19
is coded as a five-point scale (1=Democrat; 2=lean Democrat; 3=pure Independent;
4=lean GOP; 5=GOP) instead of seven-point because strength of partisanship was not
asked as a follow-up question. In addition, several dummy variables were created for the
issues respondents felt were the most important in the election: war on terror; war in Iraq;
education; and the economy. With both candidates courting the Latino vote in 2004, we
also include a measure that asked the respondents which candidate had a more sincere
Latino outreach effort; this was coded 0 for Bush, 1 for “the same,” and 2 for Kerry.
Finally, the key independent variable in the baseline model is military service.
Respondents were asked whether they were a veteran of the U.S. armed forces or were
currently serving on active duty in the U.S. military or national guard. Overall, 12.5
percent of Latino registered voters were either veterans or on active duty, which we
combined to create the military service variable – which is very close to the 2004 ANES
estimate for the percentage of veterans in the overall population (13 percent).
Building on the baseline model, in the second regression we introduce three
additional variables that capture attitudes about the Iraq war and campaign against
terrorism:
(1) “All in all, considering the costs to the United States versus the benefits to the
United States, do you think the war with Iraq was worth fighting, or not?”
(2) “About the war in Iraq, some people say that Latino soldiers are suffering a higher
share of the casualties, while others say that all racial and ethnic groups are
suffering equal amounts of casualties. What is your view?”
(3) “On another subject, do you think the United States is winning or losing the war
on terrorism?”
Convention