As reported first by United Press International, Republicans won the two-party vote by a 53 percent to 47 percent margin. Despite the collapse of the Voter News Service exit polls, evidence has mounted that the Republicans triumphed not by broadening their tent to include more minority voters; rather, they motivated more whites to turn out and vote GOP.
With the 2000 Census showing that the non-white population was growing rapidly, GOP leaders such as Karl Rove, President George W. Bush's top political adviser, devoted much lip service and a fair amount of money over the last two years to wooing minorities, especially Hispanics.
Matthew Dowd, head of polling in Bush's 2000 campaign, told The Washington Post in 2001 in a widely cited remark, "Republicans have to increase their percentage among blacks and certainly among Hispanics. As a realistic goal, we have to get somewhere between 13 and 15 percent of the black vote and 38 to 40 percent of the Hispanic vote." Bush had won only 9 percent of the black vote and 35 percent of the Hispanic vote, according to the 2000 VNS exit poll of 13,130 voters.
Before Sept. 11, 2001, Bush frequently visited with Mexican President Vincente Fox and the administration leaked vague plans for an amnesty for many Mexican illegal immigrants. After the terrorist attacks, however, both U.S.-Mexico relations and loosening immigration laws moved to the Bush administration's back burner.
When the crunch came in the last 10 days of the 2002 campaign, the Republicans went hunting where the ducks are: among whites. Non-Hispanic whites had cast 81 percent of the votes in the 2000 election according to both VNS and the Census Bureau's post-election phone survey of 50,000 households.
It's possible that in this lower turnout midterm election, white voters were even more dominant. One of the pre-election polls that correctly registered the GOP's late surge to victory, the Ipsos-Reid/Cook poll a survey of 1,518 registered voters from Oct. 28 to Nov. 3 ( with an overall margin of error plus or minus 2.6 points), found that 85 percent of likely voters were white. Only 7 percent were black, and just 4 percent Hispanic. (In 2000, Hispanics cast 5.4 percent of the total vote according to a huge Census Bureau study, and 6.5 according to VNS.)
Those exit polls that are available from 2002 show a similarly strong white turnout as the Ipsos-Reid/Cook survey predicted. In 10 battleground states, Fox News conducted quasi-exit polls, phoning about 900 voters in each state on Election Eve or Election Day. (Please keep in mind that polls are not perfect. The Fox polls picked the wrong winner in four of the 18 major races in those 10 states, most notably in Georgia.)
In California, the L.A. Times conducted a traditional Election Day face-to-face exit poll with a sizable 3,400 respondents. On average in these 11 states covered by the Fox and L.A. Times exit polls, the average white share increased two points over the 2000 VNS findings, with increases coming in states such as Florida (+9 points), Colorado (+5), Missouri (+4) and California (+3). The most important white decline was in Texas (-2).(These figures compare the VNS exit poll turnout figures from the 2000 Presidential election to the Fox and L.A. Times exit polls for the 2002 mid-term elections. Though not perfect, these estimates are the best that can be done at present.)
In the campaign's endgame, the Republicans focused overwhelmingly on getting people in the solidly Republican areas to show up and vote GOP. These areas are also primarily white. (By their nature, techniques such as door-to-door canvassing work best in homogenous neighborhoods where the chances of reminding the other party's sympathizers to vote are low.)
The president blitzed his way through GOP rallies in battleground states across the nation. Moreover, during the last weekend, the GOP launched "ground war," its carefully crafted "72 Hour Plan" to get out the vote in Republican rural and suburban neighborhoods in about 30 states.
Led by celebrated grass roots organizer Ralph Reed, the former executive director of Pat Robertson's Christian Coalition, Georgia's Republicans unseated a Governor, a U.S. Senator, and a local legend who was Speaker of the Georgia House for the last 28 years. Reed commented to The New York Times, "The story of 2002 is not that Democrats stayed home, it was that Republicans came to the polls in historic numbers..."
The N.Y. Times reported that the winning margin in Georgia came from "white, rural, Republican-leaning voters, according to Republican and Democratic officials." Bo Harmon, the campaign manager of victorious Republican Senatorial candidate Saxby Chambliss, described his last-weekend recruits as, "They're generally conservative, but just not politically motivated. They're people who typically don't turn out."
The Gallup Poll recorded a swing in the Congressional vote from a three point Democratic advantage two weeks before the election to a six point Republican margin on Oct. 31 to Nov. 3. (This telephone survey interviewed 1221 adults, and had a plus or minus 3 point confidence interval at the 95 percent level. Of the respondents, 715 were likely voters. It's overall prediction turned out to be almost exactly correct.)
The Gallup analysts said the Republican victory "largely resulted from a greater turnout of Republicans than Democrats." From their early November poll, Gallup projected the turnout rate for Republicans voters would be 43 percent, compared with 36 percent for Democrats. (Curtis Gans of the Committee for the Study of the American Electorate estimated after the election that overall turnout was 39.3 percent.)
In California, the total number of votes cast dropped precipitously compared to 1998, probably due to ill feelings toward both parties' gubernatorial candidates. Only a little over half as many minorities showed up to vote in 2002 compared to 1998, according to the L.A. Times exit poll. Yet, the number of votes cast by whites was stable. This helped the seemingly hapless Republican candidate Bill Simon narrow incumbent Gray Davis' victory margin from 20 points in 1998 to a surprisingly narrow five points last week.
Gallup pointed out, "By far the largest divide among American voters continues to be racial." Nationally, minorities favored the Democrats by an overwhelming margin. This Gallup survey pegged Democratic candidates for Congress as winning 82-14 among nonwhites. Similarly, the Pew Poll, which incorrectly predicted a narrow Democratic victory, saw 85 percent of minorities favoring the Democrats.
While the sample sizes are much too small to be definitive, these two surveys imply that Republicans may have done even worse among minority voters than they did in the 2000 Presidential election, when VNS found the GOP winning around 21% of the nonwhite vote.
Yet, ultimately, that mattered less than many pundits had expected because whites turned out relatively heavily, and they appear to have voted more strongly for Republicans than in recent elections. Gallup discovered that right before the election whites favored the Republicans by a 20-point margin: 58 percent to 38 percent.
This was up sharply from the gap found in 2000 Presidential exit polls among whites: VNS saw Bush winning only a 12 point victory over Al Gore, while the L.A. Times national exit poll recorded the margin as 11 points.
(By the way, the L.A. Times only conducted its exit poll in California this year, so there are not yet any national exit poll results. VNS is considering releasing some of the data it collected, but nothing is for sure yet.)
Similarly, in California's gubernatorial race, the L.A. Times learned that the GOP's performance among whites improved nine points, from losing 45-51 in 1998 to winning 46-43 in 2002.
The poor performance of the GOP in California in recent years is typically attributed to the rapidly growing immigrant population, but that factor can be greatly exaggerated. Hispanics, for example, only cast ten percent of the votes in California this year, down from 13 percent in 1998 and 2000, according to the L.A. Times exit polls for each election.
The biggest problem faced by California Republican problem is not actually minorities, but that its whites are much more Democratic-leaning than in the rest of the country. In contrast, Texas Republican governor Rick Perry won reelection 58-40 by carrying 70 or 72 percent of the white vote, according to Fox's exit poll and his internal poll, respectively.
How Hispanics voted last week is crucial to a contentious dispute among Republican strategists. One camp, which the Bush Administration actively favored before Sept. 11, argues that Republicans must win a higher share of Latino votes than the 35 percent Bush captured in 2000 and that the best way to do that is to please Hispanic voters by easing immigration restrictions.
The other side scoffs that Hispanics will always vote more for Democrats than for Republicans; therefore, increasing immigration would be political suicide in the long run. In the short run, they argue, Hispanic voting power is currently small enough to allow Republicans to salvage the party's distant future by imposing an immigration cutback now.
Mexican-American turnout, often assumed to be an ever-rising tide, did not appear to be particularly strong this year.
Besides declining in California from 13 percent to 10 percent (according to the L.A. Times exit poll), Floridian Hispanics fell from 11 percent of the electorate in 2000 to seven percent. In Colorado, they fell from 14 to 10 percent. In New Jersey, from five to four percent. (These are from the Fox Election Day surveys in 10 states and are compared to the 2000 VNS Presidential election exit poll).
The one bright spot for Hispanic turnout in the states for which exit polls are available is Texas, where Hispanic gubernatorial candidate Tony Sanchez, who spent about $60 million on his own money on his campaign, appeared to help boost their share since 2000 from ten to 17 percent.
But much of that came from his home base in the Rio Grande Valley, according to the Fort Worth Star-Telegram last Sunday: "For all Texas counties that are 75 percent or more Hispanic, including populous El Paso and Cameron counties, turnout nudged upwards by 4 points."
The day after the election, the Houston Chronicle called the Hispanic turnout in urban Harris County "anemic," observing, "Voters in Harris' conservative Republican suburbs went to the polls at triple the rate of voters in most of the older barrio precincts on Houston's east side."
Did Republican outreach efforts pay off in a higher percentage of those Hispanics who did vote picking GOP candidates this election?
The evidence is mixed.
In 2000, VNS reported that Bush won 35 percent of the national Hispanic vote. In several of the states for which there is data for this year, the GOP Hispanic vote share couldn't climb out of the 20s. The L.A. Times measured Simon as winning 24 percent of the California Hispanic vote, barely up from 1998 Republican candidate Dan Lungren's 23 percent. According to two major 2000 exit polls (VNS and L.A. Times), Bush had garnered either 23 or 29 percent of Golden State Latinos.
In Colorado, according to Fox, Republican Senator Wayne Allard carried 29 percent of Hispanics versus 25 percent for Bush (VNS) in that state in 2000.
In New Jersey, the GOP share fell from 35 (VNS) to 26 (Fox) over the last two years.
In Massachusetts, a lesser-known University of Massachusetts at Boston exit poll of 1,200 Hispanic voters claimed that 87 percent favored the Democratic candidate for governor, and 92 percent opposed Question 2 calling for the elimination of bilingual education. Yet, among all voters, Republican Milt Romney won a solid victory and the Ron Unz-backed initiative for English immersion classes for immigrant children triumphed in a 68-32 landslide.
Shortly after the election, the Republican National Committee boasted that their candidates had won by big margins in three counties -- El Paso in Colorado, Gwinnett in Georgia, and Wake in North Carolina -- that they claimed had relatively many Hispanics for their state, thus trying to imply that lots of Latinos had voted for the GOP in those districts.
These examples, however, speak more of desperation on the part of the RNC at spinning the Hispanic results than of hard data. According to the 2000 Census, these three counties range from only five to 11 percent Hispanic in population. And, the fractions of the voters in each who are Hispanic are no doubt much smaller. So, little can be inferred about Hispanic voting propensities from these examples.
The RNC, however, also offered three better examples:
In Texas, the Republican senatorial and gubernatorial candidates won 33-35 percent of the Hispanic vote according to the Fox exit poll. This is down from the 43 percent that favorite son Bush won there in 2000. Further, it's not much above the average Republican share in Presidential elections over the last 40 years. Still, it was a relatively good performance against a heavy-spending Hispanic gubernatorial candidate. On the other hand, continuing to lose roughly two to one is not going to solve the long-term problem that Hispanic immigration and high birth rates poses for the GOP.
In Florida, home to many conservative Cuban voters, Jeb Bush, who has a degree in Latin American studies and a Mexican-born wife, picked up 56 percent of the Hispanic vote, according to Fox. This beat his brother's 2000 performance (49 percent according to VNS) by seven points.
Among all voters, Jeb did 13 points better than George W., but that sibling improvement was due almost completely to the white turnout increasing from a 73 percent share to an 82 percent share, combined with Jeb's margin among Florida whites (60-38) improving five points over George W.'s (57-40).
The one state in which a Republican incumbent definitely followed a "minority outreach" strategy with success was New York, where governor George Pataki moved so far to the left in wooing Hispanics, union members, government workers, and homosexuals that he earned the endorsement of the liberal New York Times.
Although the RNC claimed that Pataki won "nearly 50 percent" of Hispanics, John Mollenkopf, director of the Center for Urban Research at the Graduate Center of the City University of New York, examined precinct-level results and estimated that Pataki actually garnered 38 percent.
While this would not be spectacularly high for a Republican presidential candidate (Ronald Reagan did about as well in 1984), the Puerto Rican and Dominican-Americans prevalent in New York tend to vote to the left of the Mexican-Americans of California and Texas, and far to the left of Florida's Cubans. So, Pataki's estimated 38 percent was particularly strong compared to Bush's 18 percent in New York State two years ago (VNS).
Of course, to do this Pataki had to take stands that would make many Republicans shudder. Pro-Democratic election analyst Ruy Teixeira of the Century Foundation and the liberal The American Prospect magazine joked to UPI, "Compared to the ideological center of gravity of the Republican Party, Pataki is now a Marxist-Leninist."
Overall, African-American turnout appeared to be down compared to the strong levels achieved in 2000. In 10 of the 11 states with Fox or L.A. Times exit polls, the black share of the electorate was smaller than in the 2000 VNS exit poll.
In 2000, Democrats won about 90 percent of the black vote. There's no evidence from the eleven states with exit polls that Republicans improved on this in 2002. In most of the polled states with significant black electorates, the Democratic share of the African-American vote was even higher than in 2000.
In 1992, Bill Clinton captured only 30 percent of the Asian-American vote. By 2000, Al Gore had roughly doubled that. The good news, such as it is, for the GOP in 2002 was that in California, the only state with adequate data on Asians this year, the Republicans seem to have stopped the bleeding: Davis beat Simon 54-37, which is similar to how Gore beat Bush last time.
Among Jews, Gore mauled Bush 80-17. Many conservatives had expressed hopes this year that Jewish voters would swing sharply to the right in appreciation of the strong stand on various Middle Eastern issues taken by Republicans in this trying time for Israel. Only a small change, however, was evident in the exit poll data.
In California, Jews voted for Davis 69-22. In Florida, the President's brother lost 73-27. In New Jersey, the Fox poll found Jews voting for Frank Lautenberg, who is Jewish, at a 71-24 rate. Meanwhile, an exit poll sponsored by the New Jersey Jewish News, saw Lautenberg winning 80-19. Interestingly, the last time Lautenberg was elected to the Senate (back in 1994), the same Jewish News poll found him gaining only 65% of Jewish votes.
In California, the four percent of the electorate who identified themselves as homosexuals voted 69-10 for the Democrat, with 21 percent going for 3rd party candidates. Republicans probably did better with gays in Pataki's New York, but no data are yet available.
The gender gap remained sizable in many close elections, but ever since it was first noticed in the 1980 election, its effects have proved to be largely a wash between the two parties. In recent years, a huge gap between single women and more conservative married women has opened. An earlier Gallup poll released Oct. 11, found that only 32 percent of single women were intending to vote Republican compared to 58 percent of married women. Unfortunately, no data from a time closer to the election was available.
In summary, these stereotype-undermining results open up many new possibilities unimagined in the crabbed thinking that prevailed before last week about the demographic fate of the American political system.
It seems likely that Democrats will need to find ways to motivate more minorities to vote, without further alienating the four-fifths of the electorate that is white.
Republicans, despite their excellent showing among whites in 2002, will ultimately need to confront the realities that that the current mass immigration system is slowly reducing their white base's share of the population, and that they have not yet shown an overall ability to win more minority votes. The GOP's choice would seem to be to eventually follow Pataki's path to the left, or to alter the immigration system so that it admits fewer potential Democrats.