The Socio-Economic Dimension
Income influenced whether voters cast their ballot for Webb or Allen, but not how you might think. Webb captured a majority of the two-party vote among voters who earned less than $30,000 per year, but he also won a majority of the two-party among voters who earned an annual income of $150,000 or more. Allen captured a majority of the broad middle of the income spectrum (loosely speaking), ranging between $30,000 and $150,000.
In short, there isn't a clear correlation between income and the vote in the Senate race, because Webb appears to have captured both extremes of the income spectrum, whereas Allen captured the income categories in the middle.
To make the effect of income clearer, I'll examine this and other categories by imagining what the outcome of the election would have been like if only certain groups could vote. Based on my back-of-the-envelope calculations, here's how income affected the race:
*If only persons with an annual income of less than $30,000 could vote, Webb would have won with 61% of the vote.
*If only persons with an annual income of more than $150,000 could vote, Webb would have won with 55.4% of the vote.
*If voting was limited to persons with an annual income between $30,000 and $150,000, Allen would have won with 54% of the vote.
Webb's vote percentage among people with an annual income of $150,000+ can probably be attributed to his support among Northern Virginia professionals with post-graduate degrees.
The Religious Dimension
Allen won clear majorities of both self-identified Protestants (58% vs. 42%) and Catholics (51% vs. 49%), while Webb won even larger majorities among the groups who identified their religion as Other (78% vs. 22%) or None (73% vs. 27%). The Jewish portion of the Virginia electorate was too small to get accurate exit poll data, but Webb probably won that group handily as well, not least because of Allen casting "aspersions" on his putative Jewish heritage.
In other words, the main religious cleavage in the Virginia electorate appears to be between those who self-identify with a Christian denomination and those who do not. In fact, the "None" group appears to consist primarily of people raised as Catholics or Protestants, but who claim no religious affiliation due to their dissatisfaction with how religion has been politicized by the Christian Right. Based on my calculations, the Virginia Democratic Party owes a big thanks to its secular and non-Christian base.
*If only self-identified Christians could vote (i.e., those who identified themselves to pollsters as Protestant or Catholic), Allen would have won with 56.5% of the vote.
*If only religious "Others" and "None" could vote, Webb would have won with 75% of the vote.
The Racial/Ethnic Dimension
Webb won the votes of 75% of nonwhite men and 82% of the votes of nonwhite women. To break it down further by individual racial categories, Webb won 85% of the African-American vote and 68% of the Asian vote, which suggests that Allen's love for nooses, the Confederacy, and the N-word may have influenced more of the nonwhite electorate than the "macaca" incident (which would have had more resonance for Asians).
*If only nonwhites could vote, Webb would have won with 78.8% of the vote.
*If only whites could vote, Allen would have won with 57.5% of the vote.
When looked at in combination with religious affiliation, the effects of race on the Webb/Allen election get even more interesting.
*If only white self-identified Christians could vote (something that Allen's "base" might like to see become a reality), Allen would have won with 64% of the vote.
*If only whites in the "None" or "Other" religious categories could vote, Webb would have won 70.3% of the vote.
*If only whites who identify as "born again" or evangelical could vote, Allen would have won 80% of the vote.
*On the other hand, if the electorate were limited to nonwhites who identify as "born again" or evangelical, Webb would have won with 70.5% of the vote.
In other words, if you're white, religious affiliation (or non-affiliation) trumps race in influencing how you vote. But if you're nonwhite, race trumps religion, even among those who identify as evangelical or "born again." If I had better racial breakdowns, I could find out more, but I suspect that this is a result of how Allen's racist antics have royally pissed off evangelical African-Americans, even though they might agree with Allen on social issues.
Urban vs. Rural Conflict
Sociologists and political scientists disagree considerably about how to define the word "urban." Some definitions argue that the essence of rural life is agriculture. Therefore, an area is defined as urban to the extent that it is removed far away from any agriculture. On the other hand, with the rise of agribusiness and edge cities springing up near farmland, this definition of "urban" is not with out its critics. Other academics in the field of urban studies argue that the essence of urban life is population density. If you live in an area with high population density or you live near such an area, then you are "urban."
According to CNN exit polls, Webb captured 61% of the urban vote, while Allen captured 55% of the suburban vote and 52% of the rural vote. Since the rural vote was actually slightly lower for Allen than the suburban vote (although the difference may be statistically meaningless), I would argue that rural vs. urban issues probably had low salience in Webb vs. Allen race. Despite all the urbanites who loathe suburban life, the suburbs are actually more "urban" than they are "rural" when it comes to their lack of connection to agriculture or the physical amenities they offer. Instead, the primary factor that makes suburbia more culturally "rural" than most urban areas is the lack of racial and ethnic diversity. This suggests that race might be driving the urban vs. suburban divide in the Webb vs. Allen race more than any city vs. country conflict per se.
Regime Support
In this instance, I will use the word "regime" to refer to how a voter feels about the Bush administration. As it turns out, "regime support" might have been the most important dimension of conflict in the Webb vs. Allen election.
*If only those who approved of George W. Bush were allowed to vote, Allen would have won with 93% of the vote.
*If only those who disapproved of George W. Bush were allowed to vote, Webb would have won with 87% of the vote.
Foreign Policy
How foreign policy influenced a voter's vote can be determined by whether a voter viewed "the Iraq War" or "terrorism" as a more important issue. To be brief, voters who focused on Iraq generally favored Webb, whereas voters who focused on terrorism generally favored Allen. Among voters who rate the Iraq War as "extremely important," 58% supported Webb. Among voters who rated the Iraq War as less than "extremely important," Allen won a clear majority. Allen won among voters who rated terrorism as "extremely important" or "very important," but Webb beat Allen among voters who rated terrorism as "somewhat important" or "not very important."
I'll leave the discussion of materialism vs. post-materialism for another time, but for now it seems that the Webb vs. Allen race was driven primarily by divisions within the electorate about the Iraq War and dissatisfaction with the Bush administration, as well as underlying social cleavages related to race and religion (i.e., white self-identified Christians vs. everybody else). Economics and the urban vs. rural issue had a less predictable effect on the election that was probably driven by the other factors mentioned above.