There are two kinds of people, conservatives and liberals. This is a handy way of sorting people into social categories that are easily understood and widely accepted. As social categories, conservatism and liberalism claim awesome inductive power. Having placed Tom in one category and Dick in the other, Harry is ready to expect them to disagree on virtually everything (Tajfel & Wilkes, 1963). Psychologists, most of whom abhor stereotyping – but study it for a living, are happy with these categories. They yield all sorts of results that vindicate their own existence. When conservatives are found to differ significantly from liberals on some other, hitherto ignored, behavior, voilà, the belief in the reality and relevance of these categories becomes stronger.
Psychological Science, the flagship journal of the Association for Psychological Science, professes to publish the proudest accomplishments in contemporary research: advances, breakthroughs, and discoveries. Research that is tedious, trivial, or tendentious should be published elsewhere, or not at all – or so you’d think.
The journal’s latest offering – available online – is a paper by R. Khan and colleagues on yet another way conservatives differ from liberals. Their research suggests that conservatives have “a systematic preference for established brands” (p. 1). We immediately feel the pleasure that comes with recognition. This assertion must be true because that’s what conservatism means, “a preference for tradition and the status quo” (p. 1).
Surely, research showing that conservatives are more adventurous, novelty-seeking, and nonconformist – along with a theory to explain how and why this is so – would not have been tedious. But this? The authors seem aware of this objection, and in a maneuver that is almost elegant, they morph the triviality of the behavior (buying Kleenex instead of the no-name product) into a virtue. The power of life-guiding conservatism must be awesome indeed if it does not stop at the inconsequential. Of course, the anti-counterargument is that if political ideology radiates indiscriminately like a lump of plutonium in Kazakhstan, it will have many trivial side-effects. To find one of them adds to science – infinitesimally.
Scientific research and the reports that communicate the results are embedded in a social environment awash with preferences, goals, and agendas. This I accept as a fact of life. Researchers, individually and as a community, bring their own prisms to refract the light. Social psychology in general and the Khan lab in particular are no exception. In their hands, political conservatism becomes subtly pathologized. It is conservatives and their behavior they seek to understand. The stem “conservati-“ appears 20 times in the text compared with 7 mentions of “liberal.” This reveals the frame. “Compared with liberals, individuals who gravitate toward conservative ideology tend to score lower on measures of integrative complexity, openness to new experiences, and tolerance for uncertainty and ambiguity, and to score higher on measures of conscientiousness, dogmatism, and need for order, structure, and closure” (p. 1). This phrase could as easily have been framed as the task to understand liberals. Also notice that most of the traits attributed to conservatives are undesirable, at least from a psychologist’s point of view. Taken by itself, the trait of conscientiousness could be a redeeming quality, but in the context of low complexity, dogmatism, and a need for order, a conscientious person looks like a fascist in waiting.
If we could only find a joint scale for researchers’ ideological commitments and the relative contribution of their work! I would then propose to raise a red flag when the score for the former is higher than the score for the latter.
I could stop here as I am sure you catch my drift. But before I do, let me humor you with comments on method. Khan and colleagues study counties, not individuals. The data are aggregate rates of conservative activity (voting Republican, religiousness) and aggregate rates of buying familiar brand-name products. This is a bit of a surprise because this is Psychological and not Sociological Science, and because the authors start with a build-up of individual psychology. They talk of “psychological needs, cognitive styles, and personality traits” (p. 1), “self-expression,” and “implicit cognition” (p. 2). If the data cannot speak to these issues, the authors shouldn’t either.
The correlations between conservatism-liberalism and preferences for familiar vs. unfamiliar projects can only be interpreted down to the level of the unit of observation (here: county). Since they are not, another red flag goes up. View research with skepticism if the story is about elements smaller than the smallest unit of data analysis. The inverse difference is not as bad. It is better to infer the properties of an anthill from the behavior of individual ants than to infer the behavior of individual ants from the properties of the anthill.
At the level of numbers, Simpson’s paradox looms. Simpson (1951) showed that when two variables are found to be related across categories, their relationship can have the opposite sign within category. In the Khan case, it is possible that within counties, it is the liberals who cleave to the familiar brand. To suggest that they do not is to follow the dictate of a good story, one that liberal psychologists want to hear. But do we not need to listen to the data before throwing rocks? Khan and colleagues advise us not to worry. After noting, that “ideally, an examination of the relationship between political ideology and brand consumption would be based on data from a consumer panel with accurate measures of ideology and pur¬chase behavior for a greater variety of products,” they cheerfully proclaim that “the consistency of our results across a large set of product cat¬egories suggests that aspects of ideology may indeed be reflected in daily behavior at an unconscious level or in an implicit manner” (p. 7). Non sequitur: it does not follow.
With a possible delusion of hindsight, I seem to remember a time when political attitudes were just that: attitudes. During that mythical era it was fun to debate with individuals who saw things differently. Perhaps you could move them on one point or another, or lo!, one of your own opinions might change.
In the current era, of which I see the Khan research emblematic, political attitudes are reified as matters of character, matters that run so deep that their ripple effects are felt over a great distance. The conservative mind rumbles deep underground, and way up, at the surface, the consumer’s hand reaches for the Hershey’s bar instead of the Ritter Sport (which, trust me, is far better). Try the Alpenmilch some time. It might make you more liberal.
See here for more.
Khan, R., Misra, K., & Singh, V. (2013). Ideology and brand consumption. Psychological Science, published online 4 February. doi: 10.1177/0956797612457379
Simpson, E. H. (1951). The interpretation of interaction in contingency tables. Journal of the Royal Statistical Society, Ser. B, 13, 238–241.
Tajfel, H., & Wilkes, A. L. (1963). Classification and quantitative judgment. British Journal of Psychology, 54, 101-114.
Be sure to read the following responses to this post by our bloggers: