How did people arrive at a decision during the Democratic primary? And what predicted their movement to or away from Obama as the primary went on? In particular, what role did race play?
Simon Jackman and Lynn Vavreck have a forthcoming paper that speaks to these questions. I think Lynn for kindly sharing with me ahead of publication. Jackman and Vavreck frame their central question thusly:
bq. Are there limits to the momentum that a black candidate can build if he is running against a white opponent? In other words, does “Big Mo” become “Slow Mo” for a black candidate as attitudes about race prevent voters with high levels of racial animus from joining the bandwagon even in a Democratic primary?
To answer this question, they draw on the Cooperative Campaign Analysis Project, which interviewed the same group of voters four times over the course of the presidential primaries: December 2007, January 2008, March 2008, and the, post-primary, in September 2008. The analysis is rich and detailed. Let me highlight several findings.
First, at the outset of the primary, a measure of racial prejudice — known in the political science literature as racial resentment — is strongly associated with preferences. The graph above shows how strongly. Whites who scored high on this scale — the “racially resentful” — were much more likely to support Hillary Clinton or to have no preference, and much less likely to support Obama:
Jackman and Vavreck write:
bq. This point cannot be made too subtly: white Americans voting in the Democratic nominating contests were driven to choice in these elections in large part by their general attitudes toward race in America.
Second, Obama’s Big Mo did indeed become a Slow Mo. He gained a great deal of support from January to March. This was due to three processes:
bq. Clinton loses support to Obama over time; he holds his support better than she does; and people who are forced to make a second (or third) choice are more likely to choose Obama over Clinton.
But after March, his momentum slows considerably:
bq. …among respondents voting in the Democratic primaries held after mid-March, Obama does not gain vote share. Moreover, among this set of voters, Obama loses as much support to Clinton as she loses to him.
Third, and most importantly, attitudes toward race slowed Obama’s momentum. In particular, Clinton’s infamous remark in May — “Senator Obama’s support among working, hard-working Americans, white Americans, is weakening again” — helped “racialize” her candidacy:
bq. We suspect that Clinton’s focus on Obama’s dwindling support among hard-working, white Americans tells much of the story. That is, as Obama’s chances of becoming the nominee increase, the contest becomes more racialized and two things occur, racial liberals move toward him and racial conservatives move away from Obama.
They then analyze why people moved to Obama or away from Obama. Not surprisingly, racial identity mattered: among those who initially preferred Obama, blacks were less likely than whites to move away and support another candidate. Sex also matters: women who initially support some other candidate are more likely than men to end up supporting Clinton; sex does not affect movement from Clinton to Obama.
Surprisingly, family income, education, and a general measure of liberal-conservative ideology have little impact. As much as commentators wanted the primary to be about social class — Obama’s working class problem, etc. — the data just don’t bear out that interpretation.
What does matter is racial resentment:
bq. The combination of Obama’s losses in the March-September period and the increased relevance of attitudes about race for those who are initially unsure who to support highlight the slowing of Obama’s momentum in this period. Among those initially supporting Obama, but voting after March 21st, even those with average levels of racial resentment are much less likely (roughly 20 points) to stay with Obama than are otherwise similar respondents who voted before the March wave. The same is true for those with the highest levels of racial animus who were unsure which candidate they preferred in December; these voters are roughly 20 points less likely to choose Obama if they vote after March 21st than if they vote before this date. Obama’s momentum stalled | and attitudes about race explain a good bit of the slowdown.
Giving the spacing of the panel waves, it is difficult to attribute significance to any particular moment in the campaign, such as Clinton’s remark about hard-working white voters. But the pattern of results signals that racial prejudice was important from the beginning and continued to be so throughout the campaign. Jackman and Vavreck conclude:
bq. By far, the strongest predictor of transitions to and away from Obama are attitudes about race. Increasing levels of racial antipathy lead to lower rates of transition to Obama, across all waves of the nominating process for voters and irrespective of a respondents’ initial preference. Not until late in the process do Obama voters switch away from his candidacy with increasing levels of racial resentment, thus slowing his momentum.
Find the paper here.