Emerging as a potentially important factor that might help to explain cross-cultural differences are the gender relations in the society from which participants are drawn. Thus, Binglin Gong and Chun-Lei Yang compared risk taking on a lottery task in the matrilineal Mosuo (in which the head of the household is also traditionally female) and the patriarchal Yi.26 Although in both societies the females bet less than the men, the gap was considerably smaller in the matrilineal Mosuo. Similarly, University of Los Andes economist Juan-Camilo Cárdenas and colleagues found that the gender gap in financial risk taking was smaller in Swedish children than in children from Columbia, a country that ranks much lower than Sweden on various macroeconomic indices of gender equality.27 There’s even some evidence that single-sex environments may encourage greater risk taking in British girls and young women.28
The problems continue when you turn from lotteries to other kinds of financial risk tasks.29 By now, you may have the impression that economists approach the topic of risk taking as though they once overheard a prestigious colleague comment that “Life’s a lottery,” and took it literally. But as we all know, the majority of financial decisions do not closely resemble economists’ lottery tasks. Warren Buffett did not make his billions contemplating whether to pursue Option A with a guaranteed return of $2, or Option B with a 30 per cent chance of a $4 return. Nor do bosses fix employees with a gimlet eye and, slamming a coin onto the table with a provocative cry of “Heads or tails, Professor Massoud?” invite them to take a 50/50 gamble on a $15 per annum raise, or stick with the guaranteed $5 rise on offer.
One obvious difference is that people typically don’t know the precise odds of the different hands they can be dealt by fate. This is also the case for two risk tasks popular with psychologists. In one (the Balloon Analogue Risk Task), participants decide how many times to “inflate” a virtual balloon with a pump, with 5 cents earned for every successful pump. At an unknown point, however, the balloon will explode and all the money is lost. A meta-analysis found that males are modestly more risk taking than females on average in this task.30 However, precisely the opposite is the case for the second popular risk-taking task, the Iowa Gambling Task. Individuals choose between decks of cards that are either high risk (high rewards but also higher losses) and less advantageous in the long run, or lower-risk (low rewards and losses).31 Although over time most people shift to the lower-risk packs, women are a bit more likely than men to continue to try their luck with the high-risk packs.
Another obvious and particularly important difference between laboratory tasks and real-world financial risks is the sums of money at stake. In the Balloon Task, financial rewards are scraping around the dollar mark, while in the Iowa Gambling Task the “dollars” lost and won are purely hypothetical. Largesse also tends to be absent in studies run by economists: pay-offs are usually small, hypothetical, or restricted to one particular gamble chosen at random. It’s therefore noteworthy that in one of the few lottery experiments that compared risk preferences for trivial versus substantial pay-offs, the sex differences seen in the typical “small-change” version of the task disappeared when nontrivial sums of money were at stake.32 An interesting perspective on why this might be was offered by anthropologist Henrich and his colleagues. They suggest that
when actual economic stakes are 0 (hypothetical), all kinds of other concerns come to predominate in the decision process. Informants may be concerned with what the ethnographer will think of them or what other people will infer about them from their decisions.
In their own cross-cultural research, Henrich and colleagues therefore use large stakes “to focus the informants’ attention on the game payoffs rather than on exogenous social concerns.”33 As you’ll recall, sex didn’t predict financial risk taking when this research method was used with the Mapuche, Sangu, and Huinca communities of Chile and Tanzania. Their research protocol is also in perfect keeping with Cass Sunstein’s argument (first met in Chapter 5) that the consequences of a decision for one’s self-concept and reputation are vital ingredients in the recipe from which preferences emerge.
This aspect of the decision-making context is something that economists, in particular, have not been especially interested in. It was only at the turn of the twenty-first century, in a groundbreaking economics article written by Nobel Prize–winning economist George Akerlof and fellow economist Rachel Kranton, that the concept that social identity and norms have a motivating effect on behaviour was formally introduced to economists.34 “What people care about, and how much they care about it, depends in part on their identity,”35 they observe. These “identities and norms derive from the social setting.… [S]ocial context matters.”36
To a social psychologist, this is an almost comically belated revelation: a little bit as if only recently a landmark social psychology article introduced colleagues to the concept of money, and its remarkable influence on people’s preferences and behaviour. But better late than never, and while this area of research is still in a preliminary state, the identities and norms at play in a particular context do seem to influence financial risk taking. For instance, when negative stereotypes about female mathematical ability are made salient (this can be done both by drawing attention to the person’s female identity, as well as by highlighting the “masculine” nature of the task), it can impair girls’ and women’s interest and performance in mathematics—a phenomenon known as “stereotype threat.”37 In one study, women were more risk averse than men when they were required to record their sex before participating in a gambling task that was described as a test of mathematical, logical, and rational reasoning abilities. However, when exactly the same task was instead described as “puzzle solving” (and participants didn’t record their sex beforehand), women were just as risk taking as the men.38
While this study manipulated the relevance of gender identity by framing the task as one of cool, rational calculation, risk taking itself is of course a key stereotypical trait of masculinity. According to popular imagination, for instance, the successful entrepreneur doesn’t just have the necessary skills, resources, and business connections; he is also a masculine hero who laughs boldly in the face of financial risk. Perhaps this is part of the reason pitches made by male entrepreneurs are evaluated more positively than those given by female entrepreneurs—even when the content is identical.39 Women can also be put off by this portrait, Binghampton University academic Vishal Gupta and colleagues have found. For example, when Turkish MBA students were shown either a (fictional) general news piece about entrepreneurship, or one describing entrepreneurs in a stereotypically masculine way (for example, as aggressive, risk taking, and autonomous), the male students later evaluated a potential business opportunity more positively than the female students, on average. But when, in two further conditions of the same study, that business opportunity was preceded instead by either a gender-neutral (creative, well-informed) or feminine (caring, making relationships) description of entrepreneurs, women were as likely, or even more likely, respectively, to see a business opportunity in the complex business case they then analysed.40 Gupta and colleagues have also found that, across three countries, both women and men who reported having more “masculine” traits showed greater entrepreneurial intentions.41 Interestingly, two other studies similarly found that men and women who report being more “masculine” in personality also score higher on measures of financial risk taking. Unsurprisingly, men report a greater number of masculine traits as characteristic of themselves, and this explains some of the gap in risk taking between men and women.42 But as both research groups point out, while being biologically male or female is fixed, how masculine men and women see themselves is not. In fact, in the United States, the “masculinity gap” has been closing over time, in step with women’s changing roles and status in society.43
If risk taking is an integral part of a masculine identity, then we can predict that men should take greater financial risks when that identity, or the norms associated wit
h it, are made salient. Viennese academics Katja Meier-Pesti and Elfriede Penz found exactly that. They primed young women and men with either masculine, feminine, or (in a control condition) gender-neutral stimuli. Men primed with masculinity gave the most risk-tolerant responses on a questionnaire assessing attitudes towards risk taking in investments.44 A more recent study also explored the importance of masculine identity for financial risk taking, by exploiting a rather depressing phenomenon known as the “failure-as-an-asset” effect. It turns out that presenting men with evidence that they have done poorly at something at which women tend to excel provides a little boost to their self-esteem, because incompetence in low-status femininity helps establish high-status manliness. Remarkably, failure in feminine domains is also perceived as an asset by onlookers. Fictitious male job applicants who reveal weakness in a “feminine” domain (like dancing, or in a form of intelligence in which women supposedly excel) are seen as more masculine, and thus more likely to succeed in high-level roles, compared with men who “lack” incompetency in femininity.45 Expanding on this phenomenon, University of Kassel psychologist Marc-André Reinhard and colleagues found that giving men failure-as-an-asset feedback increased their self-reported interest in risky activities, as well as the amount they were prepared to invest in a gamble. This shift seemed to be brought about by greater identification with being male.46 Interestingly, the investments of men who were told either that they’d done poorly on a masculine test, or well on a feminine one, made investments that were no riskier than those of women.
In apparent contradiction, research conducted at the University of South Florida found that young men take greater financial risks after a threat to their masculinity. (The psychological castration was achieved by asking a group of the men to try a florally fragranced hand cream.)47 However, the contrast with Reinhard and colleagues’ findings may lie in the private versus public nature of the risk taking. Masculinity threats only had an effect on financial decisions made publicly, suggesting that costly displays of masculinity in response to a threat of manhood are only worth it if they serve a face-saving function.48
Although we have to be careful that findings like these are robust and replicable, they have an important implication, as Nelson points out:
Differences that may appear at a cursory level to be due to “essential” differences between the sexes may in fact be due (in part or completely) to some additional, confounding variable, such as societal pressures to conform to gender expectations or locations in a social hierarchy of power, or may no longer be seen when the sampling universe is broadened.49
Yet researchers may nonetheless treat results as though they reflect categorical, Mars versus Venus differences, Nelson goes on to point out. For example, a four-country comparison of the risk preferences of female and male asset managers revealed only marginal, scattered, and unsystematic differences across the four countries. Even for the most sizable difference in risk preference (which was seen in Italy), the possibility of creating a perfect match by pairing a female client with a female fund manager was only 38 per cent, compared with a 25 per cent chance of a successful match if the customer employed a male manager instead. Nonetheless, the study authors suggest that “female fund managers may be better suited to female customers.”50 As Nelson wryly notes, given that economists assume that financial risk-taking preference can be readily assessed with a few simple questions, why not simply ask clients what they want? It’s a bit like a restaurant manager learning that women are slightly more likely than men to order fish than steak, then telling waiters to use a customer’s sex as a guide to what meal to place in front of them.
Why might researchers make this kind of conceptual slide from small average differences to fundamental difference? Is it because, implicitly, they join so many others in assuming that the sexes just are essentially different? As Nelson explains:
The attribution of (on average) different psycho-social behaviors to (fundamental) sex differences in hormones and/or brain structure, further explained as caused by differences in evolutionary pressures on bodies with different reproductive roles, can currently be found in many studies.51
Sound familiar, at all? This assumption, in turn, makes it easier to neglect important features of the data: the vanishingly small size of sex differences in financial risk taking, and the dependence of those differences on who is being tested, the kind of task, and the social context. As we’ve already learned, these details matter a lot for the kind of explanations we reach for. If we say that “men are financial risk takers, and women are financially risk averse,” then men’s higher testosterone exposure looks like a plausible cause of that difference. But to echo questions asked earlier, how does the substantial sex difference in testosterone translate into such modest behavioural differences? How does T make men more risk taking when a gamble is framed in an abstract way, but not in a concrete salary context? Or when stakes are trivial, but not necessarily when they’re substantial enough for a loss to sting? How do sex differences in testosterone make young North American men more risk taking than their female counterparts, but not the men of China, Mapuche, Sangu, or Huinca? How does T make men inflate more risky balloons, but select fewer risky cards?
These are questions to bear in mind as the research seeking to link T with financial risk taking proceeds apace. One growing line of investigation tries to find links between financial risk taking and a measure known as digit ratio. Digit ratio is the relative length of the second to the fourth finger and on average men’s digit ratio is smaller than women’s.52 Digit ratio is often popular with researchers because it’s so easy to measure, and supposedly reflects prenatal exposure to testosterone, although whether or not there’s adequate evidence for this is contentious. (One set of researchers, for instance, describe digit ratio as “a putative, not yet sufficiently validated marker of prenatal testosterone.”)53 But even if it’s a reasonable measure for comparing prenatal testosterone between groups, digit ratio “may be much less useful” as an index for individuals within a group, as Herbert explains.54 (It’s just too “noisy” a measure: a bit like using a person’s height as a proxy for their early nutrition, on the grounds that people who were fed well as kids are taller, on average, than those who were underfed.)
But setting all this aside, from a Testosterone Rex perspective, it’s not hard to understand why researchers might be interested in looking for a correlation between digit ratio and financial risk taking. The traditional view of sex differences in the brain (as we saw in Chapter 4) is that the high levels of T produced by the newly developed testes of unborn boys plays a singularly important role in creating discreet “masculinized” circuits in the brain. These circuits, especially when activated by the higher levels of T at pubescence and beyond, are the basis of distinctly male sexually selected mating behaviour: like fighting off other contenders for the cosiest cave; hunting ferocious, meaty prey; and today, apparently, buying high-risk biotech stocks. Put these outdated assumptions together with another, that we met in Chapter 5—that risk taking is a stable, masculine personality trait—and the long chain of reasoning is complete. A person with a lower, more male-typical digit ratio will have a more “male brain”; a person with a more “male brain” will be more masculine; a person who is more masculine will be more risk taking; and a person who is more risk taking will be more financially risk taking. Thus, someone who, according to their digit ratio, was putatively exposed to more prenatal testosterone will, many decades later, be reliably more likely to say: “What the hell, I’ll take the 30 per cent odds of winning one dollar, rather than the twenty cents for sure.”
You’ll already be aware from earlier chapters of the weakness in several links in this chain. The assumption that resources and status—and therefore risk taking and competition—are distinctly male concerns in the struggle for reproductive success (and should therefore be wired into a “male brain”) was dissected and found wanting in the first part of the book. In line with this aband
onment of dichotomous “competitive males” versus “coy females” thinking in evolutionary biology is the shift in neuroscientific understanding of sex and the brain. The notion of discreet, T-induced “male” circuits is being replaced with a more complicated, interactive mishmash of factors, out of which emerge a variety of shifting “mosaics” of brain characteristics. This, in turn, fits nicely with what we know about sex differences in behaviour. These certainly exist but, again, in ways that create mosaics rather than categories. Put this all together, and it probably shouldn’t surprise us too much that recent meta-analyses and a large-scale study failed to find convincing evidence for correlations between digit ratio and other supposedly quintessentially masculine behaviours: aggression,55 sensation seeking,56 dominance, and both aggressive and non-aggressive risk taking in adolescents.57
Along similar lines, as we saw in Chapter 5, although you might assume that your friend, Ankush—who goes skydiving every weekend—must therefore be “a risk taker” with a “male brain” (thanks to an abundance of testosterone prenatally and/or in adulthood), you may well later discover that Ankush’s investments are all in government bonds. Looking for links between financial risk taking and testosterone exposure makes sense from the old perspective we met in Chapter 5, in which risk taking is assumed to be a stable, domain-general personality trait. But less so, however, from a more nuanced understanding of risk-taking behaviour as custom-made for each situation out of the “unruly amalgam” of factors, including social identity, norms, knowledge, past experience, social context, and the perceived risks and benefits in that particular domain. Which kind of risk taker do you expect to have the most male-typical digit ratio? And in which circumstances?
Testosterone Rex Page 16
