Published online by Cambridge University Press: 26 February 2004
The 2003 California recall election was historic in more ways than one. Not only was this the first time in California's history where an attempt to recall a statewide official successfully made it onto the ballot, but minority voters had a chance to cast the deciding votes in the election. Despite growing frustration and contempt for the governor across the state, Gray Davis, a Democrat, was counting on strong support from minority voters to retain his office. Eleven months earlier in the November 2002 gubernatorial election, majority support from Latinos, African Americans, and Asian Americans propelled Davis to a 5-point victory over Republican challenger Bill Simon, who carried a 46–43 edge among White non-Hispanic voters. In addition, Lieutenant Governor Cruz Bustamante had entered the replacement candidate race and was making a bid to become the first Latino governor of California. Thus, as the election drew near, pundits and scholars alike were left speculating what role minority voters would have in the recall of Governor Davis as well as the election of a replacement candidate. The general feeling leading up to the election was summarized by Harry Pachon, president of the Tomás Rivera Policy Institute: “the minority vote will be more critical in this election than in any election in the past 25 or 30 years” (Sterngold 2003).
Before examining the minority vote in the 2003 California recall election, it is important to briefly review Latino, Black, and Asian voting preferences over the past decade. Scholars of California politics know well that the 1990s represented a contentious and significant era for minority voters (DiCamillo and Field 2000; Fraga and Ramírez 2003; Pachon 1998; Segura, Falcon, and Pachon 1997). Since 1992, Latino, Black, and Asian voters have become accustomed to ethnic appeals, both positive and by negative (Guerra and Fraga 1996; Pantoja and Segura 2003). In 1994, 1996, and 1998, three successive statewide ballot initiatives seemingly targeting minorities in California, were backed by Republican candidates for office, including Governor Pete Wilson (Ramírez 2002). In 1994, Proposition 187 rolled back state services for undocumented immigrants; in 1996, Proposition 209 sought to end affirmative action in public institutions; and in 1998, state voters passed Proposition 227 ending bilingual education programs in public schools. The cohesive non-Hispanic bloc vote in favor of these Republican-sponsored ballot initiatives (63%, 63%, and 67% respectively) overwhelmed minority voter opposition but has been credited in galvanizing minority voter support for the Democratic Party (Barreto and Woods 2000). After the loss of the governorship and all but one statewide office by Republicans in 1998, some prematurely believed that California Democrats had consolidated minority support and turned the state into a one-party system. However, the end of the Wilson era in 1998, significant efforts to court minority voters in 2000 by President Bush, and a less than stellar performance by Gray Davis in 2002 called into question the certainty and continuation of minority voting patterns.
In Table 1 we examine the percent of the vote won statewide by Democratic candidates broken down by race/ethnicity 1992–2003.1
Based on Los Angeles Times exit poll surveys conducted on Election Day 1994–2003; Voter News Service 1992 exit poll.


While minority voters contributed to the success of nine of the 10 Democratic candidates who went on to win in California, it seems that White voters may hold the real key to victory. The only unsuccessful Democrat, Brown in 1994, saw her support dip to only 35% among White non-Hispanic voters, despite strong support from Latino and Black voters. While less likely to vote Democrat than minorities, Whites voted at least 40% Democrat in the nine elections that Democrats won. The reason White voters are so influential lies in their numbers. Although California is a majority-minority state measured by total population (53% minority in 2000 Census), Whites are, and will likely continue to make up for some time, the majority of the electorate (Citrin and Highton 2002).
However, over the past 12 years, minority voters have become a larger part of the California electorate, growing by almost 70% from 1990 to 2000, according to estimates from the U.S. Census Current Population Survey (CPS). This is in large part due to the increase of successful mobilization drives targeting Latino voters (Pachon 1998; Pantoja and Woods 2000; Michelson 2003; Ramírez 2002). Table 2 reports the total number of California voters, and the percent that each racial/ethnic group represented in statewide elections from 1990 to the 2003 recall election. In 1990, Whites accounted for 82.3% of all voters, compared to just 17.7% for Latino, Black, and Asian voters combined (see Table 2). However, each year between 1990 and 2000 the minority share of the electorate grew, standing at 29.8% by the 2000 presidential election (See Figure 2). This growth has primarily been due to increases in Latino and Asian-American voter registration and turnout, while Black voting has remained at roughly the same levels. Minority participation reached a high in the 2000 presidential election but fell significantly in the 2002 gubernatorial election. Although turnout was low statewide, minorities, feeling alienated from their traditional Democratic base, were even less likely to vote in 2002 (Barabak 2002). Thus in 2002, the minority share of the California electorate decreased for the first time in 10 years, falling an estimated four percentage points from just two years earlier. With increased mobilization and the frenzy surrounding the 2003 recall, minority turnout was up in 2003 as compared to 2002 (Tomás Rivera Policy Institute 2002; 2003). Despite increased turnout in the recall election, levels of support for Democratic candidates were down among Latinos, Blacks, and Asians.


Table 3 reviews vote preference in the 2003 recall election and candidate replacement ballot as compared to the mean Democratic vote rates from 1992–2002 in California. An interesting pattern emerges among minority voters in the recall election. Both Black and Asian voters voted “No” on the recall at virtually the same rate as they had voted for Davis in 2002. Latino support for Davis, on the other hand, was down 10% from their levels of support in 2002 and 14% lower than their average support for Democrats over the 10-year period. Conversely, even though Cruz Bustamante was the only major Democrat replacement candidate on the ballot, Black and Asian support for Bustamante fell off by 14 and 19 points to 65% and 34%, respectively. Similarly, White support for Davis and Bustamante fell off significantly in the recall election.

Two points are notable about the levels of Democratic support in the recall election. First, only Latinos maintained equal support for Democrats on both parts of the ballot, voting to retain Davis and elect Bustamante at virtually the same rate. Every other ethnic group witnessed a sharp drop in support for Bustamante as the replacement candidate for governor, as compared to their “No” vote on the recall. Even though Blacks were more likely to vote for Bustamante than Latinos, there was a large inconsistency between their vote for Davis and Bustamante, the two top Democrats in the election. Second, the difference between the average Democratic vote, and the vote for Bustamante is considerably outside the expected vote range. None of the support levels for Bustamante are in the predicted 95% confidence range that 10 years of previous vote history projects. What's more, for all ethnic groups, the vote for Bustamante was more than two standard deviations away from their mean levels of support for Democrats in state elections, including four standard deviations away for White voters. While most of the negative attention was on the state's top Democrat, Davis—so negative that he was facing a recall petition—it was Bustamante who received significantly lower than expected levels of support.
While neither Davis nor Bustamante reached the high levels of support from minority voters that they might have hoped for, the results might have been different had voter turnout and Democratic support among Latinos, Blacks, and Asians been more consistent with their exhibited patterns since 1992. In order to assess the probability that minority voters may have changed the outcome of the recall election, we simulated three projections of voter turnout and issue preference by race and ethnicity for California voters, reported below in Table 4. The first two lines represent the actual vote results and the actual percent of the electorate (i.e. voter turnout) for each racial/ethnic group. The percentage in the final column, Recall “No,” represents the predicted no vote on the recall and is the combined result that changes in vote preference and voter turnout would have had on the recall initiative.

Projection 1 holds recall support levels constant and adjusts minority turnout levels to reflect their true percents of the electorate. Had minority turnout been higher, but support levels stayed the same, the recall would have passed 53–47. Projection 2 holds turnout levels constant, and adjusts recall opposition levels to reflect the level of support that Davis received 11 months earlier in the 2002 election. Had Davis sustained the same levels of support in 2003 as he had in 2002, he still would have lost the recall, although by a very slim margin of 51–49. Finally, Projection 3 combines the adjustments made in the previous two projections by increasing minority turnout levels and using 2002 Davis support levels. In this scenario, the recall initiative would have failed by a 51–49 margin. While this scenario did not play out on October 7, 2003, it is not unrealistic to think that it might have. According to the California secretary of state, Latinos represent 18% of all registered voters, so had their turnout been equal to that of all other voters, they would have represented 18% of the voting public. In addition, had Davis and the Democratic Party confronted the recall sooner and begun their campaigning earlier, they may have been able to sustain their levels of support in 2002. Further, the perceived high turnout and strong support for Democrats Davis and Bustamante by Latinos may have been lost by the clouded and often confused messages of the disjointed and unorganized state Democratic Party. The failure of the Davis and Bustamante campaigns to effectively embrace one another may have dissuaded some Latino voters from going to the polls or supporting one or both of the Democratic candidates. While two out of the three projections yield the same result as that on Election Day, the projection that seemed to have the most support in the months before the election was Projection 3, in which case, the recall would have failed (Hernandez 2003; Gold 2003). Pollster Mark Baldassare noted, “Minority voters in California have proven the winning edge for Democratic candidates in the past … and for the recall their vote will be a huge issue” (Sterngold 2003).
While minority voters were not the crucial swing vote that many had anticipated, the circumstances of the recall election may be too unique to draw any broad generalizations about the future of California politics. With the 2004 presidential election beginning to pick up steam, Latino, Black, and Asian voters in California will no doubt be courted by the Democratic and Republican parties. With a Republican governor, the stakes may change in the otherwise safe Democratic state of California, and minority voters may find themselves in the middle of a competitive presidential contest. Whatever the case, the 2003 recall was a fascinating, if not extremely unusual election—even by California standards, and implications that the Republican Party made strides among minority voters may prove to be premature. We will have to wait for the next chapter of minority voter turnout and candidate preference in California to be written during the 2004 election.
Author names are presented in alphabetical order. The authors would like to thank Shaun Bowler and Bruce Cain for their review and helpful suggestions on an earlier draft of this article. We also benefited from discussions with Louis DeSipio, Janelle Wong, Gary Segura, Stephen Nuño, and Nathan Woods about minority voters in California, and are grateful for their suggestions.