Immigrant and Native:
Mexican American Partisanship and Candidate Preference in the 2004 Elections Across
Louis DeSipio & Carole J. Uhlaner
University of California, Irvine
If post-election media reports are to be believed the 2004 election reflected a significant
turning point for the Latino electorate (Alonso-Zaldivar 2004; Johnson 2004, two of many
examples). According to the National Exit Poll (NEP) that was the source for most media
analyses, the long-term loyalty of non-Cuban Latinos to the Democratic Party may have declined
considerably and in at least one state of traditional Democratic Latino dominance – Texas – a
majority of Latino votes went to the Republican presidential candidate (see Tables One and
Two). While these numbers are disputed and probably overstate the 2004 Republican Latino
vote (Leal et. al. 2005), the national spotlight on possible dynamics of change in the Latino
electorate raises a long overdue challenge to analyze whether change in occurring within the
Latino electorate and, if it is, whether the change is community-wide or among specific and
identifiable subsets of the Latino community.
[Tables One and Two Approximately Here]
We seize this challenge in this paper. We are interested not only in the determinants of
partisanship and vote choice of Latinos in the 2004 election but also in whether and how these
change across immigrant generations. Because of our focus on generational change, we restrict
the analysis (for reasons noted below) to Mexican Americans. Our interest in change is driven
by the increasing heterogeneity of the Latino community. Two of the most dynamic
compositional changes in the community are the growth in the number of naturalized citizens
who began to move into U.S. citizenship in large numbers in the late 1990s and the growth and
steady aging of the second generation (made up largely of the U.S.-born children of post-1965
1
We would like to express our appreciation to the Washington Post, Univision, and the Tomás Rivera Policy
Institute for access to the data analyzed here.
1