Related Publications: Video
Download this Backgrounder as a pdf
- Democrats lead Republicans by a comfortable margin in the partisan identification of Latino voters. The gap is even wider among immigrant Latinos who have not yet become citizens. As many of these non-citizens naturalize, the political affiliation of Latinos is likely to shift still further toward the Democratic party.
- Across all nationality groups except Cubans, the Democratic advantage in party identification among Latinos is greater than 20 percentage points, and this is true across nearly all states. Even among Cuban-Americans, the once large GOP lead has dwindled to just six percentage points.
- Latinos become more Democratic, not less, with increasing education and tenure in the United States. Rising income also does not appreciably change Latino partisanship.
- Contrary to the thinking within the Bush White House, there is no evidence that a significant percentage of the Latino vote is "in play."
- Current immigration policy is slowly but steadily shifting the nation’s electorate toward the Democratic Party.
- There is no indication that the Republican or Democratic positions on immigration policy explain the orientation of Latino voters. Rather, it seems that the core positions of the Republican Party are simply not as attractive to Latino voters as are those of the Democratic Party.
The fact that Republican efforts to attract Latino voters are not likely to have a large impact on their voting patterns does not mean that Republicans should abandon efforts to increase support among Latinos. As far as the GOP is concerned, even a shift of only a few percentage points is better than no shift at all. Current U.S. immigration policy, however, militates against President Bush’s goal of increasing the Republican vote share since most newly naturalized Latinos vote overwhelmingly Democratic, but do so for reasons other than the Democrats’ stand on immigration policy. It is therefore puzzling that President Bush is pinning his hopes for an Hispanic realignment on a guestworker program or a limited amnesty when most Latinos join with other Americans in the belief that the current level of immigration is too high. To make serious gains among Latinos, the GOP will have to consider a wholesale policy redirection on health care and social welfare programs, changes that will be almost impossible to pursue given the party’s fiscally conservative base.
George W. Bush and prominent members of the GOP leadership have high hopes for Latino voters, believing that a political realignment is in the making, or soon can be if the right strings are pulled. Republican party chairman, Virginia Governor Jim Gilmore, recently announced plans to expand outreach efforts to Hispanics. The Bush campaign spent considerable resources courting the votes of Latinos, and reports from the White House have indicated that this effort continues, as weekly radio messages are now being broadcast in Spanish. The new president and his advisors have been carefully balancing the costs and benefits of a new amnesty program, and are said to favor a generous guestworker law. Some of the best minds working within the new Administration are charged with the task of closing the gap between Democratic and Republican voting within this growing segment of American society. In this Backgrounder, we explore whether it is realistic for the new Administration, and the Republicans in general, to spend time pursuing the allegiance of Latinos, and speculate about whether these efforts can win the votes of this increasingly consequential electoral bloc. Using recent polls to assess the potential for such a realignment, we examine the partisan identification and voting behavior among Hispanics.
The Republican case for pursuing Latino voters rests mainly on what some in the GOP perceive as the cost of not doing so. Republicans, like the Democrats, are aware of some compelling demographic realities. As of the 2000 census, Latinos represent the largest minority group in the nation, slightly bypassing African Americans in numbers. Beyond this, Census Bureau population projections expect dramatic declines in the proportion of white non-Hispanics, particularly if current immigration trends continue. While non-Hispanic whites now comprise 72 percent of the total U.S. population, this proportion is expected to fall to 64 percent by 2020 and to 53 percent by 2050 (see Figure 1). The highest rates of population increase are expected from Latinos and Asians with annual growth rates that may exceed 2 percent per year. (To put these phenomenal rates into perspective, even at the height of the post-World War II baby boom, the U.S. population never grew by 2 percent in one year.) In absolute numbers, the largest growth is anticipated among Latinos (see Figure 1). Furthermore, given the relative disparity in birth rates between minorities and non-Hispanic whites, white children are expected to comprise less than half of the total under age-18 population by as soon as 2030. These are daunting figures for a party whose membership is overwhelmingly non-Hispanic white.
Beyond the obvious demographic imperative that motivates Republicans to create inroads into these burgeoning minority electorates, political indicators suggest that there is no time to waste. The growing Hispanic vote in California, for example, is believed to have been a major contributing factor in Republican Dan Lungren’s decisive loss to Democrat Gray Davis in the 1998 gubernatorial race. California was not as competitive for the Republicans in the 2000 presidential election as it had been in elections past, also reflecting the pronounced Democratic bias among Latino voters in that state. Other states with large Hispanic populations are tipping more Democratic as well, with Illinois and New York readily coming to mind. Bill Clinton and Al Gore ran very competitive races in Florida (Clinton winning it twice), and the once-reliable Republican advantage among Cubans has dwindled to almost nothing. If Republicans are to remain competitive in these prize states, so the argument goes, they must find a way of attracting this large and growing constituency. As evidence that it is possible to pull this off, Republican optimists often point to the 1998 Texas gubernatorial race, when George W. Bush was said to have won nearly half of the Latino vote, forgetting to add that this "impressive" feat was accomplished only in the face of exceedingly weak opposition — in this lopsided contest, Bush won nearly one-third of all Democratic votes. Does that mean Bush would benefit by devoting countless staff hours to figuring out how to win the votes of regular Democrats? Probably not, but hopes for a Hispanic realignment persist.
These hopes are buoyed by the strong evidence that Bush did outperform Dole among Latinos. According to exit polling data, Bush won about 35 percent of the Hispanic vote, compared with the meager 21 percent that Dole won. Examining the last seven presidential elections, however, does not reveal a decisive trend toward the GOP (see Figure 2). Bush did not do as well among Latino voters as Ronald Reagan did in 1980 and 1984, in spite of much more intense efforts directed their way. Unlike many of his predecessors, Bush was committed to attracting Latino voters and devoted substantial campaign resources to this task. Spanish-language advertising blitzes in Hispanic-rich communities, campaign stops in Latino barrios across the country, and a Latino-centered convention program all illustrate Republican efforts to persuade and mobilize this important constituency. Apparently Bush did halt a steady Latino slide away from Republicans that began with his father’s first election in 1988 (see Figure 2). But only future elections will reveal whether his performance among Latinos was an aberration, and whether Republican voting among Latinos can ever exceed Reagan’s margin. And it would be worth noting that Gore beat Bush among Latinos by 27 percentage points in the vote.
Hispanic Party Loyalty by Voting Intention, Citizens, and Non-Citizens
To understand what the president is up against in this determined effort to convert Latinos, we can examine some data from a July 1999 survey of Latinos conducted jointly by The Washington Post, the Henry J. Kaiser Family Foundation, and Harvard University. This is certainly not the only reliable source of recent data, but this study does have the advantage of polling a large number of citizens and non-citizens from a variety of ethnic and nationality backgrounds within the Hispanic community. Since we know from many volumes of political science research that party identification is one of the most reliable guides for vote choice, we examined the partisan preference of Latinos according to vote intention, registration, and citizenship status. The results in Figure 3 indicate that among likely voters in the 2000 election, Democrats had a 10.7 percentage-point lead over Republicans, 41.8 percent to 31.1 percent. This is a significant margin, but perhaps not insurmountable. The gap grows wider, however, as one moves across the categories in Figure 3. Democrats lead Republicans by 18.6 points among Latinos who are citizens, but are not registered voters. Among non-citizens, the partisan gap is even wider: 39.4 percent to 15.0 percent. Immigration policy has practical political implications, and in this case we can see that generous Hispanic immigration has contributed to a solid Democratic edge in partisan affiliation.
Hispanic Party Loyalty by National Origin
Hispanics are a diverse group, coming from literally dozens of countries with distinct cultural traits and traditions. The vast majority of Latinos in the United States are of Mexican ancestry, with Puerto Ricans, Cubans, Dominicans, and others making up a much smaller share of the pie. In spite of this diversity, however, the partisanship of Latinos tends to show a strong Democratic bias, regardless of national origin.
Table 1 should give President Bush’s political advisers some pause as they ponder the uphill battle they have ahead of them. When we take the Cuban population out, the Democratic advantage ranges from a low of 21.5 points among the sample of Salvadorans to a whopping 57.6 points among Dominicans. Among the most numerous group, Mexicans, Democrats lead Republicans by 22 points, with a large proportion of Mexicans claiming to be independents. Perhaps it is the data on these independents that sparks hope within the White House. But previous research has shown that many of these independents are not truly independent at all, but vote predictably Democratic, according to the prevailing Democratic political climate in the neighborhoods in which they live. They are not up for grabs, in other words — they are Democrats. For their part, Democrats have every reason to believe that they can win over Cuban votes, as the Republican advantage there is a mere six points. But by comparison, Republican hopes to turn these other nationality groups around seem far-fetched, if not fanciful.
Hispanic Party Loyalty by Income, Education, and Tenure
A persistent GOP optimist might argue that Hispanics are even less homogenous by income, education, and duration of residence in the United States than they are by national origin. While Democrats have a considerable advantage across nearly all nationality groups, and among both citizens and non-citizens (see above), surely it must be possible for Republicans to win the votes of wealthy and well-educated Hispanics, or of those immigrants who are more established. Upon this hope, George W. Bush met with Hispanic business leaders throughout his campaign, and has appointed some to important positions within his Administration. Might wealthy and well-educated Hispanics sing with a Republican accent? Our examination of the data suggests that the Democratic advantage in partisanship remains strong regardless of economic standing, and actually increases among Latinos with longer tenure in the country and with greater educational attainment, consistent with the findings of previous research on this subject (Cain, Kiewiet, and Uhlaner 1991).
Consider the figures presented in Table 2. Here we have estimated the percentage of Democrats and Republicans who identify with each of the two major parties for Hispanic voters who fall into a variety of categories. The contrasts are remarkable. For Latinos with less than an 8th grade education, Democratic partisanship exceeds Republican partisanship by 17.7 points. But for those with four years of college, Democrats have an enormous 32-point advantage. For those of short tenure in the country (less than one year), Democrats maintain a 21-point advantage. But as duration of residence rises to 30 years, the Democratic advantage increases to 27 points. These data lead to the bizarre conclusion that Republicans would be better off if Latinos remained uneducated and left the country after a few years.
Table 2 does show that rising income improves the prospects of finding Hispanic Republicans. Among those earning less than $20,000, about 17.4 percent claimed GOP identification, compared with 26.1 percent of those earning over $100,000. And as income rises, Republicans do make up ground on Democrats, but at the highest level of income, Democrats still maintain a 10-point edge in partisan affiliation. In order for the Republicans to eliminate this deficit, Latino incomes would have to stand at well over $200,000 per year. If Bush advisers can discover some means for making hundreds of thousands of Latino citizens this wealthy, then maybe their hopes for Latino realignment are not misplaced. Taken at face value, though, the results in Table 2 are a sobering dose of reality. A similar analysis of data from Knight-Ridder’s 2000 Latino Voter study indicated that income does not have a significant effect on Hispanic partisanship (Alvarez and Garcia Bedolla 2001). Latinos are strong Democrats, even at exceptionally high levels of income, education, and tenure. No realistic policy program is likely to change this in the next four years.
Hispanic Party Loyalty by State
The 1999 Washington Post/Kaiser/Harvard study is not the only resource on the political orientation of Latino voters. We can also turn to the 2000 Voter News Service (VNS) exit polls taken as voters left the voting booths in November. These figures are especially valuable because they provide insight into the voting loyalties of Hispanics who actually showed up at the polls to cast ballots. They also provide some means of studying the stability of Hispanic partisanship across various political settings. In Table 3, we present the proportion of the Hispanic electorate that identified as Republican, Independent, and Democratic in each of the 10 states in which the Hispanic population was 10 percent of the total or larger according to the 2000 Census.
The figures in Table 3 paint a grim picture for Republicans in the states where the Latino vote counts most. Hispanic Democrats who voted in last year’s election hold a commanding lead over Republicans almost everywhere. Only in Florida and New Jersey, both with large Cuban populations, is the contest for partisans even close. In New York, which has large Puerto Rican and Dominican populations, the Democratic lead is an astonishing 62 points. Even in Texas, 42.9 percent of Latinos identified as Democrats, compared with only 16.7 percent who identified as Republicans — a difference of 26.2 points. In the Rocky Mountain West, the Democratic margin ranges from 33 percentage points in Arizona to 47.8 percentage points in Colorado. With the exception of African-Americans, there are few groups in American society where the balance of partisanship is this lopsided. For the White House to make up this kind of ground by 2004 will take far more than tax cuts and school voucher programs.
From Party Identification to Voting
We have shown that the partisan bias of the Latino voter remains heavily Democratic at the turn of the new century, posing a quixotic challenge to Republican strategists bent on turning the partisan tide. While some may suggest that partisanship is not the same as voting, we also know that the Democratic bias in Hispanic partisanship translated predictably into a heavy Democratic vote for Al Gore in the November 2000 election, as exit polls clearly indicated that Gore held an almost 2 to 1 lead among Hispanics nationwide. Nationally, the Bush campaign’s attempt to woo Hispanic voters was mostly a failure, as the Latino bloc showed largely the same pattern of presidential vote choice that they had in previous elections.
Nevertheless, it is possible that Latinos departed from their traditional partisan moorings to vote for George W. Bush and the Republicans in some locations. For evidence of a regional disloyalty to party, one can turn to the state-level exit polls to examine the actual Hispanic vote in each of the 10 states (see Figure 4). According to these figures, Democrats have good reason to rest easy. Even in the nation’s one bastion of Hispanic Republicanism, Florida, where his brother is Governor, Bush eked out a tiny one-point lead. To his credit, Bush did manage to close the gap in Texas to just 10.7 points, but this was in the absence of a serious Democratic campaign in that state. New York was by far the worst state for Bush, but he had little appeal among Hispanics in the Rocky Mountain states either — bastions of conservative Republicanism where Ronald Reagan did well.
Conclusions: More Than an Uphill Battle
How much of the Latino vote is really "in play?" If past is prelude, our results suggest surprisingly little of it. Republicans can count on 20 to 25 percent of the Hispanic vote in most states, regardless of what they do. Obtaining the sizable gains to pull even with Democrats is unrealistic anytime soon. Mexicans, the largest of the nationality groups, look pretty hopeless for the GOP in the near term. Dominicans and Puerto Ricans are even less promising. Winning over Cubans is less of a stretch, but they show signs of drifting into the Democratic column as the generation that came of age in opposition to the Castro revolution is gradually replaced by younger cohorts with shorter memories. Cubans are in play today, but 20 years ago they were loyal Republicans. For Bush, the challenge may be in simply maintaining some Hispanic vote share, not making gains. Republican hopes that higher Latino turnout will help even out support seem misplaced, as our figures show that Hispanic non-voters are even more likely to be Democrats than those who vote.
Republicans may be hoping that the lack of cumulative political experience among many Latinos new to the country is an indication that strong political habits have not yet been formed. They might conclude that there is an opening for them among recent arrivals, since history tells us that Hispanics become only more Democratic with greater tenure in the United States. But they also start out very Democratic, and one wonders what it would take to attract legions of low-income immigrants into the GOP. Other research has shown that Hispanics are attracted to the Democratic Party because they care about the expansion of health care and social insurance programs (Alvarez and Garcia Bedolla 2001; Martinez-Ebers, Fraga, Lopez, and Vega 2000; Straub 2001). The large disconnect between Latino support for a wide array of social welfare programs and the Republican aversion to "big government" solutions will continue to plague Republican efforts at Latino recruitment. One must ask whether Republicans are really willing to change their traditional stands on health care and social welfare programs in order to compete for this constituency. An alliance on cultural issues may create opportunities for enhanced Latino enrollment, but only if these concerns displace social welfare issues in their importance to Hispanic voters.
Our analysis raises questions about whether the GOP attempt to woo Hispanics is worth the enormous effort that it would take to make solid gains. The GOP is in no position to make the kind of policy promises that would be required to bring Latinos over to the Republican side, much less deliver on such promises. And yet, by merely flirting with the Hispanic vote, one risks mobilizing them without converting them — clearly a recipe for Republican disaster in 2004 and beyond. And one must not overlook the fact that Republicans pay significant opportunity costs in chasing after the Latino vote. Their time may be better spent on trying to close the gender gap, or attracting the loyalties of white working-class voters who have regularly shown an independent streak.
There may also be a lesson in our data about legal immigration policy. There can be little question that high levels of immigration have been a boon for Democrats, allowing them to import large numbers of impoverished, mostly unskilled Latino immigrants who, once naturalized, add significant numbers to the Democratic base in the major immigrant-receiving states. The Hispanic population surged by 58 percent between 1990 and 2000, largely through immigration. Between 1980 and 2000, Hispanic voter registration more than doubled. About one-third of the Latino population is under age 18, meaning that many more will be entering the electorate in the next ten years. The sustained growth of the Hispanic population, coupled with its increasing political muscle, will continue to have major political ramifications for the balance of party support, tipping once-competitive states to the Democratic side and turning once-solidly Republican states competitive. Immigration reduction, from this point of view, is likely to be one of the few pathways to Republican survival in California, and even Texas.
In light of the assembled facts, the Republicans may want to think long and hard about current proposals to grant amnesty to immigrants illegally living in the country. From a Republican standpoint, legalizing the status of illegal aliens is another path to electoral defeat. A Republican vote in favor of amnesty is likely to be read by Latino political leaders as a "mixed signal" — not enough to demonstrate true commitment. Once naturalized and mobilized, these voters will be almost certain Democratic constituents, regardless of how Republicans vote on amnesty.
One might ask if Republicans run the risk of losing their significant bloc of loyal Latino supporters by upholding a more restrictionist position on immigration policy. The odds of this are small. Analysis of several public opinion polls has indicated that Hispanics in the United States are themselves divided on immigration policy. Latinos may be more favorable to liberal immigration policies than Anglos or blacks, but that does not mean they are unified. Older Hispanics, those who are better educated, and those who reside in areas with large illegal immigrant populations are especially likely to favor more restrictive legal immigration policies (Hood, Morris, and Shirkey 1997). David O. Sears and his colleagues, working from Los Angeles-area polls from the mid-1990s, show that nearly half of Latino respondents favored a decrease in the number of legal immigrants and "many wanted to delay access to government services for legal immigrants and force children to make the transition into English-language instruction" (Sears, Citrin, Cheleden, and van Laar 1999, 52). A large proportion of Latinos do not take "group-interested" positions on policy issues, preferring to think of themselves as "Americans" and not "Hispanics" or "Latinos." The upshot of this line of research is that President Bush and the Republicans are probably overestimating the extent of Latino consensus in support of high levels of immigration. Latinos are strongly attracted to the Democratic Party not because of that party’s positions on immigration policy, in isolation, but because the Democrats are in line with Latino policy preferences on education, health care, and social services, firmly entrenched positions that neither the Democratic Party nor the GOP will change in the foreseeable future.
If the past tells us anything, then the prospects of a widespread Latino conversion to the Republicans are more fantasy than reality. The GOP is not operating in a vacuum. There is another party out there actively courting the Latino vote with a platform far more consistent with their interests. For the GOP to shift its policy ground to the extent necessary to compete for this vote would be tantamount to trying to reverse the New Deal realignment of the 1930s. Presidents and party leaders are, of course, free to pursue their dreams, but this one could well turn into a nightmare — leading to another one-term Bush presidency.
R. Michael Alvarez and Lisa Garcia Bedolla. 2001. "The Foundations of Latino Voter Partisanship: Evidence from the 2000 Election." Paper presented at the Annual Meeting of the 2001 Western Political Science Association Convention, Las Vegas, Nevada, March 15-17.
Bruce E. Cain, D. Roderick Kiewiet and Carole J. Uhlaner. 1991. "The Acquisition of Partisanship by Latinos and Asian Americans." American Journal of Political Science 35:2: 390-422. Lourdes Cue. 2001. "Election 2000: The Latino Factor." Hispanic 14:1 (January/February): 24-27.
Rodolfo O. de la Garza and Louis DeSipio, eds. 1999. Awash in the Mainstream: Latino Politics in the 1996 Elections. Boulder, CO: Westview Press.
Louis DeSipio. 1996. Counting on the Latino Vote. Charlottesville: University of Virginia Press.
Karen Foerstel. 2000. "Parties Maneuver for Identity in Hopes of Swinging 2000." CQ Weekly Report 58:1: (January 1) 8-11.
M.V. Hood, Irwin L. Morris and Kurt A. Shirkey. 1997. "¡Quedate o vente!":Uncovering the Determinants of Hispanic Public Opinion Toward Immigration." Political Research Quarterly 50: 627-647.
Michael Jones-Correa and David L. Leal. 1996. "Becoming ‘Hispanic’: Secondary Panethnic Identification Among Latin American-origin Populations in the United States." Hispanic Journal of Behavioral Science 18: 214-254.
Valerie Martinez-Ebers, Luis Fraga, Linda Lopez and Arturo Vega. 2000. "Latino Interests in Education, Health and Criminal Justice Policy." P.S.: Political Science and Politics 33:3: 547-554.
Gregory Rodriguez. 1999. "Latino Republicans and the 1998 Election." Aztlan 24:1: 175-184.
David O. Sears, Jack Citrin, Sharmaine V. Cheleden and Colette van Laar. 1999. "Cultural Diversity and Multicultural Politics: Is Ethnic Balkanization Psychologically Inevitable?" In Deborah A. Prentice and Dale T. Miller, eds. Cultural Divides: Understanding and Overcoming Group Conflict. New York: Russell Sage.
Bill Straub. 2001. "Bush’s Latino Problem." Scripps Howard News Service. April 27.
Carole J. Uhlaner and F. Chris Garcia. 2001. "Learning Which Party Fits: Experience, Ethnic Identity and the Demographic Foundations of Latino Party Identification." Paper presented at the conference on Minority Representation: Institutions, Behavior and Identity, Claremont Graduate University.
James G. Gimpel is an Associate Professor and Karen Kaufmann is an Assistant Professor of Government at the University of Maryland, College Park.