Misguided Analysis of Electability

Several articles I have read about the electability of Senators Obama and Clinton indicate that many people (especially Clinton supporters) are misguided in their analysis.  An example is a an article from today’s issue of The Washington Post.  The article discusses that Obama’s strategy has been to rack up delegates in lots of Republican states (many of which held caucuses) to overcome the advantage Clinton has in Democratic strongholds.  Several people interviewed in the article expressed concern about the fact that Clinton has bested Obama within several demographic groups such as working-class whites and Latinos.

These people are misguided because all that really matters is whether Obama or Clinton can beat McCain in enough states to win enough votes in the Electoral College in November.  It doesn’t matter if a group of people prefers Clinton to Obama as long as they prefer Obama to McCain.  For instance, there is very little doubt that California, New York, and Massachusetts would vote for Obama over McCain in November even though these states voted for Clinton over Obama in the Super Tuesday primaries.  Likewise, it doesn’t really matter whether Latinos prefer Clinton to Obama; what matters is whether they prefer Clinton or Obama to McCain.

I have not yet seen any polls comparing how Latinos would currently vote in a McCain/Obama contest, but a CNN poll of Latinos in Texas found that 81% of Latinos in Texas a quick troop withdrawal from Iraq.  That might lead them to support Obama over McCain even though they went for Clinton in the Texas primary.    Additionally, the CNN article about the poll reported that a prominent Latino Republican adviser, Lionel Sosa, has repeatedly warned that the Republican Party will lose the Latino vote if it does not change its rhetoric on immigration.  While all Democratic candidates have actively been campaigning in Latino communities, the Republican candidates were “fighting to see who is more anti-immigration” according to Lionel Sosa.  McCain supports a border fence but has had a softer approach on the immigration issue than most other Republicans.  The poll also showed that 78% of Texas Latinos favor bilingual education programs, another thing Republicans have opposed.  Finally, even though Texas Latinos leaned toward Clinton, it is not the case that they did not like Obama; the poll showed that 76% of them had a favorable view of Clinton, 66% had a favorable view of Obama, and only 48% had a favorable view of McCain.  (Only 34% had a favorable view of Bush.)  Given all this data, I suspect that Latinos in Texas, California, and other states would strongly support Obama over McCain.  So, their preference for Clinton over Obama is no reason to worry about his losing their votes in November. 

Along these lines, a recent state-by-state poll by SurveyUSA predicted that Obama would beat McCain in the Electoral College 280 to 258, wining the Northeast states, the Potomac states, the upper Midwest, the West coast, Hawaii, Colorado, Nevada, and New Mexico but losing Florida, Pennsylvania, and New Jersey.  SurveyUSA predicted that Clinton would also beat McCain in the Electoral College 276 to 262, winning the Northeast (including Pennsylvania and New Jersey but not New Hampshire), Maryland (but not Virginia), some of the upper Midwest (but not as many as Obama), California (but surprisingly not Washington and Oregon), Hawaii, New Mexico, and Florida.  So, the margins would be quite similar for Obama and Clinton.

Here are the electoral college maps published by Survey USA:

 

Tags: 2008 Election, Clinton, Clinton vs. McCain, electability, McCain, Obama, Obama vs. McCain, poll of Latinos, SurveyUSA