Choices of books about science reflect political polarization

Can lifestyle choices correlate with left-to-right political alignment? And if any such correlations involve science, can anything be learned in a time when science and political polarization overlap troublingly? Yes, says a 3 April study in Nature Human Behaviour. Yes, say media reports about it.

But what’s being learned makes science-related polarization seem all the more troubling.

A simple, intriguing online display from Cornell University places the study into a wider scholarly context. In each of nearly 50 selectable categories, the display depicts correlations between lifestyle choices and political alignment. Wired quoted Cornell computational social scientist Michael Macy, who sees in the data “a strong correlation between ideology and cultural preferences that have seemingly nothing to do with ideology.” To get the data, the Cornell researchers methodically tracked Twitter activity, probing what they call the “cultural polarization of the American electorate.” Click “Sports” on their display, and you’ll see golf brands clustered on the conservative right, with hiking brands furthest left. “Dining”? Chick-fil-A appears furthest right. “Retail”? L. L. Bean appears furthest left.

Since the categories don’t include “Sciences” or “Science books,” you can’t find out whether, for example, left-leaning book purchasers might prefer basic physical-science topics while right-leaning buyers might choose applied or commerce-related science. But now the Cornell researchers, together with colleagues from Yale and the University of Chicago—tracking not Twitter but publicly available book-purchase data—have published the 3 April study . Their findings suggest such a correlation, and much else.

Here’s the abstract for their paper, titled “Millions of online book co-purchases reveal partisan differences in the consumption of science":

Early on, the paper questions the assumption that even in a highly polarized electorate, science can still bridge the partisan divide. If you ask for responses to survey questions, it proposes, you might simply hear an echo of “Enlightenment commitment to value-free scientific inquiry that masks underlying scepticism about science.” This can be seen, the paper asserts, “not only in conservative resistance to climate change, but in historically liberal resistance to consensus over the positive benefits of genetically modified organisms, vaccination, nuclear power and the safe storage of nuclear waste.” The researchers ask: Has “selective attention … reinforced the ‘Big Sort’ of American politics—the tendency to cluster in like-minded communities?”

Then the paper explains their research method for probing that question. Scientific American characterizes the method as “follow the money.” Rather than seek survey responses, the researchers used publicly available data from Amazon and Barnes & Noble—the “co-purchase” information about other books bought by people who also bought a given book. For both Amazon and Barnes & Noble, they began with what they call “seed books,” politics-related ones to set political self-definitions. The researchers then built up and analyzed networks of co-purchase connections, focusing on “the consumers of science, using online co-purchases of books on science and politics as a behavioural indication of preferences.” They probed the extent to which “purchasers of political books are also interested in science, and in what parts of science they are most interested.”

The researchers’ sampling, they report, extended to coverage of “hundreds of millions of online customers.” They “collected data recursively by tracing co-purchase links, iterating the search until no new titles could be identified.” They report that in total from Amazon, they “collected 26,467,385 co-purchase links among 1,303,504 books … after consolidating multiple editions.”

The researchers conjectured that a shared interest in science might help bridge political divisions, but that “selective attention to ‘convenient truths’ risks reinforcement of existing political identities.” They found that a few disciplines, “notably palaeontology,” do bridge political divisions, attracting equal interest from left and right. But most “purple disciplines (with equal likelihood to have red and blue book links),” they found, don’t bridge the divisions. “Simply put,” the researchers wrote, “even when left and right are equally likely to read books in a discipline, they are rarely the same books or even from the same topical cluster.”

After discussing “inherent limitations” to the power of their method, they summarized in part by reporting findings that

• “online co-purchases reveal a similar level of interest in science among liberal and conservative readers, but not in the same topics,” that
• there’s “little support for the view that science is either an apolitical separate sphere, largely ignored by partisans, or a public sphere in which left and right share common scientific interests,” and that
• “scientific preferences are polarized at the aggregate level, with liberals attracted to basic science, and conservatives attracted to applied, commercial science.”

They also suggest that although

• their “analysis cannot uncover the cognitive and social processes underlying these partisan differences in the consumption of science, it is consistent with research on ideologically motivated reasoning,” and that although
• science “may not be on the front lines of the culture wars … it is not above the battle, nor is it immune to the ‘echo chambers’ that have been widely observed in political discourse.”

The echo-chamber implications were emphasized in a commentary in the edition of the Nature journal containing the study, and later in an editorial at Nature itself. Wired, under a headline citing “myopic political bubbles,” concluded that while “it’s great that the reasoned examination of facts appeals to everyone, the study seems to suggest that—unsurprisingly, but depressing nonetheless—people seek out the stuff that supports their worldview.” This subhead appeared with Guardian science editor Ian Sample’s discussion of the study: “Can an interest in science unite a divided society? No, concludes research based on reading habits of those from right and left of the political spectrum.”

But whether or not Pollyannaishly, the researchers and others , including an expert consulted by the Los Angeles Times, envision remediation. Concerning the echo-chamber problem, the researchers wrote:

Yes, it is hoped.

Steven T. Corneliussen is Physics Today‘s media analyst. He has published op-eds in the Washington Post and other newspapers, has written for NASA’s history program, and was a science writer at a particle-accelerator laboratory.