This project will extend the reach of the American National Election Study by translating the instrument for the first time into Spanish and providing bi-lingual interviewers. To make the resulting data more usable for both general behavioral scholars interested in between-group comparisons, as well as Latino politics scholars interested in analyzing NES data among Hispanic Americans, an over-sample of self-identified Hispanic/Latino citizens of the United States as part of the overall NES project will be secured. The larger sample size and adjusted sampling frame will produce a statistically valid sample of US citizens of Hispanic/Latino ancestry suitable for analysis. Finally, using these newly available data, the received knowledge with respect to partisanship, sophistication, participation and vote choice will be reexamined.
Though the American National Election Studies have been conducted in one form or another for 60 years, the format and design of the study has introduced a systematic error by interviewing only in English. Over half of these Hispanic/Latino citizens are Spanish-dominant, making comfortable completion of the NES in English an imposing task, nearly impossible for some share of the electorate. By confining the interviews to the English language, the NES has entirely excluded these citizens from the resulting data, introducing a systematic bias into the study that skews the sample toward more acculturated, higher socio-economic status respondents. Using past efforts as a guide, Spanish language interviewing should have accounted for about three percent of all completed interviews and over 1/3 of all Hispanic interviews. Moreover, demographic trends suggest that this bias will be of increasing importance as the overall Latino population, and the Spanish-dominant portion, grows. In short, behavioral scholars of national politics are employing data with a systematic and growing bias.
Latinos comprise a rapidly growing share of the United States population (now at 15%) and comprise a significant and growing share of the national electorate, estimated between seven and eight percent. Moreover, the size of the Latino electorate will grow significantly in the immediate future regardless of assumptions regarding rates of immigration and naturalization. The 2000 Census reports that over 85% of Latino/Hispanic persons under the age of 18 are native-born US citizens, suggesting that the Latino voting-eligible population will double in the next 18 years. Accurate and representative survey data that allows meaningful comparisons with non-Hispanic whites are difficult to come by and have never approximated the depth and breadth of the National Election Studies. As a consequence, this data development project and its immediate empirical products have the potential to significantly enlarge the body of knowledge with respect to Latino citizens'''' political attitudes, beliefs, and behaviors, and how they might vary from other voters.
