially important variables. Our results show that the effectiveness of political institutions differ among country groups and that there is no "one size fits all" policy prescription that links institutional quality and income inequality.