Wednesday, May 5, 2010

Michael Medved gets the math wrong about white vote

Michael Medved claims that Republicans will be unable to win based on the white vote alone.

I don't have any problems with his arguing this point, but I do have problems with his use of the data.

He writes:

" First, there is no chance that white voters will ever again comprise 74 percent of the electorate. Most projections for 2012 suggest that self-identified whites will comprise 70 percent or, at most, 72 percent of those who cast presidential ballots."

I will bet with Michael Medved or anyone that when the most detailed survey is completed by the Census after 2012, whites will be 74% or more than of the voters.

Medved relies on the exit polls, rather than the more comprehensive data by the U.S Census. According to the later even in 2008 non-Hispanic whites were 76.3% of the vote.

The slow, demographically drive decline of the white vote share is not the reason Obama won in 2008. That process exists, but far too slow to effect matters much. In 2004, non-Hispanic whites were 75.2% of the adult citizen population. In 2008, they were 73.4%. The decline is only 1.8 percentage points, or 0.44 percentage point per year. If the demographic decline was responsible, McCain, who won whites by 12% and lost Hispanics by 36%, would have lost only 0.8% of vote compared to what Bush got in 2004. In fact, McCain did 5.1% worse than Bush.

Only about 0.8% out of the 5.1% McCain lost compared to Bush are due to demographic change, the force Michael Medved focuses on!


So why did McCain lose? Here are several factors, some of which Medved ignores:

1. McCain did worse than Bush among white voters. He won by 12 point, Bush won by 17 points. That is a 5 point decline right there for you.

2. White turnout was lower in 2008 than 2004 , because McCain could not motivate white voters (and because the election was close). There is no reason white voters cannot be motivated to vote more in 2012 or beyond. One factor to keep in mind is that whites are aging (and old people vote more).

3. Non-white turnout was inflated in 2008, because Obama was on the ticket. This effect will not last beyond 2012, and perhaps not even that long. White turnout decreased 1.1 percentage points between 2004 and 2008. Non-white turnout increased by 3.7%.

4. Non-white voters went to Obama by a much larger margin than 2004. There is virtually nothing the GOP could do about minority voters wanting to vote for a minority. If the 2008 election was to be won, with a African American on the ticket, it has to be done through getting more white votes. The same is true for 2012.

The claim "there is no chance that white voters will ever again comprise 74 percent of the electorate." is based on the mistaken belief that slow-working, long term demographic factors determined the decrease in white share between 2004-2008, whereas in fact it was mostly turnout.

Unless something dramatically happens (like amnesty for Hispanics) The white voting share will decline by on average about 1.5-2 percentage points per election cycle, no more. The effect on the Republican vote margin will be marginal for the next 20 years or so to come.

In 2012, non-Hispanic white voters will be about 75% of the voters, and Hispanics about 8%. Because of political correctness, Medved is advising Republicans to target the second group, and ignore the first one, which is 10 times bigger!

Furthermore, it is silly by Medved to claim that Bush's 58% in 2004 was some sort of magical ceiling for the white vote. First, Bush was a mediocre candidate. Second, as the demographics of the country change, white voters will block together, just as blacks and Hispanics engage in block voting. Lastly, the republicans can alter their agenda in order to target white voters better.

There is absolutely nothing that says that Republicans cannot win election by taking 60-65% of the white vote, just as Democrats routinely win 90% of the Black vote, and 70% of the Latino and Asian vote.

Black and Latino voters are concentrated in non-battleground states. One last figure. I have compiled data using a national survey of voters. Only 4.7% share of the battleground state voters were Hispanic in 2008, and 12% black, compared to 81% who were non-Hispanic white.

No comments:

Post a Comment