Monday, August 06, 2007

Candidate Trait Bias

The NY Times's July 22, 2007 "Week In Review" focused on the relationship between religion and politics, and particularly Mitt Romney's Mormonism, in an article by Michael Luo titled "God ’08: Whose, and How Much, Will Voters Accept?".

Mrs. Van Steenis wanted Mr. Romney to say where the Book of Mormon would figure in his decision making as president.

“Where would the Bible be?” she asked. “Would it be above the Book of the Mormon, or would it be beneath it?”

Although the Constitution bars any religious test for office, if polls are to be believed, Mr. Romney, the former Massachusetts governor, faces a serious obstacle to winning the presidency because of his faith. Surveys show a substantial percentage of Americans would be less likely to vote for a Mormon, or for that matter a Muslim or an atheist. But how rigid is that sentiment?
I found the article to be particularly fascinating because it includes a chart summarizing a Pew Research Center survey which asked respondents whether they were more or less likely to support a candidate based upon individual traits, such as: Doesn't believe in God, Never held elected office, In his/her 70s, Homosexual, Muslim, No college education, Used drugs in past, Had extramarital affair, Taken antidepressants, Mormon, Has been a minister, Smokes cigarettes, Long-time Washington politician, Hispanic, etc.

For example, I find it interesting that there is 60% net support for belief in God, but 10% net opposition to having been a minister. Similarly, the article highlights belief in God but, considering the ranking of Muslim and Mormon, it actually seems to be a rather weak trait by itself and perhaps only significant by association with other more relevant qualities.

I think the chart is also a bit deceptive, which the last question in the blockquote above highlights. How much weight do we give each of these traits? I'd be interested to learn if there's a survey in which respondents attempt the difficult task of weighting them.

Does the fact that the "Constitution bars any religious test for office" imply a broader moral imperative for voters to ignore religion in evaluating a candidate? Do you think the article suggests this?

More generally, do you think the trait biases evident in the survey are justifiable or are they representative of the unjust prejudices in our society? For example, how much of the support (4%) and opposition (14%) to a Hispanic candidate represents racism versus, for example, the expectation of a correlation between being Hispanic and specific views on immigration? Or does the latter count as racism as well?

(hat tip to Andrew Jackson at RedBlueChristian)

4 comments:

MarkC said...

Kevin,

A few good questions here, and like you I am particularly fascinated by that chart. But, back to that in a minute. First, the easy stuff:

Does the fact that the "Constitution bars any religious test for office" imply a broader moral imperative for voters to ignore religion in evaluating a candidate?

Definitely not. The Constitution prohibits a codified religious test. Individuals can vote based on whatever factors they think are important.

do you think the trait biases evident in the survey are justifiable or are they representative of the unjust prejudices in our society?

I think the survey is at best suggestive of general trends, but is useless at explaining the underlying reasons for those trends. It does nothing to differentiate between racist anti-Hispanics and those who are suspicious of Hispanics because of political views often held by Hispanic politicians.

Now, back to the chart.

I was fascinated (and surprised) by what I saw when I flipped the chart around. The NY Times sorted the chart in order of highest "Less Likely to Support" answers. When you sort it instead by highest "More Likely to Support" percentages, the results are interesting. Only five traits got "More Likely to Support" percentages above 20% (compared to eleven less-likelies!). In order, they are:

* Military service (48%)
* Christian (39%)
* Long-time Washington politician (35%)
* Former business executive (28%)
* Attended prestigious university (22%)

Not only did all of those have high positive numbers, none of them had over 15% negative numbers, so there aren't any highly-divisive traits here. These are traits that are generally seen as strong indicators of a good President.

Military service is the most obvious and unanimous (it has both the highest positive percentage and the lowest negative percentage on the list). But I would have thought that "long-time Washington politician" would have elicited a much more ambivalent response.

Clearly, American voters are concerned about experience. I'd hypothesize that most American voters hate that politics are a game, and polls show that Americans deeply distrust politicians... but at the same time, as long as the game is being played at such high stakes, we want somebody in the office who knows how to play it.

I'd wager that the high results for former business executives come primarily from people looking for financial accountability, since there is at least a perception that private industry is better at managing money and controlling spending than the public sector. We've seen in recent elections, though, that prominent business leaders, though able to garner initial support, generally cannot make it far even in their party primaries. (Also, the 13% negative response shows a corrollary distrust of private sector leaders, probably because of the perception of rampant corruption.)

I'm personally surprised to see so many people who care about whether a politician is "Christian". I say that not because I think a candidate's religion is unimportant... but because I think a candidate's actual religious beliefs are nearly impossible for us to know! On policy positions, a candidate has a voting record and other various forms of accountability. On religion, though, a candidate can say they are whatever religion they want, go to a corresponding church, and it tends to show nothing of their true heart belief. (Note that the Mormon church, as a highly-organized niche religious group, is an exception to this rule, so we do know something of Romney's core religious principles.)

Anyway, thanks for pointing out this study. It's very fascinating.

Mark

Kevin said...

Mark,

Thanks for your excellent analysis of the preferred traits. You make a lot of sense to me. I agree throughout. And you cut through my "easy stuff" like butter. :)

I agree regarding the Constitution, though I do get whiffs from the article and elsewhere that it is politically incorrect to consider religion. And in that vein, you do a good job of highlighting our inability to learn a candidate's true religion and the weakness of broad terms such as "Christian". I imagine it would have a small weight for that reason.

You're right that my question regarding prejudice probes the reader's perceptions of society or his own hypothetical responses as opposed to the data of the survey. It's also loosely related to some thoughts on (e.g. racial) profiling and the broader morality of discriminating based upon protected attributes when they statistically correlate with more relevant factors.

Kevin

MamasBoy said...

Mark,

I agreed with much of what you said, but find myself hesitating at this statement.

"On religion, though, a candidate can say they are whatever religion they want, go to a corresponding church, and it tends to show nothing of their true heart belief. (Note that the Mormon church, as a highly-organized niche religious group, is an exception to this rule, so we do know something of Romney's core religious principles.)"

I'm not sure that a candidate being a Mormon is much more telling than a candidate being a Baptist or a Catholic. Where "heart belief" is concerned, I find hypocrites on all camps. Moving beyond simple heart belief to the predictive nature of religious affiliation, both Mormons and Catholics could be referred to as highly organized groups, whose members beliefs you can nail down fairly well by observing whether or not they actively participate in religious services and to some extent who within their religious group they hang out with. It is trickier with Baptists, because their beliefs are not documented as formally as the other two, but there are definite camps with highly refined belief systems on a less formal scale. In their case, though, I think one can predict with equally reliable accuracy what they believe on beliefs that relate to politics by observing the beliefs and statements of of the Baptists that they associate with.

In some ways, it is less predictive to know that a candidate is a Mormon than that a candidate is a strong Baptist or Catholic. There are many things in Mormonism that are handled with more subtlety than may appear at first glance (e.g., abortion and divorce). I've talked with Mormon missionaries about those example issues and had them issue strong condemnations of them. However, when I read the book of covenants that they gave me, I found the views to be much more nuanced and ambiguous. Honestly, I've only met a couple mormons missionaries who understood and could accurately explain what their church teaches on those topics, and I've talked to many. As a side note, it can be rather disappointing to finally find one who understands what their church teaches and can accurately answer questions, only to have them moved along to a different locale a few weeks later and replaced by another "less insightful" person. If this is what the missionaries are like, I find it frightening to imagine what the average Mormon knows about their church... It's almost as scary as the ignorance of most Catholics.

Perhaps you were simply going on the predictive nature of the title Mormon as opposed to the title Christian. In that sense, perhaps you would be correct, since even Mormons use the title Christian. Would you happen to have any examples of what you meant up there?

MB

MarkC said...

MB,

I agree that Mormonism is less predictive of specific political positions than many other religions. You're right there. That wasn't quite my point.

What I mean is that, because Mormonism is still a niche religion in our country that is not itself politically advantageous (except maybe in Utah), if a politician is openly Mormon, you can assume that they actual view themselves as Mormon.

Catholicism (or other prominent versions of Christianity) on the other hand is clearly a huge benefit to a politician (especially over, say, atheism). So, there is strong motivation for non-Christian politicians or very nominal Christians to maintain their external Christian identity and the basic forms of their religion purely for show.

No politician would maintain external Mormonism purely for show. What would be the point?

That's what I mean. I'm pretty confident that Romney is actually a practicing and believing Mormon. I have really no idea if Giuliani is a practicing and believing Catholic, or if McCain is a practicing and believing whatever-he-claims-to-be.

Mark