Americanizing Latino Politics, Latinoizing American Politics by Rodolfo O de La Garza

Author:Rodolfo O de La Garza [Garza, Rodolfo O de La]
Language: eng
Format: epub
Tags: Political Science, General
ISBN: 9781351054645
Google: 5Ru_DwAAQBAJ
Goodreads: 49293375
Publisher: Routledge
Published: 2019-11-01T00:00:00+00:00

We now turn to the multivariate results from binary logistic and multinomial logistic regression models that control for the variables described in Table 5.1. As we did in our Chapter 4 analyses of political participation outcomes in the 1990 LNPS and 2006 LNS, we obtain predicted probability differences between Puerto Ricans and Mexicans, Cubans and Mexicans, and Cubans and Puerto Ricans,17 holding all else at their mean values. As before, the coefficient estimates on the relevant national origin dummy indicators tell us whether there are statistically significant differences in political participation, holding all else constant, relative to the base category at the conventional 0.05 significance level.

Consistent with our bivariate findings, we find no significant national origin difference in self-reported registration in 2008, while the large bivariate Cuban-Mexican and Cuban-Puerto Rican differences in 2012 noted earlier are no longer statistically significant after controlling for factors such as SES, partisanship, and Americanization level. However, holding all else constant, Puerto Ricans remain significantly more likely to be registered compared to Mexicans in both 2010 and 2014 (10 and 12 percentage point higher predicted probabilities, respectively),18 while in 2016, Puerto Ricans were significantly less likely to be registered than both Mexicans and Cubans (14 and 12 percentage points lower, respectively). In short, on the question of registration, our findings are mixed—while there are no consequential national origin differences in both 2008 and 2012, there are a number of moderately large differences, holding all else constant, in 2010, 2014, and 2016 (see ET-5.4).

Turning to voting intention, we find that the moderately large bivariate Puerto Rican-Mexican difference in 2008 is no longer significant. However, when holding all else constant, we find that in 2012, Cubans are now predicted to be substantially less likely to vote than Puerto Ricans (15 percentage points lower), and that Puerto Ricans now have a lower predicted probability of voting compared to Mexicans in 2016 (7.6 percentage points lower). With these exceptions, we find no other significant national origin differences of consequence.

Our multivariate analysis of vote choice in Chapter 4 showed no significant difference between our three national origin groups in voting for George W. Bush over John Kerry in 2004. This represented a substantial change from the 1990 LNPS results, which showed that both Puerto Ricans and Cubans were substantially more likely (compared to Mexicans) to vote for George H.W. Bush over Michael Dukakis in 1988. Our analyses of prospective voting preference in both presidential and midterm elections using more recent NSL data shows more mixed results.

Interestingly, even after controlling for an extensive set of control variables, many of the significant bivariate differences examined above remain statistically significant and comparable in magnitude (see ET-5.5).19 While there are large national origin differences in both 2008 (with both Cubans and Other Latinos substantially more likely to prefer McCain compared to both Mexicans and Puerto Ricans) and 2012 (with Cubans markedly more likely to prefer Romney compared to both Mexicans and Puerto Ricans (19 and 46 percentage points higher, respectively)), we find, holding all

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.