Americans are more exposed to difference than we think: Capturing hidden exposure to political and racial difference

Abstract Use of name generators (NG) in assessing exposure to political difference has led to the conclusion that disagreement is uncommon. Data from two representative surveys tested a NG modification to probe for exposure to difference. Over half for whom NG results suggest no exposure to difference were able to name an alter who supported an alternative candidate. Over a quarter whose NG results indicate no exposure to a racial difference were able to name someone of the opposite race with whom they’d talked politics. These findings demonstrate that mixed or diverse networks are more common than prior NG research suggests.