Won\u27t You Be My Senator?: Nuanced “Friends and Neighbors” Voting in U.S. Senate Elections, 1968–2018

Recent experimental scholarship shows that local “place identity” continues to matter to voters\u27 assessments of candidates, echoing early findings by Key (1949) and others demonstrating electoral boosts in candidates’ home areas. But it is yet to reexamine these “friends-and-neighbors” effects in the observational context, nor does it parse differences between effects of local candidate roots and current candidate residency. This paper closes these gaps using an original county-level dataset of incumbent reelection races for the U.S. Senate from 1968 to 2018 (N = ∼34,000 county-years). I show that the friends-and-neighbors effect is more nuanced than previously thought, applying to candidate roots like birthplace and local educational attainment, in addition to current residency; that these effects are in some cases heterogeneous; and perhaps most surprisingly, that local candidate effects are as strong in the U.S. Senate today as in earlier decades of “low partisanship” like the 1970s