Confession: I love reading soft and even pseudo-science. I read it like Greek tragedy where a hero (the mind) untethers from his or her principles and descends into madness. Often there are brilliant insights even in that madness.

Confession two: I read this article because it sounded fun. It wouldn’t surprise me if political differences were expressions of much deeper psychological traits that might even show up in consumer choices like beer. I think there is the seed of insight here, but early on I realized the article wouldn’t be as much fun as I hoped.

Vishal Singh has published an article in the Journal of Psychological Science. He teaches in the New York School of Business which should be our first clue that this study may not be the hardest of science. But my beef isn’t with him or his study. I am responding to an article about the study on

My suspicion that the author of the article had a bias against conservatives and was cherry picking science to prove those assumptions occurred early on. The article mentions “an ever growing body of research” which is always a very dangerous phrase. That phrase can tempt us to base our conclusions not only on the tenuous evidence the author presents, but also on an imagined landslide of future research that will fill in the gaps. The author did not go far down that route, but such sloppy terms were often the backbone of his argument.

“The study hypothesis was simple: In counties characterized by strong Republican voting or religiosity, generics and new products would fare considerably worse at the checkout counter than established brands. Just picture a liberal and a conservative at the laundry detergent aisle. The basic idea is that the conservative more often reaches for the pricier but more established Tide (a Procter & Gamble product), rather than the cheaper in-store generic variant (more often favored, presumably, by the liberal).”

So a “scientific” study lumped together those who vote Republican and those who are religious into one category called “conservative.” That group would then would provide the sample for this test. But how does one scientifically measure a trait like “religious?” Wouldn’t that depend on what behaviors the researcher thought to be “religious?” And how is it in any way scientific to assume a person identified as “religious” can then be casually thrown in the same bucket with those who voted Republican?

 “Conservatives, after all, are known to be more uncomfortable with uncertainty, and less open to new experiences. There’s every reason to expect this to translate into consumer behavior — particularly with respect to one’s allegiance to brands. “A major function of branding,” write the study authors, “is to reduce uncertainty and simplify decision-making.” Their study design therefore sought to test whether “aspects of conservative values — such as preference for tradition and convention, and dislike of ambiguity and complexity,” would be “reflected in higher reliance on national brands as opposed to generics.””

Wait a minute. It may prove to be true that conservatives are more uncomfortable with uncertainty, but is it really something we “know” to begin with? Doesn’t this method of using stereotypes to group people run the danger of hiding your conclusion in the yardstick you will later use to measure your sample?

“So what did the study find? Across 26 different product categories — coffees, diapers, household cleaners, mayonnaise, and much else — the study found that in Republican or religious counties, generic and new brands captured considerably less market share. “More conservative counties did not have higher penetration of new products in any of the categories,” write the authors. As for generics: In 19 out of 26 product categories, they fared worse in more religious counties. (The study controlled for demographic factors, as well as store size and variations in the quality of generic products.)”

If I fit anywhere on the political spectrum it would be way to the left of liberal, but I can’t help but be concerned with the way conservatives are portrayed in this article. That the author begins with a bias is suggested by the fact that he has already written one book called “The Republican War on Science,” and a second book coming out in April, “The Republican Brain: The science of why they deny science- and reality.” I might agree with much of what he says, but to present his biased opinion as science is a disservice to the conservatives he demeans and to the science he misuses to do it.

I suspect there is an insight buried in all this bad science. It may be true that conservatives are more likely to drink brand beers and liberals to drink more exotic brands. I suspect it may be true that being conservative or liberal may say more about our temperaments than our mere political positions. That possibility is important to explore, but the approach used in this article will hide more than it reveals. Science can answer our questions of truth but it is a deadly when we try to scientifically measure human values. It is one thing to say conservative and liberal are different, but once we say one value is better than another, we have left the domain of legitimate science and entered the smokey and dangerous illusions of pseudo-science.