An exit poll on the 2004 political contest between Democratic candidate Obama and Republican candidate Alan Keyes is revealing about contemporary voter patterns. It may show some problems for presidential candidate Obama as well. At the last minute, Alan Keyes was drafted by Republicans to replace the late Republican dropout Jack Ryan just 86 days before the election; Keyes had run twice and lost for U.S. Senator in Maryland. Both candidates are African Americans, probably the only occurrence of such a Senate contest in the history of U.S. politics, at least since Reconstruction.
Keyes, a former diplomat, well-known media commentator, and perennial political candidate (now including the 2008 race for president), took arch-conservative positions on issues such as abortion and gay rights, at one point causing controversy by asserting that Jesus Christ would not vote for Obama because of positions on abortion.
Not surprisingly, given the lateness of the Keyes entry, his Maryland residency, and the blue state record of Illinois, Obama, a veteran of the Illinois State Senate and former community activist and top Harvard law student, won the election with nearly three quarters of the total vote (nearly a record for Illinois) and became only the third African American member of the U.S. Senate since Reconstruction (and only the fifth ever elected). Obama is widely considered a viable potential Democratic nominee, if not (yet) a front runner.
Nonetheless, details of the voting have implications for how Obama might be viewed by key segments of the voting population in election battles coming up soon in 2008.
According to a 2004 CNN exit poll taken on election day, women voters outnumbered male voters, and Obama got a larger percentage of female than male voters (73 percent versus 67 percent). Whites made up some 78 percent of the voters, with African Americans at 10 percent and Latinos at 9 percent. Among the three racial groups tabulated, whites gave Obama the lowest percentage of their votes (66 percent), with African Americans and Latinos giving him much higher percentages (92 percent and 82 percent respectively). Among four racial-gender groups, white men gave him the lowest percentage (64 percent) of the racial-gender groups tabulated, and women of color gave him the highest percentage (88 percent). Clearly, even against very weak opposition, the liberal Obama lost 36 percent of the white male vote. Apparently some whites saw Keyes as “essentially white” in his extremely conservative social and political views, those very characteristic of arch-conservative white candidates like George Bush.
The poll found little difference among age groups, with all but those 60 and older giving Obama at least 70 percent of their votes. Those in the highest age group gave him 65 percent of their votes. There was also little variation by income or union membership.
There was considerable variation by ideology and religion, with those saying they were liberal (23 percent of those sampled) giving Obama 91 percent of their votes, and those saying they were conservative giving him just 33 percent of their votes. Those who said they were white conservative Protestants gave him only 18 percent of their votes, while those who said they were Catholics (presumably mostly white) gave him 75 percent of their votes.
The voters were asked what the most important election issue for them was, and there was considerable variation in voting along these lines. The four top issues were Iraq, the economy, moral values, and terrorism, all issues accented by 18-21 percent of these voters. Obama got more than 88 percent of those who accented Iraq or the economy, but only 51 of those who accented terrorism and just 38 percent of those who accented moral values. On a question about the most important quality of a candidate, those who said “will bring change” gave 97 percent of their votes to Obama, which contrasted with only 26 percent of those who said “religious faith.”
As one might expect, there was significant variation by place of residence, with city voters giving Obama 80 percent of their votes, and rural voters just 58 percent of their votes, with suburban voters falling in between at 70 percent.
John Jackson, a 2006 visiting professor as Southern Illinois University (Carbondale), has noted that the significance of this lower rural vote in the fact that eight of the ten Illinois counties Obama lost to Keyes were together in a block in southern Illinois, an area that often votes like the South:
“These include Effingham, Jasper, Clark, Clay, Richland, Wayne, Edwards, and Wabash counties. These are rural and small town counties where cultural conservatism has always been strong and where religious fundamentalism is growing in its impact on politics.”
These are countries that are hard hit by corporate outsourcing of good-paying, often union jobs, but they still tilt rather Republican in their voting patterns. Given that 2004 was an extremely good year for Democratic candidates, and that Obama ran against an extraordinarily weak Republican candidate who had only recently entered the race and was widely expected to lose, this voting pattern suggests that in many parts of the South, white voters still are not inclined to vote for Democrats, especially black Democrats, in part because they see such Democrats as “essentially black” and as “weak” on moral and “family values.”
Jackson put the dilemma for an Obama national candidacy in these savvy words:
“Barack Obama has taken some initial steps toward doing exactly that in building coalitions during his first Senate race in Illinois, which is a large and diverse state. He successfully appealed to virtually all parts of the state in the general election against a controversial opponent. However even against Alan Keyes, Obama lost ten counties in areas indicative of the places where Republicans have recently won nationwide recently. The results in his first primary against Dan Hynes and Blair Hull [Democratic Candidates he ran against in the primary] indicate that rural and small town voters in Middle America are not his most natural constituency. Whether he can use that experience, build on it and take it to success on the national level is an intriguing question at this point.”
It remains to be seen if the candidates in 2008 have learned anything about race and gender from the 2004 campaign.
It will be interesting to see what happens in the upcoming election; will people vote for political issues first or along the lines of gender and/or “race?”