immigrants) who are now growing into political maturity.
Since few of these post-1965 second
generation children have yet produced an adult third-generation, the third generation continues to
be made up largely of the descendants of earlier Latino migrants, most of whom were Mexican
American. Only Mexican Americans among Latinos have significant shares of the adult
population who are immigrants, second generation, and third and beyond generations.
Consequently, we limit our analysis here to the Mexican Americans.
Our analysis examines individual and attitudinal characteristics that have previously been
shown to shape partisanship and candidate choice. We supplement these, however, with
particular attention to immigration generation. To the extent that we do find predictable
differences in partisanship and, to a lesser degree, candidate choice across the generations, we
will be able to speak not just to the 2004 election, but to potential long-term changes in the
Mexican American and, by extension, Latino contribution(s) to national elections.
We rely on a unique survey of electoral attitudes and candidate preferences among Latino
registered voters conducted by the Washington Post, Univision, and the Tomás Rivera Policy
Institute (TRPI) (we discuss these data in greater depth later in the paper). We should state at the
outset that the results of this poll, conducted in early October 2004, conflict with the results of
the National Exit Poll. Nearly 60 percent of Latino survey respondents indicated that they would
vote for Senator Kerry and just 30 percent indicated a preference for President Bush. The gap
between the candidates widens among Mexican Americans, 65 percent of whom indicate a
preference for Senator Kerry compared to 25 percent who supported President Bush (the NEP
does not disaggregate its Latino respondents by national origin). Even in Texas, Senator Kerry’s
margin among Mexican Americans was 59 percent to 32 percent. Most respondents indicated
2
We should note at the outset that the second generation are the children of all Latino immigrants, not necessarily
the children of the naturalized citizens who make up our first generational sample. The second generation, thus,
includes children of immigrants who may have been less politically engaged than naturalized citizens, registered to
vote. In the Mexican American community, at least one-quarter of the first generation parents would likely have
been unauthorized migrants, at least at the time of initial migration.
2
Convention