["Social Science Quarterly, Volume 102, Issue 4, Page 1439-1450, July 2021. ", "\nAbstract\n\nObjectives\nThis study addressesfour questions posed by V.O. Key's friends‐and‐neighbors (F&N) voting thesis that have been neglected in the scholarly literature.\n\n\nMethods\nPooled regression of 101 primary and 66 general elections in Mississippi, South Carolina, Pennsylvania, and Wisconsin for the offices of U.S. senator and governor over the period 2002–2016 yielded a total of 11,263 candidate‐county cases.\n\n\nResults\nF&N effects are most pronounced in primary elections, for Democratic candidates, and for nonincumbents. Also greater F&N effects for candidates whose “home” county populations were less well educated. The relative importance of F&N effects is displaced by social cleavage variables in elections where they are relevant to voter choices, namely general elections rather than primaries.\n\n\nConclusion\nThis study begs the larger question for future research of why F&N has greater effect for Democrats than Republicans and whether more Democrats tend to be low‐information voters.\n\n"]