Hispanic voters in the 2000 and 2004 presidential general elections.

Kenski, Kate ; Tisinger, Russell

In the highly contested 2004 presidential campaign, Republican George W. Bush captured a slim majority of the popular vote (Federal Election Commission 2005). Bush received 50.73 percent of the vote compared to 48.27 percent obtained by Democratic presidential nominee John Kerry. The 2.46 percent difference was larger than the 0.51 percentage difference that separated Bush and Al Gore in 2000 (Federal Election Commission 2001). Consequently, it is of interest to determine the groups with whom Bush specifically and the Republican party more generally made gains in 2004 over the level of support they received in 2000. Using data from the 2000 and 2004 National Annenberg Election Surveys (NAES), we examine the extent to which Bush and the Republicans made gains among an increasingly important group and growing segment of the U.S. population, Hispanic Americans. Results from the NAES demonstrate that Bush made inroads with Hispanics in 2004. This did not mean, however, that Hispanics had more favorable impressions of Bush in 2004 than they did in 2000. Party identification among Hispanic registered voters was comparable for the two time periods.

According to the National Elections Pool (NEP) exit poll, President Bush garnered 44 percent of the Hispanic vote (Kenski and Kenski 2005). That marked the second straight presidential election in which the Republican party significantly increased its share of Latino voters. In 1996, Robert Dole won 21 percent of these voters, and in 2000, Bush won 31 percent (Connelly 2004). There has been disagreement among political researchers and activists over the level of support Bush received from Hispanics in 2004 (Dunham 2004; Fears 2004). In an exit poll conducted by the William C. Velasquez Institute, Bush won 34 percent of the Latino vote, results similar to those found in 2000 (Dunham 2004).

If the NEP exit poll figures are correct, then the story these numbers tell is one that questions whether Hispanics can be considered a voting bloc. After the 2004 election, political strategists argued that Hispanic voters may be breaking from their traditional Democratic voting patterns, no longer resembling a monolithic voting bloc at all. Instead, this "bloc" may become a demographic group that splits its votes more evenly between Democratic and Republican candidates (Kasindorf 2004). Democratic candidates may no longer be able to count on a majority of these voters to show up to the polls and vote for them. Despite evidence of an "emerging Democratic majority" (Judis and Teixeira 2002) fueled partly by Latino voters, other research suggests that a large portion of these voters has not yet settled on a political party. Unlike African Americans, who Democratic candidates rely on as a voting bloc, Hispanic voters may be more open to appeals from both political parties (Shapiro 2005).

This line of argument suggests that there are still large segments of Latino voters that could be mobilized--arguably by either party--but simply have not been. A theme of the research in the recently published volume titled Muted Voices edited by Rodolfo O. de la Garza and Louis DeSipio (2005) is that Latino voters could be a much more powerful bloc if elites and Latino interest groups would only mobilize them. One implication of this situation for political parties is that with the right message, either party could appeal successfully to a large number of these voters. The research in this area sketches a kind of "field of dreams" scenario: if a political party can offer the right message to mobilize Hispanic voters, then they will respond. In other words, "if you ask them, they will come."

The latest chapter in the debate about shifting Latino voters involves the extent of Bush's gains among Hispanics in the 2004 presidential election. Statisticians have challenged the NEP exit poll results. Some public opinion researchers have charged that those numbers are biased because they interviewed too few Latinos and too many suburbanites. Other exit polls reported significantly lower numbers of Latinos voting for Bush. In short, the nature and extent of the 2004 shift is an open empirical question. In this article, we use the data from the NAES to help answer questions about the 2004 election results. While no single study is ever definitive, our results provide strong evidence to suggest that Bush support among Hispanics improved over the four-year period. In this study, we draw on 831 interviews in 2000 and 990 interviews in 2004 conducted with Hispanic registered voters. These interviews took place during the final eight weeks preceding the general elections and two weeks immediately after them. (1) These data help determine (1) whether Hispanics indeed moved toward the GOP during 2004 and (2) if so, how dramatic the shift was.

Before we examine the NAES data, it is worth noting that the political stakes of winning over these voters appear to be quite high. Political observers and strategists are increasingly convinced that winning over Hispanic voters is an essential part of capturing the White House. One news weekly boldly declared that "the Hispanic vote has become the El Dorado of American elections" (Campo-Flores and Fineman 2005, 7). Because electoral votes (not the popular vote) determine who wins the presidency and Hispanic populations tend to be concentrated in some states more than others (e.g., Florida and New Mexico), some analysts argue that in a close presidential election, the votes of Latinos could determine the outcome. Both the Republican and Democratic parties seem to behave as though this is the case, competing more aggressively for the hearts and minds of these voters. Evidence of this battle played out on our television screens during the 2004 presidential campaign. Spanish was spoken early and often from the podiums of both party conventions. And, ad makers targeted specific subgroups of Latinos rather than the general demographic group (Campo-Flores 2004).

Trends in Hispanic Voting Behavior

Survey research from the mid 1990s to the present has outlined a number of demographic and political factors that have made Latino voters generally sought after by both political parties: the current number of Latinos living in the United States, the projected growth of the community, the growth of registered voters in particular, and the concentration of Hispanic voters in states with large numbers of electoral votes (DeSipio 1996; DeSipio and de la Garza 2005; Marbut 2005).

Long-term demographic trends can be seen as fueling an "emerging Democratic majority" based in part on a booming Hispanic population (Judis and Teixeira 2002). The latest U.S. Census shows that African Americans are no longer the largest identifiable racial and ethnic group--Latinos are. Long-term demographic projections suggest that there may be mass changes in politics and participation in the United States involving increasing numbers of Latino voters (Shapiro 2005). At the very least, the potential exists for Latinos to follow in the footsteps of African American voters who, in the 1960s, responded to mobilization efforts and became a powerful voting bloc for the Democratic party.

On the other hand, research on Latino voting taken as a whole shows a much more complicated picture of these voters--one in which Republican gains among these voters is plausible. It is not a certainty that Hispanic voters will behave as African Americans did in the 1960s. In fact, a consistent theme of research since the late 1990s has been that many Latino voters are ready to be mobilized, but elites have failed to do enough to energize them. One implication of this research is that the Democratic party does not have a monopoly on mobilizing these voters (Shapiro 2005).

Other Republican candidates in recent years managed to win unexpectedly large percentages of Latino voters. In New York, Republican Governor George Pataki and New York City Mayor Michael Bloomberg garnered higher proportions of Latino voters compared to past levels of support. This is more evidence of the possibilities open to the Republican party (Shapiro 2005).

If large numbers of Latino voters are open to mobilization, how might the Republican party appeal to them? Republican strategists have been increasingly focused on this question since the mid 1990s, when efforts to appeal to white voters led Republican candidates to come out in favor of immigration restrictions. Those efforts appeared to fuel Hispanic allegiance to the Democratic party, especially in California (Marbut 2005). Specifically, the 1994 Contract with America proposed restraining benefits to illegal immigrants, and in that year's gubernatorial race in California, Republican candidate Pete Wilson campaigned on a similar platform.

After these campaigns, a number of Republican analysts managed to convince candidates that while attacking immigration might increase support among white voters, gains would be outweighed by losses among Latinos and moderate white voters (Marbut 2005). Republicans have pursued Latino voters with middle-class economic issues as well as socially conservative "value" issues. Middle-class issues include education reform, prescription drug programs, Social Security reform, middle-class tax cuts, and small business promotion. Values appeals were about commitment to families, self-reliance, and a strong military (Marbut 2005).

At least in the case of values appeals, research supports the idea that Hispanics might be attracted to Republicans on the basis of these issues. Recent research on Latino voters in Texas found that Latino voters largely identified the Republican party as the party of morality and values (Connaughton 2005). Hispanics did perceive real differences between the Democratic and Republican parties. Democrats were seen as the party of the poor and Republicans were for a more laissez faire approach and for business, but also with morality and values. The argument is that while an economic message might work well for Democrats, an appeal to Catholic or even Protestant values would work well for Republicans.

In addition to values appeals, Republican candidates could emphasize a core Republican idea of individualism, de la Garza, Falcon, and Garcia (1996) found that compared to Anglos, Mexican Americans were just as or more supportive of values such as economic individualism and patriotism.

Finally, an ominous demographic sign for Democrats is that younger Hispanics are less committed to the Democratic party. Connaughton (2005) surveyed Texas Latinos and found that while most affiliate with the Democratic party as opposed to the GOP, the strength of their identification is "questionable." She concludes that "partisan identification of young Texas Latinos is not set in stone" (p. 124).

The 2004 Shift to the GOP: The State of the Debate

The 2004 NEP exit poll, the one most cited by journalists, reported that Bush garnered an unexpected 44 percent of the Hispanic vote. But the true percentage of Hispanics voting for the Republican presidential ticket in 2004 is still an open question. The William C. Velasquez Institute, a nonpartisan think tank, charged that the NEP exit poll as well as other media polls interviewed too few Hispanics and too many suburbanites. The numbers, they argue, are biased as a result. The Velasquez Institute conducted its own exit poll and reported Bush garnered just 34 percent of the Latino vote, basically the same percentage he received in 2000 (Dunham 2004). More recently, a Business Week analysis of actual election returns from sixty-two jurisdictions in thirteen states--mostly places where Hispanics comprised 75 to 93 percent of the population--found that Bush improved on his 2000 performance in 85 percent of these heavily Hispanic areas (Dunham 2004). Overall, however, the exit poll found that Bush won closer to 37 percent of the Latino vote. Warren J. Mitofsky charges that the Business Week study undersampled more conservative Latino voters living in mostly Anglo areas (Dunham 2004).

Methods

Data from the 2000 NAES (for a copy of the 2000 data set, see Romer et al. 2004) and the 2004 NAES were analyzed in this study. The Annenberg Public Policy Center at the University of Pennsylvania conducted yearlong rolling cross-sectional surveys, the largest academic surveys on political attitudes and behavior conducted of the U.S. population. For 2000, the survey began November 8, 1999 and ended January 19, 2001. For 2004, it began October 7, 2003 and ended November 16, 2004. For both surveys, telephone interviews were conducted with adults in the United States each day. Households throughout the nation were randomly selected using random-digit dialing. Random individuals within each household were selected for interviewing. Callbacks and refusal conversions were employed to maximize responses. Each night a set number of replicates, random subsamples of the sampled telephone numbers, was released for interviewing to maximize the representativeness of each daily cross-sectional survey. The response rates (RR1) for the yearlong national surveys were 25 percent in 2000 and 22 percent in 2004 (for a description of response rate calculations, see American Association for Public Opinion Research 2004).

In this study, we used subsets of the NAES data. Because Latinos make up a small, although growing, percentage of the U.S. population, we used ten-week sections of data from 2000 and 2004 to obtain estimates of Latino support for the presidential candidates. The data collected between September 12 and November 21, 2000 provided 15,897 interviews with registered citizens of whom 831 self-identified as Hispanic. (2) The data collected between September 7 and November 16, 2004 yielded 17,809 interviews with registered citizens of whom 990 were Hispanic. (3) The data from each time period were weighted to take account of the household size and number of telephone lines in the residence and to adjust for variation in the sample relating to geographic region, gender, race, age, and education.

The primary independent variable in this study was ethnicity. In 2000, respondents were asked, "Are you of Spanish or Hispanic origin or descent?" Of registered citizens, 5.8 percent identified their descent as Spanish or Hispanic. In 2004, respondents were asked, "Are you of Hispanic or Latino origin or descent?" Of registered citizens, 7.8 percent reported that they were of Hispanic or Latino origin.

The major dependent variable was presidential vote preferences in the eight weeks preceding the general elections and the two weeks after them. Because substantial absentee and early voting takes place prior to Election Day (Kenski 2004), during the pre-election period, respondents were asked whether they had already voted. If a respondent had already voted, she was asked for whom she had voted for president. If a respondent had not yet voted but said she planned to vote in the general election, she was asked for whom she intended to vote. Our vote preference variable, therefore, encompasses both vote intentions and behavior.

In addition to examining overall changes in presidential vote preferences, we also examined general favorability ratings toward Republican presidential candidate George W. Bush in 2000 and 2004 and toward Democratic candidates Al Gore in 2000 and John Kerry in 2004. The favorability ratings were measured through so-called feeling thermometer scales. In 2000, survey participants were asked to rate the candidates on a scale from 0 to 100, and in 2004, they were asked to rate them on a scale from 0 to 10. Ratings of 0 to 49 in 2000 and ratings of 0 to 4 in 2004 were considered unfavorable. Ratings of 51 to 100 in 2000 and ratings of 6 to 10 in 2004 were considered favorable. The midpoints (50 in 2000 and 5 in 2004) and "don't know" responses were categorized as neutral ratings. Respondents' party identification and ideology were also assessed.

We disaggregated Hispanic vote preferences by gender, religion, language of interview, age, education, household income, marital status, region of country, and urban status of the area to determine which Hispanic voters had changed their preferences from 2000 to 2004. Finally, we compared 2004 Hispanic to non-Hispanic registered citizens on several political issues and attitudes to provide a stronger portrait of Latino voters.

Results

In both 2000 and 2004, Hispanics were less supportive of Republican presidential candidate George W. Bush and more supportive of the Democratic presidential aspirants than were registered citizens overall, as shown in Table 1. Nevertheless, Bush garnered support from Hispanics in 2004 (36.0 percent) above the level of support he received from them in 2000 (28.2 percent). A two-sample proportions test demonstrated that this increase was statistically significant (p < .001). This does not necessarily mean, however, that the Democrats lost a lot of ground with Hispanics in 2004. The NAES data show that 52.7 percent of Hispanics supported Al Gore in 2000 and 33.9 percent supported John Kerry in 2004. A two-sample proportions test revealed that these percentages were not statistically distinguishable from one another. A further look at the data suggests that in 2004, Bush obtained support from those Hispanics who had voted for third-party candidates (such as Nader or Buchanan) or had said that they did not know for whom they would vote, when asked in 2000. The Democrats lost ground in 2004 to the extent that the "don't know" (DK) responses in 2000 broke for Gore.

While Bush captured a higher percentage of Hispanic vote preferences in 2004, it is important to note that Hispanic party identification remained roughly the same between 2000 and 2004. No statistically significant differences in party identification were detected. While some subtle shifts in ideological identification occurred (e.g., some Hispanics shifted from "conservative" to either "very conservative" or "moderate"), no major changes in ideological composition were apparent.

General candidate favorability ratings for the presidential contenders declined between 2000 and 2004 for both party candidates, suggesting that Hispanics were not enamored with either Bush or Kerry. Hispanic opinions toward Bush were less favorable overall in 2004 when compared to 2000 (29.7 percent unfavorable in 2000 versus 41.6 percent in 2004). A two-sample proportions test showed that the difference between these percentages was statistically significant (p < .001). Favorable impressions toward the Democratic candidate decreased significantly as well, and unfavorable impressions increased. While 64.6 percent of Hispanics expressed positive feelings toward Gore in 2000, these favorable opinions decreased to 52.8 percent in 2004 when Kerry was assessed (p < .001). Consequently, while Gore had a 16.1 percent lead in favorability ratings over Bush in 2000 (64.6 percent favorable toward Gore versus 48.5 percent favorable toward Bush), this Democratic lead in favorable impressions decreased to 8.1 percent when Kerry was compared to Bush in 2004 (52.8 percent favorable toward Kerry versus 44.7 percent favorable toward Bush).

We examined changes in vote preference and political orientations by three Hispanic subgroups (Mexicans, Cubans, and Puerto Ricans) because prior research suggested that when disaggregated by national origins, Latino partisanship varies, with Mexican Americans and Puerto Rican Americans supporting Democrats and Cuban Americans supporting Republicans (DeSipio 1996). Because of the small sample size for Cuban Americans, the results for this group should be read with some caution. As shown in Table 1, our results from 2000 and 2004 support the characterizations of Hispanic subgroups offered by DeSipio's work (1996). Cuban Americans tended to support Bush, while Mexican Americans and Puerto Rican Americans tended to support Gore and Kerry.

After the results showed that Hispanics had increased their support for Bush in 2004, we disaggregated Hispanic presidential vote preferences by several demographic characteristics to pinpoint the types of Hispanics who changed their support to Bush. Table 2 reveals that support for Bush increased from 27.8 percent in 2000 to 40.9 percent in 2004 among Hispanic males (p < .001). Support for Bush among Hispanic females, however, did not change significantly over the four-year period. Females composed 55.1 percent of the Hispanic registered voter sample in 2004 and composed 52.3 percent of it in 2004.

A greater percentage of Hispanic Catholics voted for Bush in 2004 than did in 2000 (two-sample proportions tests significant at p < .05). In both years, however, Hispanic Catholics supported the Democratic candidate more than they supported Bush (p < .001). In 2000, statistically comparable percentages of Hispanic Protestants supported the major party presidential candidates. In 2004, a significantly greater percentage of Hispanic Protestants supported Bush (51.3 percent) over Kerry (40.8 percent, p < .001).

It is important to note that a greater percentage of Hispanic voters reported that they were Catholic (62.7 percent in 2000 and 59.1 percent in 2004) than said they were Protestant (27.8 percent in 2000 and 29.7 percent in 2004).

The sample percentages for those Hispanics interviewed in Spanish and English suggested that support for Bush increased, as shown in Table 2, but these increases did not reach conventional levels of statistical significance when analyzed by language of interview. Hispanic voters interviewed in English were more inclined to support Bush than were those interviewed in Spanish in 2004 (p < .05). In 2000, 12.9 percent of interviews with Hispanic registered voters were conducted in Spanish. In 2004, this figure increased to 19.1 percent.

In all age groups, the percentage of Hispanic voters supporting Bush increased between 2000 and 2004. These increases were only statistically significant, however, for two of the age groups. Among Hispanic voters 65 years and older (who composed 9.9 percent of the Hispanic registered voter sample in 2000 and 9.0 percent in 2004), Bush support increased by 15.6 percent (p < .05). Among 35 to 50 year olds (who composed 37.8 percent of Hispanic voters in 2000 and 35.9 percent in 2004), Bush support also increased (p < .01). Hispanic voters age 18 to 34 years made up the largest proportion of Hispanic registered voters (38.0 percent in 2000 and 40.6 percent in 2004). The percentage of support for Bush among the youngest Hispanic voting group was statistically comparable for 2000 and 2004.

When Hispanic vote preferences were analyzed by education, the results showed that Hispanics with a high school degree or less (p < .01) or with some college or vocational education (p < .05) increased their support for Bush between 2000 and 2004, but their support for the Democratic presidential candidate was comparable for the two elections. Among those with a four-year college degree or more, the result was the opposite. While the level of support for Bush remained the same, Hispanics with a four-year college degree increased their levels of support for the Democratic candidate by 11.4 percent (p < .05) between 2000 and 2004. Among Hispanic registered voters, 17.4 percent in 2000 and 18.4 percent in 2004 had received a Bachelor's degree or higher.

Bush obtained additional support in 2004 from lower income Hispanic voters and those who were married or living as married. Among Hispanic voters with household incomes of $25,000 or more, the levels of support from 2000 to 2004 did not change significantly. Among those Hispanics with household incomes less than $25,000, Bush received an additional 11.0 percent in support in 2004 over the 18.8 percent he had received in 2000 (p < .01). When it came to those Hispanics who were married or living as married (who composed 59.7 percent of Hispanic registered voters in 2000 and 60.3 percent in 2004), 30.3 percent supported Bush in 2000 and 41.3 percent supported him in 2004. This 11.0 percent increase was statistically significant (p < .001).

Looking at Hispanic vote preference by region of the country and type of locale (urban, suburban, or rural), the results show that Bush gained support from Hispanics in the Northeast (p < .01) and South (p < .001). Among Hispanics in the West, however, Democratic presidential support increased (p < .001), while Bush support remained comparable between 2000 and 2004. When Hispanic vote preference was disaggregated by urban status, the data showed that in urban locales, Bush support among Hispanics increased (p < .01). Support for the Democratic presidential candidate was statistically comparable.

Finally, we compared the political opinions of Hispanic and non-Hispanic registered voters in 2004 to gain a better understanding of the ways in which Hispanics differ from other voters. As shown in Table 3, the political opinions of Hispanics differed from non-Hispanics in several ways. When it came to opinions about the economy, non-Hispanics reported that the country was in better shape than did Hispanics (Chi-square = 48.981, df = 5, p < .001). Non-Hispanics also rated their own personal economic situation more favorably than did Hispanics (Chi-square = 96.482, df = 5, p < .001). When asked to assess the situation in Iraq, Hispanics were more likely to contend that Iraq was not worth it compared to non-Hispanics (59.4 percent Hispanics compared to 48.8 percent non-Hispanics; Chi-square = 28.224, df = 3, p < .001). While a majority of both Hispanics and non-Hispanics supported making the recent federal tax cuts permanent, Hispanics were less favorably disposed toward the position than were non-Hispanics (Chi-square = 8.717, df = 3, p < .05). On other issues, Hispanics and non-Hispanics gave comparable responses. Both groups had similar impressions of the general direction of the country and opinions on Social Security in the stock market, allowing reimporting drugs from Canada, and abortions laws.

Conclusion

The source of the 2004 Bush victory is an important matter to political researchers and strategists. Political analysts and activists have hotly contested the extent to which Hispanics were a part of this source. Data from the NAES reveal that Hispanic support for Bush did increase from 2000 to 2004. While we agree that Bush support increased in 2004, the magnitude of this increase is more modest than the 44 percent offered by the NEP exit polls. Our data suggest that Bush support among Hispanics in 2004 was 36.0 percent--up 7.8 percent from the 2000 election. Interestingly, the increase in support for Bush cannot be attributed to stronger overall favorability ratings or shifts in party identification or ideology. Bush's favorability ratings were lower in 2004 than 2000. Luckily for Bush, Kerry's ratings were lower than Gore's ratings had been in 2000. Hispanic party identification and ideology were comparable over the four-year period. So while Hispanics contributed to the victory of the Republican presidential candidate, their general political orientations did not change.

Some segments of the Hispanic population changed their levels of support for Bush, while other segments did not. The increased Hispanic support for Bush came from males, those with less than a four-year college degree, those living in households with yearly incomes less than $25,000, and those who were married or living as married. Bush support also increased among those Hispanics living in the Northeastern and Southern regions of the country. Support for Bush did not change among those Hispanics living in the West. But support for the Democratic presidential candidate did increase among Hispanics in the West.

Democratic presidential candidate John Kerry was not able to capitalize on Hispanics' issue orientations, which should have favored the Democratic party. A comparison of Hispanics to non-Hispanics reveals that in 2004, Hispanics had less favorable impressions of the economy than did their non-Hispanic counterparts. And when it came to assessing the war in Iraq, Hispanics were much less likely to think that the war in Iraq was worthwhile. Hispanics and non-Hispanics had similar opinions on several issues, such as Social Security in the stock market and abortion laws. A slim majority of Hispanics and non-Hispanics felt that the country was seriously off on the wrong track.

A limitation to our study should be noted. The results presented in this study include data collected prior to Election Day. Consequently, it is possible that some of the individuals interviewed changed their opinions after they were interviewed. In order for our sample size to include a large number of Hispanics, it was necessary to use our pre-election data. Therefore, our vote preference estimates may be different from the final vote tallies. While our estimates may differ from the final vote tallies, our results provide compelling evidence to suggest that significant shifts in vote preference took place between two comparable ten-week time periods in 2000 and 2004.

Hispanics compose a growing and important segment of American society. They are a group that should not be taken for granted by either party. Democrats should not assume that Hispanic voters will necessarily vote for Democrats when they go to the polls. Republicans should not assume that because Bush made gains with Hispanics in 2004, these gains represent an allegiance to the Republican party.

AUTHORS' NOTE: We would like to thank Kathleen Hall Jamieson for providing the resources fir this study. We would also like to thank Adam Clymer for providing insights on this topic and Henry C. Kenski for giving comments on an earlier draft of this article.

References

American Association for Public Opinion Research. 2004. Standard definitions: Final dispositions of case codes and outcome rates for surveys, 3d ed. Lenexa, KS: AAPOR.

Campo-Flores, Arian. 2004. Leave no Hispanic behind. Newsweek, November 8, p. 9.

Campo-Flores, Arian, and Howard Fineman. 2005. A Latin power surge. Newsweek, May 30, p. 7.

Connaughton, Stacey L. 2005. Inviting Latino voters: Party messages and Latino party identification. New York: Routledge.

Connelly, Marjorie. 2004. How Americans voted: A political portrait. New York Times, November 7, Section 4, p. 4.

de la Garza, Rodolfo O., and Louis DeSipio, eds. 2005. Muted voices: Latinos and the 2000 elections. New York: Rowman & Littlefield.

de la Garza, Rodolfo O., Angelo Falcon, and F. Chris Garcia. 1996. Will the real Americans please stand up: Anglo and Mexican-American support of core American political values. American Journal of Political Science 40(2): 335-51.

DeSipio, Louis. 1996. Counting on the Latino vote: Latinos as a new electorate. Charlottesville: University Press of Virginia.

DeSipio, Louis, and Rodolfo O. de la Garza. 2005. Between symbolism and influence: Latinos and the 2000 elections. In Muted voices: Latinos and the 2000 elections, edited by Rodolfo O. de la Garza and Louis DeSipio, 13-60. New York: Rowman & Littlefield.

Dunham, Richard S. 2004. Did Hispanics really surge to Bush? Business Week, November 29, p. 51.

Fears, Darryl. 2004. Pollsters debate Hispanics' presidential voting: Discrepancy in estimates vs. results examined. Washington Post, November 26, p. A04.

Federal Election Commission. 2001, December. 2000 presidential popular vote summary for all candidates listed on at least one state ballot. Retrieved March 1, 2005 from http://www.fec.gov/pubrec/fe2000/prespop.htm.

--. 2005. Federal elections 2004: Election results for the U.S. president, the U.S. Senate and the U.S. House of Representatives. Retrieved August 10, 2005 from http://www.fec.gov/pubrec/fe2004/federalelections2004.pdf.

Judis, John B., and Ruy A. Teixeira. 2002. The emerging Democratic majority. New York: Scribner.

Kasindorf, Martin. 2004. Hispanic voters paint a new picture. USA Today, November 11, p. 04A.

Kenski, Kate. 2004. Visualizing data across the campaign. In Capturing campaign dynamics: The National Annenberg Election Survey, edited by Daniel Romer, Kate Kenski, Paul Waldman, Christopher Adasiewicz, and Kathleen Hall Jamieson, 66-79. New York: Oxford University Press.

Kenski, Henry C., and Kate M. Kenski. 2005. Explaining the vote in a divided country: The presidential election of 2004. In The 2004 presidential campaign: A communication perspective, edited by Robert E. Denton, Jr., 301-42. Boulder, CO: Rowman & Littlefield.

Marbut, Robert G. 2005. [??]Un nuevo dia? Republican outreach to the Latino community in the 2000 campaign. In Muted voices: Latinos and the 2000 elections, edited by Rodolfo O. de la Garza and Louis DeSipio, 61-83. New York: Rowman & Littlefield.

Romer, Daniel, Kate Kenski, Paul Waldman, Christopher Adasiewicz, and Kathleen Hall Jamieson, eds. 2004. Capturing campaign dynamics: The National Annenberg Election Survey. New York: Oxford University Press.

Shapiro, Robert Y. 2005. Introduction: Awaited voices: Latinos and U.S. elections. In Muted voices: Latinos and the 2000 elections, edited by Rodolfo O. de la Garza and Louis DeSipio, 1-12. New York: Rowman & Littlefield.

(1.) These samples are drawn from larger pools of Hispanic adults who were interviewed over the entire course of the 2000 and 2004 election campaigns as a part of the NAES's national rolling cross-sectional surveys. In 2000, the NAES conducted 4,676 interviews with Hispanic adults in the United States. In 2004, 6,155 interviews with Hispanics were conducted.

(2.) Of these 831 Hispanic registered voters, 385 were Mexican, 43 were Cuban, and 120 were Puerto Rican.

(3.) Of these 990 Hispanic registered voters, 472 were Mexican, 75 were Cuban, and 168 were Puerto Rican.

Kate Kenski teaches political communication at the University of Arizona and was a member of the 2000 and 2004 National Annenberg Election Survey teams at the Annenberg Public Policy Center of the University of Pennsylvania.

Russell Tisinger is a doctoral candidate at the Annenberg School for Communication at the University of Pennsylvania and was a member of the 2004 National Annenberg Election Survey team at the Annenberg Public Policy Center of the University of Pennsylvania.

TABLE 1

Vote Preference and Political Orientations by Ethnicity
in 2000 and 2004 (%)

 All
 All Hispanics Mexicans
 2000 2004 2000 2004 2000 2004
Vote preference
Bush 41.2 46.5 28.2 36.0 26.8 36.9
Gore (2000) 41.2 -- 52.7 -- 53.8 --
Kerry (2004) -- 44.5 -- 53.9 -- 54.0
Nader 5.2 1.8 4.9 2.4 5.1 2.5
Buchanan 1.4 -- 0.8 -- 1.4 --
Other 0.5 0.4 0.2 0.5 -- 0.5
Did not/would not 1.0 0.7 1.1 1.0 0.6 0.2
 vote for president
Don't know 6.5 4.0 9.7 4.8 9.3 3.6
Refused 3.1 2.0 2.5 1.4 3.0 2.2
Opinion of Bush
Favorable 52.6 50.8 48.5 44.7 51.6 45.8
Unfavorable 30.3 40.0 29.7 41.6 27.1 38.1
Neutral/don't know 16.6 9.0 20.8 13.5 21.1 15.7
Refused 0.4 0.2 1.0 0.2 0.2 0.3
Opinion of Gore
 (2000)/Kerry
 (2004)
Favorable 50.9 44.3 64.6 52.8 68.7 53.3
Unfavorable 33.7 40.8 18.2 26.7 15.7 26.7
Neutral/don't know 15.0 14.8 16.7 20.5 15.6 20.0
Refused 0.4 0.1 0.5 -- -- --
Party identification
Republican 29.3 32.2 19.8 21.4 17.1 18.8
Democratic 33.4 34.7 46.7 43.9 52.6 46.1
Independent 27.3 24.8 21.7 24.1 22.3 23.4
Other 7.5 5.3 7.8 6.6 5.4 6.9
Don't know 1.9 2.0 3.3 3.5 2.3 4.1
Refused 0.5 1.1 0.7 0.6 0.3 0.6
Ideology
Very conservative 6.8 9.5 6.0 7.6 3.9 6.5
Conservative 31.3 29.5 31.6 26.3 32.8 28.4
Moderate 38.2 37.5 37.0 39.8 42.7 38.6
Liberal 17.5 16.4 18.5 18.7 14.6 20.1
Very liberal 3.4 4.8 3.1 6.1 2.8 5.6
Don't know 2.4 2.1 3.0 1.4 2.9 0.7
Refused 0.4 0.2 0.8 1.0 0.4 0.1

 Cubans Puerto Ricans
 2000 2004 2000 2004
Vote preference
Bush 46.4 52.5 26.4 31.9
Gore (2000) 43.5 -- 54.6 --
Kerry (2004) -- 39.4 -- 53.4
Nader 4.5 4.1 5.2 1.2
Buchanan -- -- -- --
Other 4.4 0.8 -- 1.1
Did not/would not -- -- 0.4 2.3
 vote for president
Don't know 1.2 2.0 13.4 9.0
Refused -- 1.1 -- 1.0
Opinion of Bush
Favorable 53.4 55.9 42.1 39.8
Unfavorable 29.9 34.8 32.1 45.7
Neutral/don't know 16.7 9.2 24.6 14.4
Refused -- -- 2.1 --
Opinion of Gore
 (2000)/Kerry
 (2004)
Favorable 48.2 43.8 62.5 50.0
Unfavorable 30.5 39.7 13.7 22.5
Neutral/don't know 21.3 16.5 21.7 27.5
Refused -- -- 2.1 --
Party identification
Republican 41.9 48.6 15.9 17.9
Democratic 35.3 28.2 47.6 49.6
Independent 22.8 21.1 20.1 22.4
Other -- 1.8 6.9 4.1
Don't know -- 0.3 9.1 5.7
Refused -- -- 0.4 0.2
Ideology
Very conservative 0.6 6.2 7.2 12.2
Conservative 39.2 32.1 40.4 26.0
Moderate 30.8 35.9 29.3 37.4
Liberal 28.9 13.8 14.6 18.0
Very liberal -- 6.9 4.2 4.5
Don't know 0.6 4.8 2.3 1.9
Refused -- 0.3 2.1 --

Note: Column percentages are given.

TABLE 2

Hispanic Presidential Vote Preferences in 2000 and 2004
by Demographic Characteristics (%)

 2000 2004
 Bush Gore Bush Kerry

All Hispanics 28.2 52.7 36.0 53.9

Men 27.8 55.0 40.9 49.3
Women 28.6 51.0 31.7 58.3

Protestant 42.7 40.5 51.3 40.8
Catholic 23.2 58.1 29.9 58.9

Interviewed in English 29.8 52.5 37.9 52.7
Interviewed in Spanish 16.8 54.2 27.8 59.8

18-34 years old 30.8 50.6 35.3 54.0
35-50 30.1 49.2 39.7 52.6
51-64 21.4 57.9 29.8 58.5
65 and older 20.0 67.1 35.6 52.5

Education

High school or less 24.6 55.9 34.3 55.0
Some college or vocational school 29.8 52.5 39.9 50.9
Bachelor's degree or more 36.6 43.1 35.5 54.5

Household income under $25,000 18.8 55.4 29.8 58.0

$25,000 to less than $50,000 32.1 52.8 37.2 53.9
$50,000 to less than $75,000 35.5 47.7 43.5 50.5
$75,000 or more 37.4 47.5 41.7 49.4

Married or living as married 30.3 52.0 41.3 48.9

Single 21.2 56.4 27.4 64.6
Northeast 17.2 62.6 31.7 56.5
Midwest 25.4 46.5 39.6 45.9
South 34.9 50.8 50.1 42.2
West 28.3 50.3 26.7 63.4

Urban 22.1 57.6 31.7 56.4
Suburban 32.9 48.7 38.7 51.9
Rural 33.1 48.8 42.4 52.3

Note: Row percentages are given. Only major party response
options are shown.

TABLE 3

Political Opinions in 2004 by Ethnicity (%)

 Non-
 All Hispanics Hispanics

Things in this country Right direction 40.6 38.2 40.9
 are generally going Wrong track 52.4 54.2 52.3
 in the right Don't know 6.3 7.0 6.2
 direction or are
 seriously off on the
 wrong track

Economic conditions in Excellent 2.2 3.1 2.2
 this country today Good 25.2 19.1 25.7
 Fair 41.1 40.7 41.1
 Poor 31.0 36.0 30.5
 Don't know 0.5 1.1 0.4

Personal economic Excellent 9.0 6.4 9.2
situation today Good 42.0 32.5 42.8
 Fair 35.2 40.7 34.7
 Poor 13.6 20.1 13.1
 Don't know 0.2 0.3 0.2

Situation in Iraq was Worth it 46.6 36.8 47.4
 worth going to war Not worth it 49.6 59.4 48.8
 over, or not Don't know 3.5 3.7 3.5

Allowing workers to Favor 55.8 59.7 55.5
 invest some of their Oppose 37.1 35.0 37.3
 Social Security Neutral/don't know 6.8 5.1 6.9
 contributions in the
 stock market

Changing the recently Favor 76.2 73.3 76.4
 passed Medicare Oppose 16.1 18.6 15.9
 prescription drug law Neutral/don't know 7.5 8.2 7.4
 to allow reimporting
 drugs from Canada

Making recent federal Favor 56.0 51.2 56.3
 tax cuts permanent Oppose 33.4 37.5 33.1
 Neutral/don't know 10.4 10.9 10.4

Making additional stem Favor 62.0 58.3 62.2
 cell lines from human Oppose 30.8 31.1 30.7
 embryos available for Neutral/don't know 6.9 10.3 6.6
 federally funded
 research on diseases
 such as Parkinson's

Laws making it more Favor 38.5 39.7 38.4
 difficult for a woman Oppose 54.1 53.6 54.2
 to get an abortion Neutral/don't know 7.0 6.4 7.0

Amendment to the U.S. Favor 40.9 41.2 40.9
 Constitution saying Oppose 50.1 47.2 50.3
 that no state can Neutral/don't know 8.5 11.6 8.3
 allow two men to
 marry each other or
 two women to marry
 each other

Note: Column percentages are given. Refused responses are not shown.