Bigotry or No Bigotry? That is the question!


Mitt and Ann Romney

The day before Mitt lost the primaries in Alabama and Mississippi on 13 March, the Internet-based marketing firm,, published an article by Michael Tesler entitled “Evangelical Opposition to Romney NOT Rooted in Anti-Mormonism.” Based in the UK, seems to be a serious organization. Nevertheless, it appears that this author falls prey to the classic problem of confusing correlation with causality, an issue I raised earlier.

Following the results in the South and a couple of days before seeing the YouGov piece, I wrote a post that included this paragraph:

While mobs no longer rage in the South, feelings of ill will are obliquely mentioned in the press (“Romney fails to attract the ‘very conservative’ in South Carolina”). People looking at that will often exclaim that this does not mean that these people oppose Mitt on religious grounds, punctuating their objections that the most religious people happen to also be the most conservative. Indeed, as any statistics professor will explain early in any course, correlation does not indicate causality. The evidence that the correlation means something more in this case, however, seems clearer than many will accept.

Just as one cannot use correlation to prove causality, however, one cannot use correlation to prove the opposite case, which would involve trying use correlation to prove that people do not reject Mitt Romney because of his religion. It appears that Tesler’s approach in his YouGov article falls into this same category of problem. In reporting results from a study he conducted to support that notion, he cites three different sources of media speculation to the contrary and then continues:

Media speculation abounds that Mitt Romney’s poor performance among Evangelical voters in the 2012 Primaries is rooted in anti-Mormonism—a sentiment that will surely intensify if the former governor loses this week in Alabama and Mississippi. My analysis of seven surveys conducted by YouGov from late January to early March 2012 (pooled n=7,000, with 1,791 likely Republican primary voters), however, suggests that Romney’s religion is not the main reason why he has not won over these voters.

Setting statistical measures aside, let’s take a look at some specifics that I feel provide a different picture. First, let’s consider what members of various Evangelical groups and congregations have said and are saying. While some of these developments will no doubt have minimal impact, taken as a group they counter the assertion by Mr. Tesler and others that Mitt’s religion is a problem for his electoral chances. Rather than taking the time to discuss each instance in detail, I will mainly provide a list of statements with links to the various sources.

2008 Election Cycle

2012 Election Cycle

Page 1 of 3 | Next page