(covering polls and related articles from the week of January 16-January 22, 2006)
In this edition of Public Opinion Watch:
• Values to the Left of Me, Values to the Right of Me and Nary a Strategy in Sight
• Party Identification Shifts Toward Democrats
Values to the Left of Me, Values to the Right of Me and Nary a Strategy in Sight
Sometimes I think there are more values than voters out there. At least, one might be forgiven this thought, given all the head-scratching about values taking place in progressive circles and the many, many nominees for the values progressives should be stressing – right now! – in their efforts to build a majority coalition.
I’ve always felt quite ambivalent about this values obsession. On the one hand, I can only applaud the general concept that values should be taken seriously as the prism through which voters view policies and politics. Just thinking about issues and how well different issues poll is certainly an inadequate way to formulate political strategy.
On the other hand, discussions about values tend to become awfully squishy awfully fast. Instead of the suspect assertion that simply talking about the right issue(s) will take progressives from Hell to hallelujah, values-talk tends toward the equally suspect assumption that simply talking about the right value(s) – linkage to actual, feasible politics unspecified – will lead progressives to the promised land. Well, I don’t believe that and neither should you.
Let me illustrate my concerns by discussing one recent offering in this ongoing values discussion, Garance Franke-Ruta’s article, “Remapping the Culture Debate,” in the latest issue of The American Prospect. Franke-Ruta’s article starts by discussing the values work of Ted Nordhaus and Michael Shellenberger, well-known in progressive circles for their essay, “The Death of Environmentalism.” Nordhaus and Shellenberger, now principals in American Environics, the American branch of the Canadian market research firm, Environics Research, have been pushing a values scheme based entirely on their analysis of an Environics in-home consumer survey that has been conducted since 1992.
Their presentations of their work to various progressive organizations and politicians have met with a generally favorable reception and Franke-Ruta’s views on their work are no exception. She portrays their analysis as path-breaking empirical work that will (or at least should) completely recast the way progressives look at politics.
I am not so sure. Begin with the fact that their data are drawn from only one survey series – their own – and no attempt has been made so far to compare their findings to those from other series. This does not inspire confidence. Take, for example, two of the few actual data points that are mentioned in Franke-Ruta’s article:
“Between 1992 and 2004, for example, the percentage of people who said they agree that ‘the father of the family must be the master in his own house’ increased ten points, from 42 to 52 percent, in the 2,500-person Environics survey. The percentage agreeing that ‘men are naturally superior to women’ increased from 30 percent to 40 percent.”
Could be, but check out this finding from the premier American academic political science survey, the National Election Study (NES). The NES asked respondents to place themselves on a 7 point scale relative to the following statements: “Some people feel that women should have an equal role with men in running business, industry and government. Others feel that women's place is in the home,” where 1 is the strongest support for the women’s equal role and 7 is the strongest support for women’s place being in the home. In 1992, 51 percent selected 1, the strongest support for women’s equal role; in 2004, 57 percent selected 1. So support for women’s equal role appears to be strengthening in the NES. Indeed, in the 2004 survey, a total of 78 percent of respondents picked 1, 2 or 3 on the 7 point scale, indicating they felt closer to the equal role statement that to the women’s place in the home statement.
But at the same time we’re supposed to believe that 40 percent now believe men are superior to women and that 52 percent believe the father should be the master of the house – increases of ten points in each case over the same period covered by the NES data. I guess we could reconcile the data from the two surveys by positing a trend toward believing women are equal but dumb and subservient. But pardon me if I’m a little skeptical – a skepticism that’s reinforced by trend data from the General Social Survey, the premier academic sociology survey, showing fewer, not more, people believing that women should take care of home, not the country, and fewer, not more, people believing that men are better suited for politics than women.
This illustrates the perils of relying on one survey for one’s data about Americans’ values – or anything else for that matter. Especially when that one survey is a consumer market research survey designed not for political research, but for very different purposes.
So why are observers like Franke-Ruta and others so captivated by the Nordhaus/Shellenberger analysis, when it relies on only one data source – a data source, moreover, whose superiority over other sources is simply an assertion lacking any supporting evidence? Several reasons:
1. The very fact that it is a consumer marketing survey actually adds to the survey’s cachet. We now realize values are important, the thinking runs, and who’s been paying attention to values all these years? Why, corporations and market researchers, of course, so they (or their data) might already have the answers we’re so frantically looking for.
2. Since the Environics survey tracks over 100 different values, there’s a ton of value trends to look at and everyone can find at least one trend (or several) that confirms their suspicions, based on pop culture/reading/hunches/whatever, about where the country is really going. In effect, the Nordhaus/Shellenberger presentation of these data functions as a sort of Rorschach test for progressives interested in values, where people see in the data what they wanted to believe to begin with.
This is especially the case since they cluster- and factor-analyze their data to death, showing in various “maps” how all these values relate to underlying value dimensions (survival vs. fulfillment; authority vs. individuality) both overall and for a multiplicity of different values-defined “constituencies of opportunity” for progressives. The result is many complex grids – some of them for groups whose sample size cannot be more than 25 or so in their data – with dozens of multicolored values sprinkled in different patterns on each grid.
Well, if you can’t find something you agree with or find significant with this much to choose from, you’re just not looking hard enough! And my sense is people do just that, hence the recent popularity of their analysis.
But the question must be asked: what, exactly, are we getting out of this analysis that we couldn’t get elsewhere? Here’s an example from Franke-Ruta’s article:
“By focusing on ‘bridge values,’ [Nordhaus and Shellenberger] say, progressives can reach out to constituents of opportunity who share certain fundamental beliefs, even if the targeted parties don’t necessarily share progressives’ every last goal.”
In other words, when reaching out to swing voters, emphasize beliefs swing voters and your base have in common, rather than beliefs they don’t. Now, if one were going to choose to talk about values instead of, or in addition to, issues in one’s political work, I can’t imagine one would take any other approach. So I’m not sure we need a zillion color-coded values maps to make that case.
But perhaps Franke-Ruta was entranced by the specific bridge values Nordhaus and Shellenberger advocate using? I can find no evidence of this in her article, comparing its content to either this public Nordhaus/Shellenberger document or what I generally know of their work. Indeed, the best insight of the article – which I really do recommend, despite her excessive enthusiasm for Nordhaus/Shellenberger – is this, which bears little, if any, relation that I can see to their analysis:
“The growing conflation of the economic and the cultural in the minds of voters has been a cause of great perplexity for thinkers who have long seen the two realms as distinct, and the cultural realm as the secondary concern of unserious men who don’t know where their self-interest lies. Thomas Frank, in his 2005 What’s the Matter with Kansas?, sketched a portrait of lower- and middle-income voters who, socially at odds with a liberal elite they accuse of moral dissipation, have forged an alliance with a conservative fiscal elite whose economic policies, paradoxically, do little to support their worldview or shore up families. Yet the broader social reality suggests that the focus of these middle-income voters on cultural traditionalism is not entirely separate from their economic aspirations. Social solidarity and even simple familial stability have become part of the package of private privileges available to the well-to-do. Behavioral surveys consistently show that, regardless of their political leanings, the better-off and better-educated live more traditional personal lives: They are more likely to marry, far less likely to divorce, less likely to have children outside of marriage, and more likely to remarry when they do divorce than their less accomplished peers. In addition, their kids are more likely to be academically successful and go to college, repeating the cycle.
The new Puritanism and cultural conservatism Frank described can also been seen as symptoms of how, in today’s society, traditional values have become aspirational. Lower-income individuals simply live in a much more disrupted society, with higher divorce rates, more single moms, more abortions, and more interpersonal and interfamily strife, than do the middle- and upper-middle class people they want to be like. It should come as no surprise that the politics of reaction is strongest where there is most to react to. People in states like Massachusetts, for example, which has very high per capita incomes and the lowest divorce rate in the country, are relatively unconcerned about gay marriage, while those in Southern states with much higher poverty, divorce, and single-parenthood rates feel the family to be threatened because family life is, in fact, much less stable in their communities. In such environments, where there are few paths to social solidarity and a great deal of social disruption, the church frequently steps into the breach, further exacerbating the fight.
American voters have taken shelter under the various wings of conservative traditionalism because there has been no one on the Democratic side in recent years to defend traditional, sensible middle-class values against the onslaught of the new nihilistic, macho, libertarian lawlessness unleashed by an economy that pits every man against his fellows. Yet in private conversations, progressives recognize that there is a need to do something about broad social changes that they, too, find objectionable….”
I think this is remarkably astute and potentially points progressives in a very fruitful direction. And if it took the Nordhaus/Shellenberger Rorschach test to get Franke-Ruta thinking along these lines, that is certainly a point in their favor. However, if we wish to be really guided by values research in formulating political strategy, we will have to go beyond the Rorschach test stage and engage critically with the widest possible range of data. If values are truly important – and I think they are – we just can’t afford any other approach.
Party Identification Shifts Toward Democrats
I’ve expressed a lot of skepticism in the past that short-run swings – sometimes quite large – in party identification (ID) between polls are necessarily indicators of change in underlying public sentiment. As I, Alan Abramowitz and others have argued, there are good reasons to suspect sampling bias in some of these swings and a good case to be made for “dynamic party ID weighting” to smooth out these swings and give a more reasonable sense of underlying political change (see this article by Abramowitz for a clear statement of our position).
But changes in party ID measured by survey data aggregated over fairly lengthy time periods are a different matter. Here one can be fairly certain that the measured changes in party ID reflect meaningful changes in underlying political attitudes. In this spirit, one can peruse with interest a recent Gallup report on party ID shifts in 2005. The news bodes well for Democrats in 2006 and poorly for the incumbent GOP. Here are the basic findings:
1. In the Gallup data, the Democrats had a 4.5 point lead in party ID (including leaners) in 2005. That’s up from a 2.7 point lead in 2004 and .1 and .4 point deficits, respectively, in 2003 and 2002. So the trend lines are pretty clear.
2. The report also breaks down the data by state, showing that the Democrats now lead in party ID in 29 states. That’s up from 22 states in 2004 and just 17 and 16, respectively, in 2003 and 2002.
3. The states also indicate that the Democrats now have a party ID lead in virtually every state that is generally put in the contested column in American politics: Oregon (15); New Hampshire (14); West Virginia (13); Nevada (12); Michigan (12); Maine (12); Washington (12); Arkansas (11); Minnesota (11); Missouri (8); New Mexico (8); Ohio (7); Iowa (6); Louisiana (6); Wisconsin (4); Pennsylvania (4); Colorado (3); and Florida (1).
As further food for thought, here are the same states, with Bush’s current approval rating in those states, as measured by the latest SurveyUSA polls: Oregon (38); New Hampshire (44); West Virginia (46); Nevada (39); Michigan (37); Maine (37); Washington (35); Arkansas (38); Minnesota (39); Missouri (41); New Mexico (42); Ohio (38); Iowa (43); Louisiana (48); Wisconsin (44); Pennsylvania (39); Colorado (43); and Florida (42). Hmmm…
Ruy Teixeira is a joint fellow at the Center for American Progress and The Century Foundation.