The political environment that historically marginalised Latinos has given way to a regime that includes significant obstacles to electoral access but also offers extensive opportunities for them to become significant political actors. Taking advantage of this new environment requires understanding the new Hispanic demographics, the viability of the ethnic and pan-ethnic as mechanisms for mobilising Latinos, the real and potential impact of the Latino vote, and how Hispanic political leaders should confront the future.

The New Demographics

A continuously increasing population is making Latinos a major political constituency that will only become more significant in the near future. Because of immigration, Hispanic numbers increased from 9.6 million in 1970 to 38.8 million in 2002, when they became the nation’s largest minority. This growth is the foundation supporting the claim that Latinos are an electorate whose demands have to be met; however, the assertion illustrates a misunderstanding of how immigrants affect American political life.

The Constitution requires that all US residents be counted in the Census whether they are native born or naturalized citizens, legal resident aliens or undocumented immigrants. Immigrants, consequently, have influenced reapportionment to the benefit of Hispanics (Poston et al., 2003). Since 1980 their presence has led to increases in the number of congressional seats allotted to Arizona, California, Texas, Florida and New Jersey, and thanks to the Voting Rights Act (VRA) many of these new districts have been designed to virtually ensure that they would elect Latinos.

These increased numbers have transformed Hispanics from a regionally concentrated but relatively isolated population into a national minority (de la Garza, 1992, p. xiii), but even though their share of the national population almost doubled during the past twenty years, Latinos only totalled less than 13% of the national population in 2000. Moreover, even that number overstates Latino electoral clout because it includes immigrants who cannot or do not vote and who make up 40% of this population. The Hispanic electorate as a political entity is further reduced by the presence of immigrants who do not identify as or interact politically with Hispanics, as well as by increasing numbers of the new Latino mestizos, ie, the children of Hispanic-Anglo marriages who have only the most tenuous ties to the group.

Although immigrants have also been the key element in the increase in the number of state legislative and local districts likely to elect Latinos, it must be recognised that local level officials and state legislators are elected from two types of districts. One consists of towns and districts in areas such as Northern New Mexico and South Texas, where Latinos have long been numerically dominant and thus do not owe their positions to recent population increases; the other is made up of jurisdictions from California to Massachusetts in which immigration has made Latinos the majority population.

While the latter exemplify why the number of district-based elected officials has and will continue to expand in future decades, the increase in such districts has been based more on VRA requirements than on politics and has not substantially increased Hispanic influence in major political arenas. This is because immigrant-based districts do not require Hispanic officials either to mobilise their constituents or develop strategies for reaching out to non-Hispanic constituencies. The experiences gained from campaigning in such homogenous districts do not serve Hispanic officials well when they seek higher office with heterogeneous constituencies, as is illustrated by the losses Latinos experienced in recent mayoral races in Los Angeles and New York, cities that had the demographic and partisan characteristics that should produce Hispanic victories.

This problem also affects Latinos when they seek major state-wide offices. As Table 1 shows, only New Mexico has enough Latino citizens to regularly play a major role in the election of governors or US senators which, except for the Presidency, are the most significant elected offices in the nation. Thus, it is not surprising that the only Latino currently holding a top level state-wide office is the Governor of New Mexico. Latinos are, however, elected to important but lower level state-wide positions (such as the Attorney General of Colorado), or to offices that are either symbolic, lack political authority or are of very low public saliency (such as the Lieutenant Governor of California and the New Mexico Secretary of State).

Table 1. Hispanic and Foreign-Born in States with Large Hispanic Populations in 2000

State	Total Population	% Hispanic	% Adult Non-citizen Hispanics
Arizona	5,130,632	25.3	32.3
California	33,871,648	32.4	61.8
Colorado	4,301,261	17.1	27
Florida	15,982	16.8	41.5
Illinois	12,419,293	12.3	48.1
Massachusetts	6,349,097	6.8	40.7
New Jersey	8,414,350	13.3	12.0
New Mexico	1,819,046	42.1	36.9
New York	18,976,457	15.1	28.1

The continued failure to seriously contest or win state-wide elections is a clear indicator of the limited political impact of population growth.

In short, while increased numbers have been instrumental to local and district-based electoral victories, they do not automatically enhance Hispanic abilities to win major state-wide offices and become influential in national elections. Rather than point to their increased numbers as proof of new-found political power, therefore, Latino leaders would do well to convert this growth into a genuine political asset. Among other things, this requires developing naturalisation campaigns, a need only recently recognized by Latino leaders (Pachon et al., 2004) but one that would greatly increase the size of the Latino electorate and would therefore provide a powerful incentive for non-Latino groups to build strong bridges to Latinos.

To summarize, demography is not destiny. Increased population will not automatically make Hispanics major political players in American politics. Instead, their role will depend on factors such as the extent to which they act cohesively, engage the political system and develop coalitions with other groups. Also, factors beyond their control, such as the willingness of other groups to work with Hispanics, will also weigh heavily on the political future of Latinos.

Ethnicity and Electoral Mobilisation

Immigration diversified the Hispanic population as it increased it. Before the 1980s, the overwhelming majority of those with Spanish surnames in the nation belonged to the Mexican-origin population in the Southwest, with a much smaller number of Puerto Ricans in the northeast, and even fewer Cubans in Florida. Primarily because of geographical factors, these groups interacted at very low levels (de la Garza et al., 1992, p. 67-68). Beginning in the 1980s, immigrants from Central and South America and the Caribbean began arriving and settling primarily in cities such as Los Angeles, Houston, New York and Miami, which had well established communities of Mexican, Puerto Rican or Cuban origin. Their presence changed the character of the Hispanic community at both the national and local levels and increased inter-group interactions which gave rise to Hispanic/Latino politics, ie, claims that a pan-ethnic Hispanic community that superseded communities built around distinctive nationalities such as Dominicans and Cubans had developed a political agenda that all nationalities endorsed and towards which they would mobilise.

Those who advance this perspective fail to recognise that Latino identity is constructed in the US, not in Latin America. The immigrants we identify as Hispanics/Latinos come with well developed national identities. Assuming that they will quickly or easily abandon their traditional nationalistic identities in favour of a homogenised American identity fails to acknowledge the emotional strength of nationalism (Greenfeld, 1992). Asking individuals to subjugate their historical national identity for a pan-ethnic one, ie, to cease identifying themselves as Dominican in favour of identifying as a Latino might succeed over time and especially across generations as home-country attachments are attenuated (de la Garza et al., 1997; Pachon et al., 2000), but for the immigrant generation they stand as enduring impediments to the creation of a new identity capable of serving the same psychological functions as ethnicity (Smith, 1991).

The evidence regarding the limited efficacy of pan-ethnic claims is abundant. As of 1990, US citizens of Mexican, Puerto Rican and Cuban origin were more likely to state that the three groups were ‘not very similar’ culturally than that they were ‘very similar’, and each group reports feeling closer to Anglos (non-Hispanic whites) than to either of the other two groups. This reflects both the historical differences among the groups and their minimal interaction with each other within the US (de la Garza et al., 1992, p. 67-69).

The salience of nationalistic versus pan-ethnic attachments is further illustrated by the extent to which country-of-origin labels are preferred over terms such as Hispanic or Latino. In 1990, 73% of Mexican Americans, 75% of Puerto Ricans and 77% of Cuban Americans preferred to be identified in terms of their national origin and only 21%, 15% and 13% of the respective groups preferred pan-ethnic terms (de la Garza et al., 1992, p. 40). By 2002, among the US born, national labels declined in popularity and ‘American’ rather than pan-ethnic labels became the identity of choice (see Table 2). As in 1990, however, the ethnic label most Latinos preferred referred to their country of origin, a pattern that is especially prevalent among immigrants.

Table 2. Preferred Identity of Latinos in the United States, 2002

Preferred Identity (%)	Total Latinos	Native-born Latinos	Foreign-born Latinos
Respondent/parent country of origin	54	29	68
Latino/Hispanic	24	23	24
American	21	46	6

Source: Pew Hispanic Center/Kaiser Family Foundation 2002 National Survey of Latinos.

It is also important to note that those who in 1990 identified exclusively in pan-ethnic terms were likely to be the most assimilated and the least likely to maintain strong ethnic attachments (Jones-Correa and Leal, 1996). For example, those who identified exclusively as Hispanic or Latino are more supportive of using English and less supportive of bilingual education than those who primarily or exclusively utilise national-origin labels. This suggests that ‘Hispanics/Latinos’ would be the least likely to mobilise in support of pan-ethnic political appeals. It is reasonable to expect pan-ethnic identifiers in 2002 to share this proclivity. This pattern, combined with the dramatic increase in those who identify as Americans, raises questions about the long term efficacy of using pan-ethnic appeals to politically mobilise Hispanics.

The limited utility of pan-ethnic appeals may be related to the more fundamental weakness of ethnic appeals per se. Latinos do not automatically rally in support of co-ethnic candidates. In 1990 over 60% of Mexican Americans, Puerto Ricans and Cuban Americans said they were not more likely to vote when a co-ethnic was a candidate, and that they voted for the ‘best candidate’ rather than the ethnic candidate when a co-ethnic and an Anglo compete. Nonetheless, when given an option between a co-ethnic and an Anglo, at least 77% of each Hispanic group support the former (de la Garza et al., 1992, p. 138). This choice, it must be emphasized, does not necessarily contradict their initial preference. Rather, it may be that co-ethnic candidates also share the policy preferences signalled by other factors such as the partisanship of the several Latino electorates. Thus, such co-ethnic candidates would be preferred because of shared political characteristics rather than because of shared ethnicity. This explains the failure of even well funded Hispanic Republicans who run against Anglo Democrats, as illustrated by the case of a Cuban mayoral candidate in 2003 and of a well known Mexican American Republican in 2000 in California’s heavily Mexican-American 20th district (Michelson, 2002b).

Further evidence of the limits of ethnic appeals is the inability of Latino candidates to rely on co-ethnicity to mobilise voters who will lead them to their election. Examples include Antonio Villaraigosa’s 2001 mayoral campaign in Los Angeles, Cruz Bustamante’s gubernatorial effort in California’s recall election of 2003 and Tony Sanchez’s 2002 gubernatorial campaign in Texas. All of these candidates seem to have assumed that Hispanics would mobilise on their behalf and provide the foundation for victory. While each received a majority of the Latino vote, in no case did Latinos turn out at unusually high rates. Analysing these and other cases from as far back as the 1980s leads researchers to conclude that while ethnicity directly affects partisanship, it does not have a direct impact on vote choice (Cain et al., 1991; Graves and Lee, 2000).

When ethnic and partisan appeals compete, the latter trumps the former; when they reinforce each other, there is a strong likelihood of high and unified ethnic or pan-ethnic mobilisation as in 2001 with Fernando Ferrer’s mayoral campaign in New York. Indeed, to the surprise of virtually all political analysts, Hispanic support for Ferrer was so substantial that he unexpectedly was able to force a run-off to determine the Democratic nominee.

Ferrer’s success, limited though it was since he did not win the run-off, suggests other factors that are needed to make pan-ethnic appeals effective. One is a political culture that emphasises historical exclusion. Latinos in New York, especially the Puerto Ricans, have long been at or near the bottom of the city’s social pyramid, and they have never seriously contested major city-wide or state level elections. Ferrer’s campaign emphasised class rather than ethnic divisions in ways that highlighted Puerto Rican and Dominican exclusion without explicitly raising it. Moreover, his approach made it unnecessary to him to downplay his Puerto Rican origins in order to win Anglo support in the way that Villaraigosa was accused of doing. It also prevented ethnic divisions because it did not cater to one specific group. Neither Bustamante nor Villaraigosa or Sanchez implemented this type of outreach.

Combining all of these examples suggests the conditions under which pan-ethnic (or ethnic) appeals are likely to be most effective. First, they must be reflective of constituency experiences, interests and objectives rather than be primarily instrumental. Historically, ethnic appeals focus on responding to discrimination. An authentic pan-ethnic movement today could successfully emulate such efforts if discriminatory exclusion were at the core of Hispanic life. This is not the case, however. While in 1990 a majority of Mexican Americans and Puerto Ricans and 47% of Cubans reported that their group experienced a lot or some discrimination (de la Garza et al., 1992, p. 94-95) less than half as many reported that they themselves had been victims of discrimination. Even this pattern was attenuated by 2002 (see Table 3), when most Hispanics either identified discrimination as a minor problem or did not consider it a problem at all, and fewer than half indicated they had been discriminated against (Pew Hispanic Center, 2002, p. 32). Other recent sources report even lower incidences of discrimination (Uhlaner & Garcia, 2002; Michelson, 2000a). In sum, discrimination does not appear to be so pervasive as to motivate Hispanic citizens to unite in response to generalized pan-ethnic appeals. Thus it does not foster a sense of linked fate among Hispanics sufficiently strong to energise them politically as it does among African Americans (Dawson, 1994). However, given that immigrants are much more likely than the native born to see discrimination as a major issue there is the potential for such a shared attitude to develop.

Table 3. Perceptions of Discrimination, 2002

	Problem	Not a	Don’t
(%)	Net	Major	Minor	Problem	Know
Total Latinos	75	38	37	21	4
Native-born Latinos	71	26	45	25	4
Foreign-born Latinos	77	45	32	18	5
Non-Latino Whites	54	13	41	30	16

Source: Pew Hispanic Center/Kaiser Family Foundation 2002 National Survey of Latinos.

Further dampening the prospects of rallying Hispanics around discrimination claims is that they describe inter-group discrimination as more pervasive than discrimination by Anglos. In 1990 only 55% of Mexican Americans and Puerto Ricans reported that co-ethnics helped rather than ‘pulled each other down’ (de la Garza et al., 1992, p. 132). By 2002 83% of Latinos reported that discrimination by other Latinos was a problem, and of those who voiced this view, the majority described such discrimination as a major problem (see Table 4). By comparing Tables 3 and 4 we can see that inter-ethnic discrimination was perceived as more pervasive and significant than discrimination by Anglos.

Table 4. Perceptions of Discrimination by other Latinos, 2002

	Problem	Not a	Don’t
(%)	Net	Major	Minor	Problem	Know
Total Latinos	83	47	36	16	1
Native-born Latinos	73	29	44	24	3
Foreign-born Latinos	88	57	32	11	1

Source: Pew Hispanic Center/Kaiser Family Foundation 2002 National Survey of Latinos.

Two aspects of these patterns are puzzling. First, why do majorities of Latinos report widespread discrimination by Anglos while so relatively few indicate that they personally have been victimised by such practices? My hypothesis is that among many of the native born it reflects the deep history of discrimination experienced by family members, friends or the group per se. In my research, when asked to describe their discriminatory experiences, many Mexican American respondents would say things like, ‘when my father applied for a job’, or, ‘look at how the Dallas police treated the young Chicano they arrested’. This suggests that a sense of discrimination is part of Hispanic political culture rather than part of the lived experience of Latinos. The second anomaly is that immigrants are more likely than the native born to consider discrimination a major problem. This may be because of the anti-immigrant climate within which they live and work. California’s Proposition 187 is an extreme but not isolated manifestation of this. It may also be that they are largely restricted to the lowest tiers of the labour market and suffer all the abuses associated with such jobs. Without in any way diminishing the negative consequences of discrimination on any individual or the validity of how discrimination is perceived, what must be emphasised is that at a societal level the overall extent and impact of discrimination does not seem to be sufficient to serve as a rallying point for ethnic or pan-ethnic mobilisation.

A final indicator of why pan-ethnic appeals are unlikely to be very successful is that the issues about which Latinos are most concerned are not explicitly linked to ethnicity. From 1990-2004, surveys indicate the issues about which Latinos were most concerned were the economy and education rather than with discrimination or similar issues (Washington Post/Univision/Tomás Rivera Policy Institute, 2004; Tomás Rivera Policy Institute, 2000; de la Garza et al., 1992, p. 88-89). Table 5 illustrates a typical list of priorities.

Table 5. Hispanic and Anglo Issue Priorities, 2000

Issues (%)	Total Latinos	Native-born Latinos	Foreign-born Latinos	White Non-Latinos
Education	40	35	48	20
The economy	17	18	15	21
Social security	7	8	7	8
Taxes	3	4	1	5
Crime	4	4	3	2
Moral values	6	8	5	13
Abortion	2	3	–	4
Health care and Medicare	7	7	7	9
National defence/defence readiness	6	7	5	10
Defending civil rights	4	4	3	3
HIV/AIDS	2	1	2	–
Don’t know	3	3	3	2

Source: Pew Hispanic Center/Kaiser Family Foundation 2002 National Survey of Latinos.

More noteworthy, not only do ethnic issues in general have very low salience, but issues that would seem especially relevant to Latinos like immigration and affirmative action also are of low priority (de la Garza et al., 1992, p. 88-89; Pew, 2002, p. 42; Tomás Rivera Policy Institute, 2000; Washington Post/Univision/Tomás Rivera Policy Institute, 2004). The most notable exception to this is that in 2000 12% of Latinos in key states cited race relations as a major concern (Tomás Rivera Policy Institute, 2000).

It is also important to recognise that pan-ethnic claims are most likely to be effective in national elections which involve all national origin groups or in states like New York and Florida where no one nationality group dominates the Hispanic population. When pan-ethnic appeals are made in states like Texas, where one group dominates, it is because of two reasons. First, the political rhetoric of such states (and the nation) is much more accepting of neutral labels such as Hispanic than of group-specific terms such as Puerto Rican or Mexican American that are linked to local historical conflicts. Second, by using pan-ethnic terms, Latino political leaders not only hope to rally all Hispanic nationalities behind their cause but they also seek to invoke the spectre of a national electorate that will mobilise within their specific contexts to developments affecting Hispanics in other states. In other words, the expectation is that politics conducted within a Hispanic framework will link Colorado’s Mexican Americans to Puerto Ricans in New York, so that if New York Democrats act against Puerto Ricans, Mexican Americans will mobilise against the Democratic Party in Colorado. This is no evidence that this occurs, however. To the contrary, state and local institutions seem to be more relevant to the evolution of Latino political life than national origin or major demographic characteristics (de la Garza, 2004).

To be most effective, then, pan-ethnic appeals should be camouflaged within campaigns that emphasise the partisan preferences of constituents. In other words, in Democratic areas, such appeals are most effective if they are linked to Democratic candidates and issues; the same pattern might not hold among Republican Cubans in Florida because their agenda remains dominated by anti-Castro issues. Thus whichever party’s candidate deals most effectively with Cuban policy is likely to be supported by Cuban American voters. This is why President Clinton, who supported Cuban American views towards Castro, had so much success in Miami in 1996 (Moreno and Warren, 1999). Second, pan-ethnic appeals are likely to be most successful in areas where the political culture emphasises ethnic-based exclusion. This is why in districts where Latinos have long been candidates but never won, turnout often greatly increases the first time ethnic candidates have a real chance of winning. This was seen across Texas in the late 1970s and early 1980s as well as in New York in 2001. Once Latinos begin to win offices in such areas, the effectiveness of ethnic appeals declines. Finally, as is true of most successful political efforts, such appeals are most successful when led by effective politicians. In sum, as traditional barriers to voting have come down and the number of elected Latinos has increased, the political relevance of ethnicity and pan-ethnicity has declined and candidate skills and issues have become more important.

Hispanic Partisanship

Latino leaders claim that the Hispanic vote is so large that it can determine the outcome of any given national election (de la Garza, 1996). Whichever party ignores Latino demands therefore jeopardises its electoral prospects. This implies Hispanic partisanship is primarily instrumental rather than ideological, that is, that it is driven by temporary self interest rather than by strong policy preferences. Partisanship among Anglos, however, is much more ideological than instrumental (Green et al., 2002), and there is no reason to think that Latinos are exceptions to that pattern. Given that Latinos perceive the Democratic Party to better represent the working class and minorities and that the great majority of Latinos have historically identified with the Democrats, they are likely to continue to do so, just as Cubans will likely remain within the Republican camp because Republicans are more supportive of their ideological opposition to the Castro regime.

The assertion that Latinos vote instrumentally gives rise to the quadrennial prediction that Latinos will be the decisive swing vote in the presidential election. The meaning of ‘swing vote’ is unclear, however. It may mean that the Latino vote ‘swings the election’ to a specific candidate and thus determines the outcome of the election. In this scenario, Hispanics are assumed to adhere to established partisan preferences and influence the election with a highly cohesive vote. ‘Swing vote’ may also refer to a group that casts its votes instrumentally so that in any given election it will vote for whichever candidate has made the strongest commitments to the group. In 2000 ‘soccer moms’ became such a group.

The argument that the Latino vote can determine national electoral outcomes is intended to persuade party leaders to court Hispanic voters. Our analysis of election outcomes from 1988 through 2000 indicates this is rarely the case (de la Garza and DeSipio, 2004). The role that Latinos have played in presidential elections may be conceptualised into four types of elections, three of which reflect actual outcomes, and one which indicates the extreme and unrealistic conditions which would have to be met for the Hispanic vote to have affected the actual electoral results in states where they could have been influential:

·         Type 1. Elections in which Hispanic votes were arguably the decisive bloc contributing to the actual election result. This occurred in Arizona in 1996.

·         Type 2. A very close election where Latinos can claim to have contributed the winning vote as was the case in Florida and New Mexico in 2000 (and Florida in 2004). While such claims generate political capital, elections like these are so close that virtually any group can claim to have been the key to the final outcome.

·         Type 3. Elections in which Latinos did not influence actual results, or in which the Hispanic vote could not be reconfigured theoretically in any way so as to influence the results of the election with the specific states analysed.

·         Type 4. Hypothetical situations which will never become real but which illustrate types of extreme conditions that would have to be in place for the Latino vote to matter. These include elections in which no Latino votes, or elections in which the outcome would have changed if only those Latinos who supported the losing candidate had voted.

It must be emphasised that the patterns shown in Table 6 are primarily a function of how non-Latinos vote. If state-level elections become much more competitive within the Anglo electorate, Latinos will be in a better position to be significant players. While such changes may come at some time in the future, there is no sign they are likely to be widespread in the foreseeable future. The Republican strategy for 2004 was designed to focus on states that included few Hispanics (Seelye, 2003). This indicates that Republicans are unlikely to need or seriously compete for their vote in those states with large Hispanic populations outside of Florida, and in the absence of such competition Democrats are unlikely to need to invest much to once again win their support in states such as California, Illinois or New York. Thus, politically, 2004 will probably be a repeat of 2000, and that means that Hispanics will again be on the sideline (as preliminary analyses of the election’s results confirm).

Table 6. Influence of the Hispanic Vote in Presidential Elections, 1988-2000

	1988	1992	1996	2000
Arizona	Type 3	Type 2 Latino Rep voters ensure Bush win	Type 4Rep victory in no Latino votes	Type 3
California	Type 3	Type 3	Type 3	Type 3
Colorado	Type 3	Type 2Latino Dem voters ensure Clinton’s victory	Type 3	Type 3
Florida	Type 3No Latino influence	Type 4 Democrat victory if no Latino voted	Type 3No Latino influence	Type 2 Republican 537 vote win includes Latino majority
Illinois	Type 3No Latino influence	Type 3No Latino influence	Type 3No Latino influence	Type 3No Latino influence
New Jersey	Type 3 No Latino influence	Type 4Rep victory if no Latino voted	Type 3No Latino influence	Type 3No Latino influence
New Mexico	Type 3No Latino influence	Type 4Rep victory if no Latino votes	Type 4Rep victory if no Latino votes	Type 2Dem 365 win includes Latino majority
New York	Type 4Reps win if no Latino votes	Type 3No Latino influence	Type 3No Latino influence	Type 3No Latino influence
Texas	Type 3No Latino influence	Type 4Rep Latino voters ensure Bush win	Type 3No Latino influence	Type 3No Latino influence

Note: type 1 – decisive influence; type 2 – contested influence in very close election; type 3 – no real or hypothetical influence; type 4 – extreme conditions required for Latinos to have influence.

Source: DeSipio and de la Garza, 2004.

The claim that Hispanics vote instrumentally rather ideologically is also unsupported by data. Hispanic partisanship has remained quite consistent over time. Cuban Americans have been Republicans since the 1960s while Mexican Americans and Puerto Ricans, who make up approximately 80% of all Latinos, have been Democrats. Thus, Latino Democratic identifiers have hovered at around 60% of the electorate since 1980 while Republican identifiers have never exceeded 30%. Indeed, no survey finds more than 25% Republican identifiers within the total Hispanic population.

Several factors help explain why it is unlikely that Latino will swing from one party to another. Most significant of these is that Latino partisanship primarily reflects social and political rather than economic factors (Uhlaner & Garcia, 2002; Alvarez & Bedolla, 2001). In other words, Hispanics do not tend to become Republicans as their incomes rise. This is clearly illustrated by the analysis of Hispanic policy and partisan preferences in Southern California (Barreto & Woods, 2003). Latino attitudes toward issues such as abortion, illegal immigration, affirmative action, government sponsored health insurance and gun control, which are central to the Republican Party and on which Hispanics and Republicans disagree, also indicate that policy preferences rather than socioeconomic status explain Latino partisanship. For these reasons it seems clear that current partisanship patterns are likely to persist unless the parties substantially change their positions on these issues, which is unlikely (Alvarez and Bedolla, 2001).

Hispanic immigrants add to the number of Hispanic Democrats. This was evident in the 1980s (Cain et al., 1991) and continues today. The recent increase in registered voters in Southern California which is driven in part by the increase in naturalised citizens adds to the gap between Latino Democrats and Republicans. In California in 1998, Latino registered Democrats outnumbered Republicans by almost 4:1, a substantial increase from their 3.4:1 advantage in 1992 (Barreto & Woods, 2003). This pattern is so clear that scholars supported by the Center for Immigration Studies, a conservative think-tank, have argued that Republican efforts to woo Hispanic Democrats are a resounding failure (Gimpel & Kaufmann, 2001).

Also illustrative of the strength of these partisan patterns is that they reflect Hispanics’ knowledge of the ideological differences between Democrats and Republicans. The majority of Hispanics describe Democrats and Republicans in the same terms as does mainstream America (de la Garza et al., 1992, p. 128-129), and Latino citizens are informed about the positions candidates hold on key policy issues (Nicholson et al., 2002). Given their historical ties to the Democratic Party, it is not surprising that 70% of Gore voters held views on issues such as abortion, gun control and school vouchers that were consistent with those of their preferred candidate, compared with 51% of Bush voters who held positions consistent with Bush’s. Furthermore, compared with Gore’s supporters, Bush voters were less knowledgeable about his policy positions. This suggests that a substantial proportion of the support Hispanics gave to Republicans in 2000 was driven by Bush’s personalistic appeal rather than because of his political agenda. To the extent that this is the case, future Republicans will have difficulty emulating his modest success.

Further complicating claims that Latinos are likely to swing to the Republicans is that except in Florida the Hispanic electorate has moved closer to the Democrats in recent years, as Table 7 shows. In several states, increases in pro-Democratic sentiment triple the changes in pro-Republican evaluations.

Table 7. Changing Affinity towards Political Parties

(%)	California	Florida	Illinois	New York	Texas
Much closer to Rep. than before	4.75	21.55	4.7	4.7	8.6
Somewhat closer to Rep. than before	4.75	10.53	3.96	3.96	7.86
Much closer to Dem. than before	18.25	10.78	21.29	21.29	17.2
Somewhat closer to Dem. than before	12	5.01	9.16	9.16	10.32
No change in feelings	58	46.12	55.2	55.2	50.37
Don’t know	2.25	6.02	5.69	5.69	5.65
Total	100	100	100	100	100

Source: de la Garza & Cortina, 2004 with data from TRPI (2000).

Additionally, in California, Illinois and New York Republican identifiers are more likely to have voted for Gore than Democrats are to have voted for Bush. In Texas, where Latino Democrats outnumber Republicans, the reverse is true, as is also the case in Florida where Latino Republicans outnumber their Democratic counterparts.

Table 8. Latino Vote in the 2000 Presidential Election by Partisanship

(%)	StrongDemocrat	Not VeryStrongDemocrat	IndependentClose toDemocrat	Independent	IndependentClose toRepublican	Not VeryStrongRepublican	StrongRepublican	Total
California
Bush	4.52	18.28	10.00	50.00	100.00	78.26	95.00	27.12
Gore	95.48	81.72	90.00	50.00	0.00	21.74	5.00	72.88
Florida
Bush	6.25	36.17	15.79	72.41	100	94.05	98.94	66.85
Gore	93.75	63.83	84.21	27.59	0	5.95	1.06	33.15
Illinois
Bush	4.69	16.8	25	36.36	76.92	75	90.91	24.3
Gore	95.31	83.2	75	63.64	23.08	25	9.09	75.7
New York
Bush	3.42	4.72	0	40	66.67	63.64	73.68	11.73
Gore	96.58	95.28	100	60	33.33	36.36	26.32	88.27
Texas
Bush	11	36.11	25	66.67	76.92	100	95.65	44.12
Gore	89	63.89	75	33.33	23.08	0	4.35	55.88

Source: de la Garza & Cortina, 2004 with data from TRPI (2000).

Additionally, when we analyse the candidate preferences in California, New York and Illinois of voters who reported they felt much closer to Republicans than they had previously, more would have voted for Gore than for Bush. In Florida, regardless of their feelings towards political parties Latinos were more likely to vote for the Republican candidate than for the Democratic candidate (de la Garza and Cortina, 2004).

In sum, it is unlikely that in the foreseeable future Latinos will swing from their Democratic preferences to support Republicans. Based on the results of 2000, the only state where Latinos might be considered swingers is Texas. However, those results may reflect support for a fellow Texan rather than swing voting. Future analysis will clarify this question.

The Hispanic Vote

There is no doubt that the principal political resource available to Latinos is the vote. Given their concentration in key states, they can influence the outcome of presidential and state level elections when these are competitive provided they turn out in large numbers in support of one candidate (Guerra, 1992). There are few such opportunities, however, and even when they present themselves, Latino turnout is seldom maximised so that they are seldom central to the outcome of the election. Thus, a fundamental question about Latino voting is why it persists at levels significantly lower than that of Anglos.

Hispanic voting is puzzling for two reasons. First, even after controlling for socioeconomic characteristics that are the principal factors explaining Anglo voting, Latinos vote less than Anglos (Michelson 2003b). Also, the influence of factors such as age and education, whose impact on Anglo voting is consistent over time, is much less clear regarding Hispanics (Arvisu & Garcia, 1996; DeSipio, 1996b; Hritzuk & Park, 2000).

To explain Latino turnout it is necessary to consider the role of additional variables. Puerto Ricans and Mexican Americans, like Anglos, involved with organizations vote at much higher rates than those who are not (Diaz, 1996). However, Latino organisational membership rates are so low (de la Garza et al., 1992; Verba et al., 1995; DeSipio et al., 2003) that while organizational membership boosts individual voting rates, it does not significantly increase turnout for Hispanics as a group.

Overall, one possible explanation for the group’s low voting rates is the continued expansion of the electorate by naturalised immigrants and their US-born children. From 1976 to 2000 increases in the number of non-voting citizens outpaced increases in voters (see Table 9). The total citizen population includes naturalised Hispanics who vote at lower rates than the native born (Bass & Casper, 1999). This is predictable given that they have the demographics associated with low turnout, they reside in neighbourhoods that candidates tend to ignore (de la Garza et al., 1994) and consequently they and their children are not well socialised into American electoral politics. In New York, as the immigrant population increased, turnout declined, a pattern that was even more prevalent in Los Angeles (Mollenkopf et al., 2001).

Table 9. Latino Adult Voters, Adult Citizen Non-Voters and Adult Non-US Citizens, 1976-2000

	Latino Adult
	Latino US Citizen
Year	Latino Voters	Latino Non-Voters	Non-US Citizen
1976	2,098,000	2,620,000	1,876,000
1980	2,453,000	3,112,000	2,645,000
% change 1976-80	16.9	18.8	41.0
1984	3,092,000	3,622,000	3,027,000
% change 1980-84	26.0	16.4	14.4
1988	3,710,000	4,368,000	4,815,000
% change 1984-88	20.0	20.6	59.1
1992	4,238,000	4,540,000	5,910,000
% change 1988-92	14.2	3.9	22.7
1996	4,928,000	6,281,000	7,217,000
% change 1992-96	16.3	38.3	22.1
2000	5,934,000	7,224,000	8,440,000
% change 1976-00	182.8	175.7	349.9

Source: Louis DeSipio and Rodolfo de la Garza 2004; Muted Voices: Latinos and the 2000 Elections, Rowman and Littlefield, Boulder, Colorado.

Among Cuban Americans, on the other hand, naturalisation seems to stimulate voting to the extent that naturalised Cubans have voted at higher rates than Anglos (DeSipio, 1996). This may reflect their high income and education levels as well as their intense ideological motivations.

As has already been noted, the political practices that historically served to deny Hispanics access to the political arena have essentially been eliminated. Also, there is no systematic evidence that historical or contemporary discrimination affects contemporary turnout (Uhlaner, 1996; Leighley, 2001; Michelson, 2000a; Clark & Morrison, 1995). There have also been very few protests regarding election officials’ efforts to deny Hispanic citizens access to the polls as used to be common (Garcia and de la Garza, 1977). Nonetheless, although not designed to discriminate against Hispanics, there are still practices that disproportionately reduce Latino turnout. For example, the lack of Election Day registration disproportionately reduces Latino turnout relative to Anglo turnout (Alvarez & Ansolabehere, 2002). Similarly, changing the naturalization process either by making it more difficult or more expensive slows the pace at which immigrants become citizens. This, in turn, prevents their enfranchisement which has a negative impact on the ability of Latinos to turn out in large numbers (DeSipio and de la Garza, 2003).

The promise of a large and engaged Hispanic electorate remains unrealised because of institutional, political and demographic factors. The first include regulations such as election day registration and more stringent rules governing naturalisation. Political obstacles are centred in the unwillingness of political parties to invest in mobilising Latino voters. The demographic factors include the youthfulness of the population, its low level of income and education, and the large number of naturalised citizens who need to be socialised into American electoral realities. Until some combination of these are addressed, the potential of the Latino vote will remain unrealised.

New Leadership, New Issues

Implicit throughout the argument I have made is the suggestion that Hispanic leaders have a different role today than they had historically. In the 1970s and 1980s, the system was so discriminatory that while leaders such as Cesar Chavez, Willie Velasquez, Congressmen Edward Roybal and Robert Garcia, Antonia Pantoja, Jose Angel Gutierrez, Vilma Martinez and Raul Yzaguirre faced daunting and sometimes threatening obstacles, their objective was clear: they sought to end anti-Hispanic discrimination, achieve equal treatment before the law and institutionalise equal access to the political process. Although not all such problems have been eliminated, the relatively low levels of discrimination Latinos perceive both at the societal levels and in dealing with government officials, including the courts (de la Garza et al., 1992, p. 92; de la Garza & DeSipio, 2001), suggests the achievements that that generation of leaders helped to forge.

The challenges facing today’s leaders are more complex in that they deal with issues that are less explicitly ethnic and much more embedded in mainstream political problems. Compounding the difficulty of addressing these problems is the increased heterogeneity of the Hispanic population. Together these developments so fundamentally changed the Hispanic political community that it is clear that the perspective that developed out of the civil rights struggles cannot guide Latinos to a successful political future. Thus, Latinos will be well served if the shift to a new generation of leaders such as that begun with Raul Yzaguirre’s resignation from the National Council de la Raza and Antonia Hernandez’s retirement at MALDEF, both of whom were major contributors to Hispanic political struggles for decades, brings with it new perspectives on how to increase Hispanic political clout.

Illustrative of how Hispanic socio-political reality has changed are the issues about which Latinos are now most concerned. As shown in Table 10, economics and education top this list. Mirroring this move away from issues emphasising discrimination and the elimination of barriers to political access is the ‘national Hispanic agenda’ released in 2000 by the Hispanic Leadership Agenda (NHLA). NHLA, an organization made up of 31 Hispanic groups from across the country, defines Hispanic priorities as consisting of six broad categories: education, civil rights, government accountability, economic empowerment and health. Clearly, these are general issues with limited specific ethnic relevance other than the extent to which they address problems that are especially acute among Latinos.

Table 10. Latino Issues in the 2004 Presidential Election

(%)	Nation	Latino	Mexican	PuertoRican	Cuban	CA	TX	FL	NY
US campaign against terrorism	19	15	12.5	13	52.1	11.2	12.4	31.4	14.6
War in Iraq	20	12.7	14	13.6	7.5	9.7	16.2	11	10.4
Economy and jobs	28	33.3	32.8	32.7	20.2	35.5	32.4	31.4	27.1
Education	12	17.5	19.6	16.1	6.4	22.7	15.5	7.3	22.2
Health care	11	11.8	11.9	11.1	7.5	11.2	13.5	11.5	10.4
Crime	1	2.7	2.7	1.2	2.1	3.3	2.7	2.6	2.8
Other	8	3.8	3.5	6.8	2.1	3.5	4.2	2.6	6.3
Don’t know	1	3.2	3.1	5.6	2.1	2.9	3.1	2.1	6.3

Source: Washington Post/Univistion/ Tomás Rivera Policy Institute, 2004.

It is noteworthy that, having issued the agenda, NHLA invited the presidential candidates to discuss how they would implement its proposals. NHLA, in other words, sought to gain commitments supporting its policy priorities in exchange for Hispanic electoral support. NHLA’s negotiating clout, however, was diminished because of the state-based distribution of the electorate and the reality of low Latino turnout. This is why I argue that the fundamental political problem Latinos now face is how to increase voter turnout. As I have noted, population growth increases the number of Hispanic voters, but this growth rate is slowed when it is substantially driven by immigration, Thus, the rate of this expansion will remain incremental unless it is spurred by registration and GOTV campaigns. Although both are important, to date more resources have been put into voter registration on the assumption that if Latinos register, they vote. That may have been true of African Americans in the South during the civil rights movement and in Texas and other states with large Latino populations in the 1970s and 1980s when political authorities made it difficult for Hispanics to register (Garcia and de la Garza, 1977). Today, registration and voting have become uncoupled as registration has soared while turnout remains relatively unchanged (de la Garza et al., 2001). This is not primarily because of systemic obstacles since there are effectively no major barriers to registered Latinos or voting. Instead, it reflects the disproportionate investment in registration relative to voter mobilization.

Although the increases in registration have had little impact in presidential and other state-wide elections, these potential voters constitute a significant mobilisable resource. To capitalise on it requires de-emphasizing future registration campaigns in favour of GOTV efforts (de la Garza et al., 2002) like those practiced by political machines historically (de la Garza & DeSipio, 1994; Michelson, 2002a; Ramirez, 2002).

This strategy may be difficult to implement, however. Hispanic elected officials, like other elected officials, resist changes such as shifts in district boundaries or in the composition of their electorate that result from GOTV campaigns targeting new voters who might be open to appeals from other candidates. This is why Latino legislators in California opposed the re-districting lawsuit by the Mexican American Legal Defense and Education Fund (MALDEF) following the 2000 census (Cain & MacDonald, 2003). MALDEF’s proposal would have slightly reduced Latino majorities in many majority Latino districts and redistributed them into new districts where they would have the possibility of influencing electoral outcomes. Some of these were held by Anglo Democratic incumbents who feared MALDEF’s proposal would lead to their being challenged by Latino candidates. Thus by opposing MALDEF, Hispanic incumbents were not only defending their incumbency but were also strengthening their relationship with their party at the cost of reduced clout for Hispanic voters.

It will therefore be up to grass roots organizations such as San Antonio’s Communities Organized for Public Services and other groups affiliated with the Industrial Arts Foundation (Warren, 2001) and organizations such as the Southwest Voter Registration and Education Project (SWVRP) and the National Association of Latino Elected and Appointed Officials (NALEO) to promote GOTV campaigns as a primary strategy. Unlike elected officials, organizations like COPS have long recognised that to achieve their goals they must be able to hold officials accountable, and the best way to do this is through the ballot box. SWVRP and NALEO are likely to increasingly engage in GOTV efforts not only because of their concern for Hispanic well being but also because they are funded to do so. Thus they can pursue their institutional interests as they advance community interests.

Labour unions that are increasingly made up of Hispanic workers are also deeply involved in GOTV campaigns. Because they are much better funded and organized, they are likely to implement the most effective Hispanic GOTV campaigns (de la Garza et al., 2002). It must be noted, however, that unions, like Hispanic elected officials, emphasise their institutional objectives over generalised Latino interests. This is illustrated in the 2002 New York gubernatorial campaign when the state’s largest health workers’ union, SEIU 1190, led by then Democratic National Committee member Dennis Rivera, endorsed the incumbent Republican governor in exchange for union-specific benefits. Although the union received those benefits, it may be argued that the agreement did not serve the city’s Latino interests as a whole.

Another way to mobilise the Latino electorate is to develop tactics that particularly target Hispanics without explicitly employing ethnic appeals. As I have shown, these no longer inspire Latino turnout and could spur counter-mobilisation among non-Hispanics which could broaden rather than narrow the turnout gap that separates Latinos from Anglos. Latino leaders must therefore develop new techniques to increase Hispanic voting. One tactic entails Latino legislators coordinating from within their respective states a national campaign in the name of increasing voting among all citizens to enact election day registration. In California, this would increase Latino turnout by 11%, compared with the 3%-6% gain it would produce among Anglos (Alvarez & Ansolabehere, 2002). While this would increase turnout in each state, its actual impact would depend on the demographics of each state’s Hispanic population.

Such a campaign would have three positive outcomes for Latinos: (1) it would disproportionately increase Hispanic voting rates without pandering to ethnic voters; (2) it would demonstrate that Latino legislators can work as a unified group to increase Hispanic influence; and (3) it would force both political parties to support the effort since each claims to seek the Latino vote. If one party opposed it, it would expose as fraudulent any claims that party might make regarding its interest in reaching out to Latinos. What should be particularly appealing to Latino legislators about this tactic is that it is relatively cost free. The only reason to oppose it is out of the same self interest that led California Assembly members to oppose MALDEF’s re-districting plan. If a large number of Latino representatives nation-wide mobilised in support of this plan, however, it is hard to imagine any given state legislator or political party successfully opposing it without paying a substantial electoral price over time.

A second tactic is to have Hispanic Congressmen join forces to advocate for legislation that would enfranchise Puerto Ricans on the island without raising the statehood question, an issue which Puerto Ricans should decide for themselves. In effect, this would make their status much more comparable to that of reservation-based American Indians and citizens in Washington DC. This would not require raising the statehood question, since neither reservations nor Washington DC are states, nor would it eliminate the special tax arrangements that Puerto Rico enjoys vis-à-vis the federal government, since reservations also have special financial arrangements. By adding over three million citizens to the Hispanic eligible voter pool, this would enhance Latino influence in presidential elections by increasing the total number of Latino voters by between one and two million voters. Again, there is no obvious reason why Latino legislators should not support this proposal. If enacted, to the extent that Puerto Ricans share common economic and social interests as well as pan-ethnic priorities, Latinos as a group benefit. If one party opposes the plan, it will expose itself as opposed to increasing the Latino electorate and, in principle, jeopardise whatever support it has among Hispanics.

This tactic, it must be emphasised, requires Puerto Ricans to hold in abeyance the long-standing issue regarding statehood. Demanding voting rights will not necessarily lead to or away from statehood, but it will give Puerto Ricans and Hispanics in general greatly increased political clout. The right to vote, for example, has not detracted from the national identities of the Navajo or Hopi. Puerto Ricans must learn from such examples in order to improve the political status of Latinos and Puerto Ricans.

It is important to emphasise that these tactics do not involve most major Hispanic organizations. This is because even those Hispanic organizations that claim to be community-based, like LULAC, cannot claim significant national membership (Marquez, 1993), and they, like advocacy groups such as NALEO and SWVRP, are dependent for their funding on foundations, government agencies that contract for specific services with them and corporations and other contractors that want access to the Hispanic community. To receive this funding they must abstain from engaging in explicit political activities or risk their tax-exempt status, which would result in their funds being eliminated. Also, like many elected officials, these groups are more accountable to their funders than to the Hispanic community. Indeed, corporate funding is such a vital part of the funding of the major national organizations that more than one attendee at their annual meetings has described those events as little more than bazaars which corporate America and the alcohol and tobacco industry in particular use to gain access to the Latino market.

An example of the influence that such corporations have over Hispanic organizations is evident in the relationship the Coors Corporation has been able to cement with them. The historic Mexican-American Coors boycott came to an end with an agreement involving the creation of the Hispanic Association for Corporate Responsibility whose mission is facilitating the recruitment of Hispanics into senior management, and a contribution of up to US$350 million in advertisements in Latino media, investments, grants and scholarships to promote Hispanic businesses and education. The actual amount to be turned over depended on how much Coors beer was consumed as a result of the relationship (Acuña, 1988, p. 380). It is noteworthy that African Americans entered into a similar relationship with Coors except that their contract had no consumption-based incentives.

It is also important to note that foundations also constrain grant recipients. According to Angelo Falcon, founder and President of the Institute for Puerto Rican Policy (IPR), the support he received from the Ford Foundation began to decline because of his criticism of New York’s Mayor Dinkins who Ford officials strongly supported. Eventually, he lost all their support. That loss, combined with its long-standing policy of accepting no alcohol or tobacco money, brought an end to IPR’s status as an independent organization. With that was lost what Latinos throughout the Northeast and independent analysts across the country, including me, recognised as the region’s most creative policy and advocacy institute.

Equally egregious as the example of relations with alcohol and tobacco interests is the role some major organizations played during the NAFTA debate. The Mexican government enlisted NCLR and other Hispanic groups to mobilise support for NAFTA among Hispanics which would then be used to pressure Congress to approve the agreement. These organizations pursued this objective even though Latinos in general, as well as the Hispanic Congressional Caucus, were opposed to the agreement unless amendments were added that would directly benefit the Hispanic communities in the Southwest that NAFTA would most affect. With those modifications, a slight majority of the Latino congressional delegation voted for the agreement.

These examples illustrate the extent to which major Latino groups pursue their institutional interests even when these undermine Hispanic well-being or run counter to Latino preferences. Additionally, even though none are directly accountable to Latinos per se, the directors of these groups portray themselves as national Hispanic leaders. Such claims are exaggerated if not false, and thanks to the visibility and resources available to these groups relative to others, they also impede the development of a Hispanic agenda that genuinely reflects group priorities. For example, Latinos have long been ambivalent regarding immigration, and their Congressional representatives have reflected those views. Major Hispanic organizations, nonetheless, have been unrestrained in their support of liberal immigration reform. Whatever policy preferences one may favour regarding this issue, there can be no doubt that the policies supported by these organizations are more in tune with the preferences of elite Hispanics and foundations than with the preferences of most Hispanics including resident aliens.

Nonetheless, as the careers of Cesar Chavez, the Reverends Jesse Jackson and Al Sharpton illustrate, Hispanic heads of large community-based organizations are better situated than elected officials, almost all of whom hold district-based office, to be national spokespersons. None has attained this status, however. This may be because none has grass roots support behind them. It may also be that the heads of major Latino organizations have emphasised relationships with corporations and foundations rather than community ties. Together, these patterns help explain why these individuals and their organizations have such low visibility among the Hispanic population (de la Garza et al., 1992, p. 135). Given their low visibility, lack of independence and consequent intermittent defence of overall Hispanic well-being, these individuals cannot credibly claim to be authentic national leaders nor should their visions for the future of the Latino community be automatically trusted.

Elected representatives, despite their self-interested opposition to expanded Hispanic influence, are more likely to provide effective and visionary leadership. They are in much closer contact with the general population and are well informed about its needs and wants. More importantly, they are in principle accountable to Hispanic voters and they put their careers in jeopardy if they do not advocate for community interests.

Conclusion

The promise of Latino politics is greater today than ever before. In part this reflects the extent to which Latinos have become responsible for their own well-being, and in part it reflects changes in their political environment. Where once the issue was how to gain access to the political process, now it is how to best capitalise on the dramatic increase in the size of the population and the electorate, on the electoral access resulting from the VRA, and on the interest that both parties claim to have in the Latino vote. This transformation requires a shift of focus from how to overcome formidable obstacles used to deny Latinos political access to the development of strategies to take advantage of the new political environment and maximise electoral clout. This is not to say that there are no more impediments to political participation; rather, that Latinos now are part of the political mainstream and are in a position to combat those barriers from within the system.

This vantage point notwithstanding, Latinos still confront major problems which, if not addressed, will darken their future. Primary among these is immigration and immigrant incorporation. A successful future for Latino politics requires they play a lead role in redesigning the nation’s immigration policy and the social contract that governs how immigrants, be they legal or undocumented, are treated. Failure to address these issues will lead to an increasingly segmented Hispanic population, much of which will be relegated to society’s lowest strata.

As these issues are debated, Latino leaders must avoid advocating simplistic open border policies and acknowledge how immigration affects their communities and the immigrants themselves. Studies have well documented that at the local level immigrants consume more in public goods than they pay for through taxes (Massey, 2002). One reason for this is that much of the fiscal revenue produced by Latino communities goes into national coffers. Latino legislators should therefore focus on developing coalitions with other lawmakers to find mechanisms for effecting fiscal transfers from the federal level, which profits from immigration, to local entities, which bear the cost of immigrants.

Additionally, Latinos should recognise that many immigrants, especially those from Mexico, do not want to be here. They come only because of the economic and political situation in their homelands. Latino legislators and interest groups, therefore, should work with the American foreign policy establishment to influence development policy in Mexico and elsewhere. For example, Latinos should demand that the Mexican government do a better job of collecting taxes, investing in education and job creation and generally redistributing wealth. Otherwise, Mexican officials will continue to use emigration as a safety valve that allows them to avoid confronting the kinds of fiscal reforms that could reduce the need for so many of Mexico’s poor to migrate. It must be emphasised that this recommendation is not anti-immigrant; instead, it is intended to make it possible for Mexicans who prefer to remain in their homeland to be able to do so.

To influence immigration and the other issues that concern Hispanic electorates requires leaders to find ways to mobilise voters. Given the limited success such efforts have had historically, the leadership would do well to focus on institutional changes such as election-day registration and streamlined naturalisation procedures.

Perhaps the most significant consequence of the new political environment is the challenge it poses for Latino leaders. Historically, systemic obstacles were so substantial that the community’s leadership was shielded from charges of incompetence or bad judgment. Now political access is so substantial that the promise of a bright future for the Latino community increasingly rests on the skills and visions of a new crop of leaders. Thus, for the first time it seems fair to conclude that if Latino circumstances do not significantly improve it will to a substantial degree reflect failures of leadership.

2004 Election Postscript

The preliminary results of the 2004 election are consistent for the most part with the analysis presented here. Most significantly, Latinos once again did not influence the results of the election. Indeed, even in Florida where they were part of the winning coalition, they cannot claim to have determined the outcome as they claim they did in 2000. Exit polls suggest that Hispanics there gave President Bush a 300,000 vote cushion. While substantial, this is less than the President’s margin of victory. In other words, he could have won the state without these votes. It is also possible that Hispanics were instrumental to the Republican victory in New Mexico even though most of them voted Democratic. This is because President Bush appears to have increased his share of the Latino vote to the point that it swung the state to him.

They played significant but not definitive roles in states that Democrats won, like California and Texas, just as they did in states that Democrats lost, like Arizona and Nevada. Since most Latinos voted Democratic, however, and the Democrats lost, it is clear that Latinos once again did not shape the results of the election.

While they did not determine the presidential results, Hispanics did play a greater role than they have historically in Senate and congressional races. This is especially true within the Democratic Party where arguably Latinos registered the Party’s two major surprises, both in Colorado. There, Ken Salazar won a Senate seat formerly held by a Republican, and his younger brother won a congressional seat Republicans were expected to win. Senator Salazar’s victory is more noteworthy than the Senate seat won by Mel Martinez in Florida who was handpicked and strongly supported by President Bush and won in a state that his party carried.

The most significant issue raised by the 2004 election is the extent to which Latinos voted Republican. The available polls present extremely contradictory results. The exit polls produced by the news consortia suggest Bush received approximately 45% of their vote. Polls taken immediately before the election predict much lower support as does a poll by the pro-Democratic but generally reliable Willie Velasquez institute. Leading specialists on Latino voting doubt both results. Whatever the truth may be, there is no doubt that Republican outreach had an impact. What is unclear is whether this is temporary or reflects a partisan switch. As that question is approached, it is imperative to recall that Latinos voted in the 40%+ range for Ronald Reagan in the 1980s and then returned to their history of strong Democratic support. Time will tell whether that pattern will repeat or if 2004 is the beginning of a new era.

Rodolfo O. de la Garza

Department of Political Science, Columbia University

Bibliographical References

Acuña, Rodolfo (1988), Occupied America: A History of Chicanos, 3rd ed., Harper & Row, New York.

Alvarez, R. Michael, and Stephen Ansolabehere (2002), ‘California Votes: The Promise of Election Day Registration’, Demos, New York.

Alvarez, R. Michael, and Lisa Garcia Bedolla (2001), ‘The Foundation of Latino Voter Partisanship: Evidence from the 2000 Election’, paper presented at the Annual Meeting of the Western Political Science Association, Las Vegas, Nevada, 14-17/III/2001.

Arvisu, John R., and F. Chris Garcia (1996), ‘Latino Voting Participation: Explaining and Differentiating Latino Voting Turnout’, Hispanic Journal of Behavioral Sciences, 18, nr. 2, p. 104-28.

Barreto, Matt, and Nathan D. Woods (2003), ‘Latino Voting Behavior in an Anti-Latino Political Context’, in Gary Segura and Shaun Bowler (Eds.), Diversity in Democracy: Minority Representation in the United States, University of Virginia Press, Charlotteville (in press).

Bass, Loretta E., and Lynne M. Casper (1999), ‘Are There Differences in Registration and Voting Behavior between Naturalized and Native-Born Americans?’, US Bureau of the Census Population Division, Working Paper nr. 28., p. 19.

Cain, Bruce E. and Karin Mac Donald (2003), ‘Victims of Their Own Success: Latino Divisions on Redistricting. Paper presented at Renewing the Voting Rights Act’, paper presented at Renewing the Voting Rights Act Conference, sponsored by the Department of Political Science, Columbia University and York University Law School.

Cain, Bruce, D. Roderick Kiewiet and Carole Jean Uhlaner (1991), ‘The Acquisition of Partisanship by Latinos and Asian Americans’, American Journal of Political Science, 35, nr. 2, p. 390-422.

Clark, William A.V., and Peter A. Morrison (1995), ‘Demographic Foundations of Political Empowerment in Multiminority Cities’, Demography, 32, nr. 2, p. 183-201.

Dawson, Michael C. (1994), Behind the Mule: Race and Class in African-American Politics, Princeton University Press, Princeton, New Jersey.

DeSipio, Louis (1996b), Counting on the Latino Vote: Latinos as a New Electorate, University of Virginia Press, Charlottesville.

DeSipio, Louis, and Rodolfo O. de la Garza (2003), ‘Reshaping the Tub: Incorporating Latino Issues into the Voting Rights Act’, paper presented at the Renewing the VRA Conference, hosted by Columbia University and New York University Law School.

DeSipio, Louis, and Rodolfo O. de la Garza (2004), ‘Between Symbolism and Influence: Latinos and the 2000 Elections’, in Rodolfo O. de la Garza and Louis DeSipio (Eds.), Muted Voices: Latinos and the 2000 Elections, Rowman & Littlefield Publishers, p. 13-60.

DeSipio, Louis, Harry P. Pachon, Rodolfo O. de la Garza and Jongho Lee (2003), ‘Immigrant Politics at Home and Abroad: How Latino Immigrants Engage the Politics of Their Home Communities and the United States’, Tomás Rivera Policy Institute, Claremont, California, p. 55.

Diaz, William A. (1996),‘Latino Participation in America: Associational and Political Roles’, Hispanic Journal of Behavioral Sciences, 18, nr. 2, p. 154-74.

Garcia, F. Chris, and Rodolfo O. de la Garza (1977), The Chicano Political Experience: Three Perspectives, Duxbury Press, North Scituate, Mass.

Garza, Rodolfo O. de la (1992), ‘From Rhetoric to Reality: Latinos and the 1988 Election in Review’, in Rodolfo O. de la Garza and Louis DeSipio (Eds.), From Rhetoric to Reality: Latino Politics in the 1988 Elections, Westview Press, Boulder, Colorado, p. 171-80.

Garza, Rodolfo O. de la (1996), ‘El Cuento de los Números and Other Latino Political Myths’, in Anibal Yanez-Chavez (Ed.), Su Voto es Su Voz: Latino Politics in California, Center for US-Mexican Studies, University of California, San Diego, California, p. 11-32.

Garza, Rodolfo O. de la, Jerome Hernandez, Angelo Falcon, F. Chris Garcia and John A. Garcia (1997), ‘Mexican, Puerto Rican, and Cuban Foreign Policy Perspectives: A Test of Competing Explanations’, in F. Chris Garcia (Ed.), Pursuing Power: Latinos and the Political System, University of Notre Dame Press, Notre Dame, Indiana, p. 401-426.

Garza, Rodolfo O. de la, Charles W. Haynes and Jaesung Ryu (2001-02), ‘An Analysis of Latino Voter Turnout Patterns in the 1992-1998 General Elections in Harris County, Texas’, Harvard Journal of Hispanic Policy, 14, p. 77-95.

Garza, Rodolfo O. de la, and Jeronimo Cortina (2004), ‘Dance with the One What Brought You: Latino Partisan Fidelity’, paper presented at the From Rhetoric to Reality: Latino Politics in 2004 Conference, Los Angeles, California, 26/II/ 2004.

Garza, Rodolfo O. de la, and Louis DeSipio (Eds.) (1999), Awash in the Mainstream, Westview Press, Boulder, Colorado.

Garza, Rodolfo O. de la, and Louis DeSipio (Eds.) (2004), ‘Muted Voices: Latinos and the 2000 Elections’, in Paula D. McClain and Joseph Stewart Jr. (Eds.), The Spectrum Series: Race Ethnicity in National and Global Politics, Rowman & Littlefield Publishers.

Garza, Rodolfo O. de la, and Louis DeSipio (2001), ‘A Satisfied Clientele Seeking More Diverse Services: Latinos and the Courts’, Behavioral Sciences & the Law, 19, nr. 2, p. 237-48.

Garza, Rodolfo O. de la, Louis DeSipio, F. Chris Garcia, John Garcia, and Angelo Falcon (1992), Latino Voices: Mexican, Puerto Rican, and Cuban Perspectives on American Politics, Westview Press, Boulder, Colorado.

Garza, Rodolfo O. de la, Louis DeSipio, F. Chris Garcia, John Garcia, and Angelo Falcon (1994), ‘Do ‘Latinos Exist? – Latino Voices: Mexican, Puerto Rican, and Cuban Perspectives on American Politics’, Contemporary Sociology, 23, nr. 3, p. 354.

Garza, Rodolfo O. de la, Carolyn Dunlap, Jongho Lee, and Jaesung Ryu (2002), ‘Latino Voter Mobilization in 2000: Campaign Characteristics and Effectiveness’, Tomás Rivera Policy Institute, Claremont, California.

Gimpel, James G., and Karen Kaufmann (2001), ‘Impossible Dream or Distant Reality? Republican Efforts to Attract Latino Voters’, Center for Immigration Studies, Washington DC.

Graves, Scott, and Jongho Lee (2000), ‘Ethnic Underpinnings of Voting Preference: Latinos and the 1996 US Senate Election in Texas’, Social Science Quarterly, 81, nr. 1, p. 226-36.

Green, Donald P., Bradley Palmquist and Eric Schickler (2002), Partisan Hearts and Minds: Political Parties and the Social Identities of Voters, Yale Isps Series, Yale University Press, New Haven.

Greenfeld, Liah (1992), Nationalism: Five Roads to Modernity, Harvard University Press, Cambridge, Mass.

Guerra, Fernando (1992), ‘Conditions Not Met: California Elections and the Latino Community’, in Rodolfo O. de la Garza and Louis DeSipio (Eds.), From Rhetoric to Reality: Latino Politics in the 1988 Elections, Westview Press, Boulder, Colorado, p. 99-107.

Guerra, Fernando (1998), ‘Latino Politics in California: The Necessary Conditions for Success’, in Michael B. Preston, Bruce Cain and Sandra Bass (Eds.), Racial and Ethnic Politics in California, IGS Press Institute of Governmental Studies, University of California, Berkeley, California, p. 439-53.

Hritzuk, Natasha, and David K. Park (2000), ‘The Question of Latino Participation: From an SES to a Social Structural Explanation’, Social Science Quarterly, 81, nr. 1 p. 151-166.

Jones-Correa, Michael, and David Leal (1996), ‘Becoming “Hispanic”: Secondary Panethnic Identification among Latin American-Origin Populations in the United States’, Hispanic Journal of Behavioral Sciences, 18, nr. 2, p. 214-54.

Leighley, Jan E. (2001), Strength in Numbers?: The Political Mobilization of Racial and Ethnic Minorities, Princeton University Press, Princeton, New Jersey.

Marquez, Benjamin (1993), Lulac: The Evolution of a Mexican American Political Organization, University of Texas Press, Austin.

Michelson, Melissa R. (2002b), ‘Competing Vote Cues and the Authenticity of Representation: Latino Support for Anglo Democrats and Latino Republicans’, paper presented at the Annual Conference of the Midwest Political Science Association, Chicago, Illinois, 25-28/IV/2002.

Michelson, Melissa R. (2002a), ‘Exploring Latino Political Efficacy and Electoral Participation in California and Chicago’, paper presented at the Annual Meeting of the American Political Science Association, Washington DC, 31/VIII-3/IX/2000.

Michelson, Melissa R. (2003b), ‘Getting out the Latino Vote: How Door-to-Door Canvassing Influences Voter Turnout in Rural California’, Political Behavior, 25, nr. 3, p. 247-263.

Mollenkopf, John, David Olson and Timothy Ross (2001), ‘Immigrant Political Participation in New York and Los Angeles’, in Michael Jones-Correa (Ed.), Governing American Cities: Inter-Ethnic Coalitions, Competition, and Conflict, Russell Sage Foundation, New York, p. 17-70.

Moreno, Dario, and Christopher Warren (1999), ‘Pragmatism and Strategic Realignment in the 1996 Election’, in Rodolfo O. de la Garza and Louis DeSipio (Eds.), Awash in the Mainstream, Westview Press, Boulder, Colorado, p. 211-237.

Nicholson, Stephen P., Adrian D. Pantoja and Gary M. Segura (2002), ‘Ich Bin Ein Latino! Sophistication, Symbolism, Heuristics, and Latino Preferences in the 2000 Presidential Election’, paper presented at the Annual Meeting of the American Political Science Association, Boston, Massachusetts, 29/VIII-1/IX/2002.

Pew Hispanic Center (2002), ‘2002 National Survey of Latinos’, The Pew Hispanic Center and The Kaiser Family Foundation, Washington DC.

Pachon, Harry, Rodolfo O. de la Garza, and Adrian Pantoja (2000), ‘Foreign Policy Perspectives of Hispanic Elites’, in Rodolfo O. de la Garza and Harry Pachon (Eds.), Latinos and U.S. Foreign Policy: Representing the ‘Homeland’?, Rowman & Littlefield Publishers, p. 21-42.

Pachon, Harry P., Matt A. Barreto and Frances Marquez (2004), ‘Latino Politics Comes of Age: Lessons from the Golden State’, in Rodolfo O. de la Garza and Louis DeSipio (Eds.), Muted Voices: Latinos and the 2000 Elections, Rowman & Littlefield Publishers, p. 84-100.

Poston, Dudley L., Steven A. Camarota and Amanda K. Baumle (2003), ‘Remaking the Political Landscape: The Impact of Illegal and Legal Immigration on Congressional Apportionment’, Center for Immigration Studies, Washington DC.

Ramirez, Ricardo (2002), ‘Getting out the Vote: The Impact of Non-Partisan Voter Mobilization Efforts in Low Turnout Latino Precincts’, paper presented at the Annual Meeting of the American Political Science Association, Boston, Massachusetts, 29/VIII-1/IX/2002.

Seelye, Katherine Q. (2003), ‘Shifts in States May Give Bush Electoral Edge’, The New York Times, 2/XII/2003, A24.

Smith, Anthony D. (1991), National Identity, Ethnonationalism in Comparative Perspective, University of Nevada Press, Reno.

Tomás Rivera Policy Institute (2000), ‘5 State Validated Latino Registered Voters Election Survey’, Tomás Rivera Policy Institute, Claremont, Califormia.

Uhlaner, Carole Jean (1996), ‘Latinos and Ethnic Politics in California: Participation and Preference’, in Anibal Yanez-Chavez (Ed.), Latino Politics in California: Su Voto es Su Voz, Center for US-Mexican Studies, University of California, San Diego, p. 33-72.

Uhlaner, Carole Jean, and F. Chris Garcia (2002), ‘Latino Public Opinion’, in Barbara Norrander and Clyde Wilcox (Eds.), Understanding Public Opinion, Congressional Quarterly, Inc., Washington DC, p. 77-101.

Verba, Sidney, Kay Lehman Schlozman and Henry E. Brady (1995), Voice and Equality: Civic Voluntarism in American Politics, Harvard University Press, Cambridge, Mass.

Warren, Mark R. (2001), ‘Dry Bones Rattling: Community Building to Revitalize American Democracy’, Princeton Studies in American Politics, Princeton University Press, Princeton, New Jersey.

Washington Post/Univisión/Tomás Rivera Policy Institute (2004), ‘Election Survey of Latino Registered Voters’.