¿Por Quién Votará? Experimental evidence about language, ethnicity and vote choice (among Republicans)

Abstract

In the summer of 2012, two candidates faced off for the Republican nomination to the US Senate from Texas: a Cuban-American with limited Spanish-language skills and a bilingual Anglo. How important were these counter-stereotypical language skills? Does fluency in Spanish (or the lack thereof) matter to voters? Does the degree to which language ability matters depend on the ethnicity of the candidates, or of the voters? We explore these questions with a series of randomized survey experiments: one with Republican voters in Texas just prior to the primary election, and later with three Internet experiments. We find that language ability matters. Latino voters, including Republican Latinos in Texas, reward bilingual ability. In contrast, Anglo Republican voters punish it, but not in Texas.