White Votes Count: the Effect of Black Candidates’ Qualifications.

Built on previous literature of deracialization and white voting behavior, this paper empirically tests how the effectiveness of deracialization, measured by white crossover vote, may be conditional, dependent on urban racial contexts. Using the newly developed ecological inference (EI) method, the paper examined 81 back candidates’ white votes in New Orleans biracial elections between 1977 and 1998. The results of OLS multiple regression show that the differences in white crossover voting in New Orleans during the last two decades were related to black candidates’ strength, measured by incumbency status and newspaper endorsement. The relatively larger effect of newspaper endorsement on white crossover voting, compared to that of black incumbency, confirms previous findings regarding the powerful influence of news media on biracial elections in general and the deracilization strategy in particular. The most important finding of this research, however, is that the deracialization strategy was most effective when white voters were no longer the majority in the urban elections. [ABSTRACT FROM AUTHOR]