From IQ to Biceps : British News Media and the Causes of Political Allegiance

PhD student Ben Whitham writes here about how the erroneous nature of many causal claims in the media and how they might best be overcome.

The Telegraph published an article last week, based on a research publication from Griffith University in Australia, under the headline ‘Strong men more likely to vote Conservative’. The Griffith University study of ‘actors known for their physical strength and formidability, among them Arnold Schwarzenegger, Bruce Willis, Chuck Norris and Sylvester Stallone’ apparently concludes that they are ‘more likely than not to support the Republican position on foreign policy’ and to support ‘military action’. Thus a pattern is ostensibly established between ‘upper body strength in adult males’ and a tendency to ‘take the more right wing view – be it Conservative in the UK or Republican in the US’. The claim, captured by the Telegraph headline, is that, to some extent, physical build causes political allegiance. Upper body strength is, the article contends, ‘a crucial variable’ that causally ‘impacts on a wide range of mental mechanisms’.

It would be easy to write-off such an article as a rather facile example of the bravado of (right-wing) Telegraph journalists, eager to establish their alpha male credentials on Fleet Street, or to question the reliability of a study looking at the political allegiance and physicality of Hollywood action stars (who might not be considered the best indicators of more general political behaviour!). However, the sort of causal claim it makes is fairly common in news media articles based on academic publications,[1] and, given the role of such media in shaping public perceptions on political issues, it is worth reflecting a little more on the way this argument is presented.

The article bore a striking resemblance (though in an almost ‘inverted’ form) to one published by the Daily Mail in February under the headline ‘Right-wingers are less intelligent than left wingers, says study’[2]. This piece predictably incensed MailOnline readers, and led to a biting and equally provocative article by the Guardian’s Charlie Brooker highlighting the bizarre MailOnline strategy of provoking its own readers (reproduced here for reasons of hilarity only):

As you might expect, many Mail Online readers didn’t take kindly to a report that strived to paint them as simplistic, terrified dimwits. Many leapt from the tyres they were swinging in to furrow their brows and howl in anger. Others, tragically, began tapping rudimentary responses into the comments box. Which is where the tragi-fun really began.

“Stupidest study of them all,” raged a reader called Beth. “So were the testers conservative for being so thick or were they left and using a non study to make themselves look better?” Hmmm. There’s no easy answer to that. Because it doesn’t make sense.

The idea behind both articles is to suggest to readers that there are certain pre-existing, ostensibly apolitical, qualities – ‘upper body strength’ in the Telegraph, ‘intelligence’ or ‘cognitive abilities’ in the Mail – which predispose individuals to hold particular political views and take part in particular forms of political behaviour. While there is nothing essentially problematic about claiming that such factors might contribute to shaping political behaviour, the appeal to a causal story is less convincing.

These articles rest on academic research which in turn relies on the philosophical and methodological doctrine of Humean empiricism, according to which, social explanation is only legitimate – or even possible – when it is based in the realm of empirical observation. But this approach, which predominates in academia, is flawed. It is rooted in the mistaken identification of social science with natural science and the consequent requirement that, to make a causal claim, one must witness a ‘constant conjunction of events’ and thus produce a statement along the lines of ‘where a, then b’. But, as critical realists like Roy Bhaskar have argued, unlike the ‘closed systems’ of natural science, social systems are inherently ‘open’ and thus no such ‘constant conjunctions’, which might reveal causal laws, obtain. The operation of a causal tendency in the social world cannot, therefore, be mapped in so linear a way, given that what we sense/witness (i.e. the ‘empirical’) can be ‘out of phase’ with the causal tendencies which shape it. Therefore, a more nuanced, less linear, account of causation is required if the study of politics is to produce more plausible theses than those presented in the Telegraph and Mail articles.

Milja Kurki’s Causation in International Relations: Reclaiming Causal Analysis (a much-simplified version of the central argument of the book can also be found online in her Millennium article, ‘Critical Realism and Causal Analysis in International Relations’) presents just such an approach to causation in political science. Kurki points out that, prior to the Cartesian and Humean revolutions in thinking about causation, it was once standard to use something closer to Aristotle’s ‘four causes’ as a framework for understanding social causation.

In addition to the ‘efficient’ or ‘mechanical’ type of cause recognised in contemporary empiricist/positivist political research, whereby causes are conjunctions of empirically observable events, knocking against one another like billiard balls (mainstream contemporary approaches do of course account for multiple causes/variables in the production of effects but nonetheless adhere to this efficient model), Aristotle posited three further forms of causation. These were: ‘material’ causes, the ‘constraining and enabling’ factors that materially determine whether an outcome, an event or an interaction is possible in the first place; ‘formal’ causes, the ideas and concepts which shape actions (or the ‘reasons’ for doing something); and ‘final’ causes, the situation or outcome aimed at when taking an action. Aristotle saw these types of cause as distinct but not discrete, he accounted for the complex and overlapping ways in which they can come together.

Kurki’s critical realist approach to politics and international relations is much closer to this Aristotelian account than the sort of efficient causal story underpinning the Telegraph and Mail stories. She argues particularly for recognition of the causal efficacy of the ‘formal’, in the sense of the ideas, values, norms and discourses that shape social action and the ‘material’, in the sense of differential access to resources. More broadly, Kurki and other critical realists would like political scientists to adopt a more nuanced social ontology – accepting the existence of unobservable social structures and the necessarily dialectical relationship between such structures and social practices and events – which in turn allows for a model rooted in ‘casual complexity’, whereby the line between cause and effect is not so clear.

To illustrate how such an approach might be applied, let’s return finally to the story from the Telegraph. So we have a pattern, a correlation at least, between male upper body strength and political allegiance. Why should we assume that it is being strong that causes right wing views, and not the opposite? Perhaps adhering to a right wing political doctrine entails a belief in a particular model of ‘manliness’ rooted in upper body strength, and hence inspires men to work out more.[3] Certainly the New Right ‘yuppie’ culture of the 1980s appeared to entail such an ideal of physique, regular gym use and so on. Such a belief, an ideal, would be a ‘formal’ cause, and – as we can see – would betray the fact that the causes of an individual person ‘being’, or voting, right wing cannot be as simple as the Telegraph story implies.

Furthermore, we might contend that ‘constraining and enabling’ material causes come into play on this issue too. Perhaps after “being strong makes you right wing” and the alternative “being right wing makes you go to the gym”, we could suggest that being born into a materially wealthier or ‘rich’ family (which already predisposes a person to ‘be’/vote conservative) might mean a healthier diet, less deprivation, more exercise (private schools focus more strongly on sports, and the rich might have more free time on their hands) and so on.

The point here is not to provide a serious or credible account of what ‘causes’ people to ‘be’ right wing, but rather to highlight the fact that better, and more plausible, causal stories might be told than those that tend to predominate in mainstream social science, and, therefore, in the news media. Such stories must necessarily account for the fact that – especially in the case of a ‘mental state’ such as ‘being right wing’ – is neither solely efficient nor uni-directional, but rather consists of multiple dialectically related agents, events, processes and structures, which are not always empirically observable.

If, as I was, political science academics are frustrated by articles which portray causal links in too simplistic a manner, perhaps, instead of lamenting the quality of contemporary journalism, they might reflect on the limitations of their own metatheoretical assumptions and methodological choices.


[1] It should be noted, in the interests of fairness to academic researchers, that news outlets may distort, reinterpret and simplify such publications for editorial, and of course political, reasons.

[2] Thanks to an anonymous colleague for bringing both articles to my attention!

[3] Thanks to another anonymous colleague for pointing this out!

  1. Alan Renwick’s avatar

    This is very interesting, Ben: thanks for writing it. I agree that journalists and politicians often make unjustified causal claims on the basis of mere correlations. One particular bugbear of mine is the claim that the fact that married couples tend to stick together for longer than unmarried couples shows that marriage has positive effects. Rarely have I heard any journalist interviewing a politician who makes this claim point out that maybe couples who are more likely to stick together are more likely to marry.

    But I do have a big problem with your claim that mainstream social science is equally at fault. That’s not true at all. Positivist social scientists understand perfectly well that a correlation between A and B does not in itself allow us to conclude that A causes B. B might cause A. C might cause both A and B. That’s why mainstream social scientists are always talking about possible spurious correlations or about endogeneity problems or about the direction of causation: these issues are absolutely fundamental to everything they do.

    So, while there may be good reasons for adopting a critical perspective, this isn’t one of them.

    Reply