Charles Murray on (the) Pew (quiz)

Criticizing a joint effort by PBS and Pew Research to construct a 12-item party identification quiz, Charles Murray makes a few questionable assertions. Firstly, his commentary on the "I never doubt the existence of God" item:

So let’s say you’re the Pew researcher making up the test, you know that conservatives tend to be more religious than liberals, and you want an item that will capture that tendency. So far so good. But then do you say to yourself, “The very most religious people are the ones who never doubt the existence of God”? “Lack of doubt in the existence of God has a statistical correlation with conservatism”? This item is clueless. Clueless about the nature of deep religious faith, clueless about the relationship of conservatism to religiosity, and clueless about the different tendencies of libertarians and conservatives regarding religious faith.

Murray's elitism (and I say that not with derision but with admiration, as the best minds are almost by necessity elitist) causes him to be overly pedantic and demanding of too much semantic precision. Blaise Pascal and Mother Theresa may have perpetually struggled with doubt, but generally speaking, the most pious tend to leave the least room for an existence in which God isn't extant, as the GSS clearly illustrates. The percentages of people who express certainty in God's existence by political orientation and then by frequency of church attendance (with no double counting, so "monthly +" includes only those who attend less than weekly but at least once a month, etc), from 2000 onward and among whites exclusively for contemporary relevance and to avoid racial confounding:

Orientation	Certain
Liberal	37.3%
Moderate	58.4%
Conservative	69.3%

Frequency	Certain
Weekly+	86.3%
Monthly+	68.8%
Annually+	49.3%
Less than annually	33.6%

If we really want to pile on, we can criticize the availability of four responses (completely agree, mostly agree, mostly disagree, completely disagree) to a clearly dichotomous question. How is it possible for one to "mostly agree" that he never doubts the existence of God? If he has ever doubted God's existence, he must choose "completely disagree" as his answer; otherwise, he must choose "completely agree".

This is the nature of most of Murray's criticism, and he could've gone a lot deeper on several of the items if he'd wanted to. For example, the item "Gays and lesbians should be allowed to marry legally" could easily be answered in the affirmative by someone entirely opposed to same-sex marriages. Yes, gays should be able to marry but no, we should not change the definition of marriage to include incestuous, multiple, or same-sex partners. A gay man is allowed to marry a woman he is not related to just as a heterosexual man is! However, the sliver of the population that might logically interpret the question and proffered response in this way is going to realize what the survey is intending to measure and thus answer that they completely disagree with the item even though, strictly speaking, they may completely agree with it.

But that's not the point of surveys like these. Look, social survey data aimed at the general public are inevitably going to be found lacking in nuance and precision, but that's more of a feature than a bug. Better to have informed (and those who read Murray are in the top decile of the population) respondents scoff at the simplicity or semantic sloppiness of a question yet answer in the same way as they would have had it been worded more cogently than risk having a substantial slice of the relatively uninformed masses become confused by the question and offer random or bemused responses that confound the entire data pool.

Murray also utters this doozy on the question about immigration, a subject for which I recall John Derbyshire having noted causes otherwise perspicacious pundits to momentarily lose several IQ points upon discussing:

Hasn’t anyone at Pew or PBS noticed that unions, mainstays of Democratic party funding, are among the most vocal critics of immigration?

The AFL-CIO's official position on immigration:

The U.S. immigration system is broken—and U.S.-born workers as well as aspiring citizens are paying a heavy price. America needs to create an immigration process that works for working people—not a system that benefits corporate employers at the expense of everyone else.

 Current U.S. immigration policy is a blueprint for employer manipulation and abuse, and both new American immigrants and American-born workers are suffering the consequences.

 We say, “¡Basta Ya!” or “Enough Already!” That’s why the AFL-CIO supports a comprehensive, worker-centered approach as part of a common-sense immigration process.

Click here to read the union movement's framework for creating a roadmap to citizenship.

To the right of the page linked to are a couple of 'statements and resolutions' explaining the union's support for "in-state tuition of undocumented youth" and "the devastating effects of state anti-immigrant laws". The language used by the nation's largest unions are indistinguishable from the words employed by La Raza. The rank-and-file might not be big fans of open borders, but, like their non-union brethren, their appointed leaders are.

GSS variables used: YEAR(2000-2010), RACECEN1(1), GOD(1-5)(6), POLVIEWS(1-3)(4)(5-7), ATTEND(0-1)(2-3)(4-6)(7-8)