A recent literature has investigated how political selections - who runs for office and who is ultimately elected - is affected by political institutions and systems of remuneration. In this paper, we use detailed data on local politicians before and after a large-scale quasi-experimental reform of Danish municipalities to examine how government architecture affects political selection. The reform allows us to examine the effect of two different aspects of government architecture: Horizontal government architecture involves the size and number of jurisdictions; vertical government architecture speaks to the allocation of responsibilities across different levels of government. Examining horizontal government architecture we find that both the field of candidates and the elected politicians are more competent in larger jurisdictions based on several standard measure of competence. Examining vertical government architecture the picture is more complex as the transfer of additional responsibilities positively affects some measures of candidate and politician competence, while negatively affecting others.