|Authors:||Jon Hernes Fiva, Askill Harkjerr Halse, Daniel M. Smith|
Geographic representation is considered to be an important factor in candidate nominations, even under closed-list proportional representation (PR), and may also matter for distributive policy outcomes. However, since nominations are determined strategically, the causal eﬀects of representation for local areas are diﬃcult to iden-tify. We study candidate nominations, voter behavior, and distributive policies in the closed-list PR setting of Norway (1953-2013). Exploiting as-good-as-random election outcomes for candidates who are marginally close to winning a seat in parliament, we ﬁnd that parties obtain higher support in subsequent elections in the hometowns of narrowly-elected candidates. This eﬀect appears to be driven by an increase in the probability of having the local candidate at the top of the party list in the next election. However, we ﬁnd no eﬀect of local representation on geographically targeted policy beneﬁts for the hometown. Our results suggest that local candidates under closed-list PR are able to attract and mobilize local voters, but either do not have the power to obtain distributive beneﬁts for their localities, or are not interested in seeking them.