Gaming Out California

In previous posts, I suggested that California will probably be the deciding battleground for Super Tuesday and for the Democratic nomination. After Saturday's results in Nevada and Obama's inability to get over 35% of the white vote in any of the contests so far, the odds on him winning California are seeming longer and longer. I did some back of the envelope calculations to see how things might work out. Based on the partisanship of various ethnic groups in the 2004 election, I estimated that Blacks will be 10 percent of the Democratic primary voters, Latinos 25 percent, and Whites, Asians, and Other 65 percent. I next estimated that Edwards will only be a small factor in California. His results since Iowa have been disappointing and he has little money to compete in California. I estimated that he'll get about 8 percent of the vote (10 percent from White/Asian/Other, 5 percent from blacks, and 5 percent from Latinos). Given this breakdown of primary voters, if Obama gets 35 percent of the White/Asian/Other vote (his ceiling so far), 25 percent of Latinos (his Nevada result), and 70 percent of the Black vote (the high end of what he's getting in polls in South Carolina), he would end with 36 percent of the vote to around 55 percent for Clinton.

Scenario 1:

	% of the Electorate	Obama %	Obama Performance	Clinton %	Clinton Performance	Edwards %	Edwards Performance
White/Asian/Other	65	0.35	22.75	0.55	35.75	0.1	6.5
Latino	25	0.25	6.25	0.7	17.5	0.05	1.25
Black	10	0.7	7	0.25	2.5	0.05	0.5
Result			36		55.75		8.25

A victory for Obama in this turnout scenario would require him to split the White/Asian/Other vote with Clinton 45-45, get at least 35 percent of the Latino vote, and 80 percent of the Black vote.

Scenario Two:

	% of the Electorate	Obama %	Obama Performance	Clinton %	Clinton Performance	Edwards %	Edwards Performance
White/Asian/Other	65	0.45	29.25	0.45	29.25	0.1	6.5
Latino	25	0.35	8.75	0.6	15	0.05	1.25
Black	10	0.8	8	0.15	1.5	0.05	0.5
Outcome			46		45.75		8.25

Of course, different turnout scenarios will lead to different outcomes. For example, support for Obama may lead to a big increase in black turnout or perhaps Edwards will manage to grab a bigger share of the White vote from Clinton. Nonetheless, these estimates indicate that an Obama victory in California will be very tough.

Update: Sorry about the tables spilling over the right margin. I don't know how to avoid that. A reader wants to know if I factored in the open/closed nature of the Democratic primary, especially, as Kevin Drum points out, the Democratic primary is open while the Republican primary is closed. Given Obama's support among independents, this obviously helps him. It remains to be seen how much.