Complexity and Social Networks Blog

« Fun with numbers: did Obama redraw the political map? | Main | 311: The Next Wave" - Harvard online event 11/13/08 »

5 November 2008

How Obama won: Conversion or mobilization?

Did Obama win by mobilizing groups who have supported Democrats in the past, or converting new people to support him?

Below is a plot of the support various demographic groups provided Obama in 2008 to the support they provided Kerry in 2004. Groups that increased their share of the electoral pie by more than 10% are represented by a larger dot. The question: did Obama's victory come from supportive groups (African Americans, young voters) turning out in greater numbers, or from moving all groups to greater support of him, as compared to Kerry?

The plot suggests that while the answer is some of both, Obama's victory may be more about conversion than mobilization. In particular, note that all groups plotted increased their support of Obama (if you have suggestions of other ways of slicing the data, please suggest and I will try to update the plot). While there was an impressive mobilization of African American voters (who have a strong tendency to vote Democratic), there were also other groups less supportive of Democrats who made up a larger fraction of the electorate in 2008 than in 2004-- most notably, evangelical voters. However, interestingly, even evangelicals were more likely to support Obama than Kerry.

Other interesting nuggets:

Young voters did not make up a notably larger share of the electorate (note, though, that turnout was up about 7% overall, so it took a really impressive jump to increase share of the electorate). They did shift more than most other voters in terms of support for Obama, although not dramatically so (here note that they had supported Kerry at high levels). The big picture here, though, for the Dems is a good one-- the young voters are settling into very Democratic voting patterns (2:1 for Obama), which research suggests is predictive of a lifelong attachment.

Jewish voters were about typical in their shift toward Obama (notable because of the issues raised by the McCain campaign regarding Obama's support of Israel, and also because of Lieberman's very visible support of McCain).

The group that shifted the most, in absolute terms, toward McCain were Hispanic voters. Also a good long run sign for the Democrats.

Frequent church goers shifted a bit more toward Obama than infrequent church goers.

Alas, no data on plumbers.

Of course, all of the above needs to be taken with many grains of salt about methodology-- e.g., about the statistical significance of the differences, the potential variations in exit poll methods iin 2004 and 2008, etc etc. To really get at these issues one would really want panel data, and so on. The above is really meant at provoking some thought as to the underlying dynamics driving the outcome yesterday.

Posted by David Lazer at November 5, 2008 3:15 PM

.
• Business (3)
.
• Citizen Relationship Management (17)
.
• Commentary (7)
.
• Complexity (2)
.
• Current Trends (48)
.
• eGovernment (21)
.
• Events/Announcements (74)
.
• Innovation (14)
.
• Internet (18)
.
• Knowledge Sharing (20)
.
• Methodology (32)
.
• Networked Governance (18)
.
• Networks in political science (5)
.
• PNG Working Paper Series (6)
.
• Powerlaws (4)
.
• Reviews (2)
.
• Simulations (1)
.
• Small group networks (3)
.
• Small Worlds (2)
.
• SN in the news (72)
.
• Social Capital (3)
.
• Social Computing (12)
.
• Social Finance (1)
.
• Social Networking Platforms (11)
.
• Social Psychology (2)
.
• Social Software (25)
.
• Societal Networks (40)
.
• Technology (35)