The Medium & the Message: A Moral Values Follow-Up

By Howard Schuman

Since I wrote my op-ed piece on the "moral values" exit poll question for the November 2004 preview issue of Public Opinion Pros, a Pew Center postelection survey has provided important new findings. First, as part of a split-sample experiment, the Pew poll replicated the original National Election Pool (NEP) closed question, and the results show that, from a list of six alternatives, the "moral values" response was the one most often selected (by 27 percent of respondents), with the selection made far more frequently by Bush voters (44 percent) than by Kerry voters (7 percent).  Second, a parallel open version of the same question yielded a smaller but still nontrivial proportion of responses coded as "moral values" (14 percent): the category was second in size, well behind "Iraq" but slightly ahead of "economy/jobs," and the distribution of responses was even more lopsided in the direction of Bush voters (27 percent) than Kerry voters (only 2 percent). 

Third, and especially valuable, the Pew poll followed the moral values response to the closed question by asking, "What comes to mind when you think about 'moral values'?"  Although sizeable proportions of respondents mentioned specific issues like gay marriage or abortion, large percentages also mentioned candidate qualities, or cited religious beliefs, or referred to traditional values.  Clearly, "moral values" covers a wide range of candidate characteristics and issues, and the term has a broad appeal to Bush supporters. 

Various commentators continue to refer to the moral values question invidiously, usually in vague critical words. In the January 13, 2005, issue of the New York Review, for instance, Mark Danner calls it "a clumsily posed exit poll question." Some poll questions are indeed seriously defective, like a hypothetical double-barreled inquiry: "Did you support the U.S. invasions of Afghanistan and Iraq?"

But the moral values question does not suffer from such a logical defect, and it is neither better nor worse than most complex survey items.  One criticism made of it is that the other five alternatives are "issues," whereas "moral values" is not an issue.  But there is no reason to require respondents to abide by a distinction that, as their open answers indicate, they feel no inclination to follow.  Nor is "terrorism" (one of the other six alternatives) a pro/con issue in the same sense as an alternative like "taxes."  "Terrorism" masquerades as an issue, but it is really a way to refer to a candidate's presumed strength of character for preventing a future attack, as suggested by its being chosen mostly by Bush voters (24 percent, versus 3 percent of Kerry voters).  Other criticisms of the moral values question are also specious.

Nothing said here implies that the moral values responses in the 2004 NEP Exit Poll explain the difference between the 2004 and 2000 election outcomes.  On the contrary, much of what we have learned thus far from exit polls and other evidence suggests that the main differences between the two years had to do with Bush's incumbency, his bonding with the public immediately after 9/11, the "swift boat" and Republican convention attacks on Kerry, and the differing personalities of the two candidates.  But it does seem clear that the various denotations and connotations of "moral values" played an important role in Bush's support in both elections, a conclusion that fits not only the poll results but also much qualitative evidence provided by other observers, for example, Thomas Frank in his 2004 book, What's the Matter with Kansas?  If so, isn't it a mistake for those interested in future elections to focus on criticizing the medium rather than on understanding more fully the nature of the message?

Howard Schuman is a professor and research scientist emeritus, University of Michigan. Most of his research has been done through the university's Survey Research Center. He is the senior author of Questions & Answers in Attitude Surveys: Experiments on Question Form, Wording, and Context (1981;1996) and of many other articles on survey questions.

Readers who wish to respond to this or other articles appearing in Public Opinion Pros, or to contribute commentary of their own in 800 words or less, should consult our author submission guidelines and editorial policies under "Letters to the editor and op-ed articles."