Testosterone Rex

by Cordelia Fine

  The same exact event may elicit different responses, depending on the way it is appraised by different individuals or by the same individual at different moments in time (e.g., in different social contexts).68

  Factors that researchers suggest could be important for how and when T responds to competitive situations include your appraisal of the skills of your opponent, the explanations you come up with to explain why you won or lost, how familiar you are with the “where” and “who” of the competitive situation, and your underlying motivations.69 This is where gender can make its entrance. We might, for instance, expect gender to influence T reactivity via the stereotypes that help create expectations or shape explanations for success or failure, the existing inequalities that create double standards for performance, and gendered experiences and social networks. After all, as we saw in the previous chapter, a different domain (masculine versus more neutral or feminine), a different cultural background (for example, patriarchal versus matrilineal), or even a different framing of the same competitive context (sports news assistant versus administrative assistant) can eliminate sex differences in willingness to compete.

  In fact, research with men has already shown the effects of culture and social constructions of gender on hormonal biology. Take the enduring effects of a successful ten-year Fast Track intervention programme in the United States. Targeted at boys at high risk of later antisocial behaviour, it “was designed to build social competencies and self-regulatory skills that enable children to respond more calmly and less vociferously to provocation.”70 Some participants received the intensive, decade-long intervention; a matched control group didn’t. Many years later, when the participants were in their mid-twenties, Nipissing University social neuroendocrinologist Justin Carré and colleagues invited about seventy men from the study into the lab, and tested the aggressiveness of their reaction to a provocation (supposedly, another participant vindictively stealing points from them in a game). The intervention group were less likely to retaliate against what they assumed was hostile behaviour on the part of a confederate, showing long-lasting effects of the intervention. But most interesting for our purposes is that they also showed less testosterone reactivity to the provocation, and this in part seemed to underlie their greater inclination to turn the other cheek in a competitive context. As Carré and colleagues conclude:

  Together, these results suggest that the Fast Track intervention creates persistent changes in psychological processes underpinning how individuals encode, interpret, and process social threat and provocation. These mental processes, in turn, influence the pattern of testosterone responses to provocation, which in turn influence aggressive behavior.71

  A similar conclusion emerges from a classic experiment by University of Illinois psychologist Dov Cohen and colleagues.72 This time the comparisons were between non-Hispanic white male students who hailed from either the northern or southern regions of the United States. In a series of experiments, groups of men were placed in a contrived challenge to their social status in which they were bumped in the shoulder by a male decoy. The decoy then added insult to injury by muttering an offensive word. To northern students, this event was of little consequence. But the southern students, raised in the lingering remnants of a culture strongly grounded in a man’s honour and the importance of the respect given to him, tended to walk away feeling worried about the effect on their reputation as a man. It was a slighted southern group, too, who increased their aggressive and domineering behaviour afterward. And again, it was only disrespected southern men who showed testosterone increases in reaction to this small challenge to their status. The discussion section of the study reassures that the experimental manipulations didn’t “produce any truly violent behavior.”73 But suppose an unfortunate confederate had been punched by an affronted Southerner. Would it make sense to blame the testosterone? Or to say that boys will be boys?

  In Testosterone, Herbert concludes that “the human brain has had to devise multiple ways of regulating, channelling, and optimizing the powerful effects of testosterone on male behaviour through laws, religion, and customs.”74 But there is no “real,” “original,” or “intended” testosterone level or reactivity with which civilization then interferes. As Wade remarks:

  Hormones, then, are not part of a biological program that influences us to act out the desires of our ancestors. They are a dynamic part of our biology designed to give us the ability to respond to the physical, social, and cultural environment.75

  The studies showcased in this chapter vividly illustrate Agustín Fuentes’s observation (echoing many feminist scientists),76 that “when we think about humans it is a mistake to think that our biology exists without our cultural experience and that our cultural selves are not constantly entangled with our biology.”77 And culture seems to enjoy the upper hand.

  OVER THE PAST EIGHT YEARS or so, I’ve taken part in a lot of discussions about how to increase sex equality in the workplace. Here, I would like to clearly state for the record that castration has never been mentioned as a possible solution. (Not even in the Top Secret Feminist Meetings where we plot our global military coup.) For organizations looking to increase the representation of women at senior levels, HR would eliminate castration immediately on obvious ethical and legal grounds. However, it can also be dismissed for scientific reasons. There’s nothing in the scientific literature to suggest that castration—even in conjunction with testosterone patches for women—would provide a powerful biological shortcut to equality. It doesn’t work for fish or monkeys, so why would it work for us? The very factors—status, experience, meaning—that become entangled with, moderate, substitute for, and override testosterone are human specialties par excellence—and no king emerges out of these complex interrelations.

  What would work, the research instead suggests, are major and sustained interventions on status, experience, and what a particular situation means to the individuals involved. This, it’s worth pointing out, poses a much more difficult challenge than administering hormone boosters or blockers.78 The “broad tapestry”79 of gender is tightly woven, and thick with redundancy: you can loosen one thread, but the others will still hold everything in place. A cichlid fish has size and flashy colouring to mark status. We have stereotypes that stain every encounter, clothing, language, salaries, titles, awards, media, legislation, norms, stigma, jokes, art, religion… the list of phenomena that make up our rich, gendered cultures goes on and on.

  That’s a lot of social construction to reconstruct. The big mistake is to confuse the persistence of the status quo with the dictates of testosterone.

  CHAPTER 7

  THE MYTH OF THE LEHMAN SISTERS

  There’s a very simple reason why most financial traders are young(ish) men. The nature of trading incorporates all the features for which young males are biologically adapted.… The whole set-up seems to have been designed for young men. All the actions of testosterone are echoed by the qualities of a successful trader. It does seem remarkable that the artificial world of financial trading should so suit the innate characteristics of young males.

  — JOE HERBERT, Testosterone1

  “IF LEHMAN BROTHERS HAD BEEN LEHMAN SISTERS, RUN BY WOMEN instead of men, would the credit crunch have happened?”2 This question, posed by a Guardian business editor, triggered a “frenzied engagement in the international media with the gender question in international finance.”3 Some commentators, drawing on research reporting links between testosterone levels and risk taking, argued an urgent need for a greater “diversity of hormones”:4 more women (and older men) would make for less testosterone. Article headlines and interviewees repeatedly invoke T-Rex, calling for more “mistresses of the universe”5 to bring some much-needed financial conservatism to the “testosterone rules”6 world of business.

  By now, the argument will be tediously familiar. Men, thanks to past evolutionary pressures, take risks in order to acquire the resources and status that led to reproductive success in our a
ncestral past. But fast-forward those Stone Age “male brains” to twenty-first-century global finance and the “evolutionary hangover” creates havoc, as Nicholas Kristof summarizes the view in the New York Times.7 Enter subprime mortgages and credit derivatives, and with the benefit of hindsight it became clear just how dangerous it was to have all that testosterone in charge with barely a woman in sight. (Fully clothed ones, that is.)

  It may seem a nice tribute to women to suggest that the world’s financial system might not have been brought to its knees if only more of their representatives had been around. We should certainly take a moment to appreciate the contrast with a New York Times article published about a century earlier, reporting a movement among broker firms to forbid women from frequenting their offices.8 As one, apparently typical, letter sent out to female clients by a Broadway firm of stock brokers explained, this was because their more valued customers “consider it undignified for women to frequent brokers’ offices.” Indeed, a woman is “a nuisance anywhere outside of her own home,” one broker pointed out. Women, it was noted, not only lacked the “business instinct”—males presumably bursting forth from the womb in a pin-striped birthday suit with an innate understanding of finance—but were also incapable of acquiring it. It’s certainly a long way from this to the May 2010 Time magazine cover featuring the female financial regulators—Elizabeth Warren, Sheila Bair, and Mary Schapiro—“charged with cleaning up the mess”9 made by you know who.

  There is a drawback, however, to belonging to the sex biologically suited to play the immobilizer of the “Big Swinging”10 organs of finance. A risk-averse temperament may be good for the world economy, but bad for one’s own personal finances. You don’t have to be an analyst of distributions of wealth to be aware that the exercise of liquidating one’s assets into a heap of dollar bills then gloatingly hurling oneself onto it would be a more comfortable experience for men, on average, than for women. Annual “rich lists” unfailingly confirm that it is almost exclusively men who would be at highest risk of suffocation during such an exercise. Historically, this inequality was easy enough to explain. Beyond advising a female client to marry well, even the most talented financial adviser would struggle to assist the wealth building of someone excluded from higher education, legally unable to own property and securities, and restricted to only the most low-paid occupations. However, these external barriers were dismantled some time ago, and many researchers have started to look to internal factors—like prenatal and circulating testosterone—to explain what has been described as “fundamental differences” in risk preferences, as one much-cited review put it.11

  But by this point in the book you may be justifiably sceptical that T can create a “fundamental” divergence in the financial decision-making styles of women and men. As we’ve already seen, a major challenge to a T-Rex view of sex differences is the typical extent of the overlap in the ways that men and women, on average, behave. Of all the qualities we possess, was it really on those relating to finance that sexual selection acted with the most vigour, way back in the Pleistocene?

  IN A HELPFULLY FORENSIC analysis of the scientific literature, the economist Julie Nelson reviewed eighteen studies of sex differences in financial risk taking from (and representative of) the economics literature.12 Some of these studies used the lottery tasks so beloved of economists, in which people make a series of choices between, say, a sure $5 or a 50 per cent chance of winning $10. Others asked people to report on their preferences for financial risk taking in real life, or looked at how they allocated their actual financial assets among more and less risky options (like stocks versus bonds). As you might already anticipate, the effect sizes for these differences were, with a few exceptions, generally quite modest, with several null results (that is, no sex differences), and even two findings of greater female financial risk taking.13

  How do we get from this overlap between the sexes to claims of a fundamental difference? By way of partial explanation, researchers often summarize results from earlier studies in an inaccurately stereotype-consistent way, observes Nelson. Researchers also tend to emphasize their own findings that are consistent with the stereotype of male risk takers, while downplaying (even sometimes to the extent of more or less ignoring) results that aren’t. This aroused Nelson’s suspicion that researchers are “tending to ‘find’ results that confirm socially held prior beliefs”14—a classic case of “confirmation bias.” If these results are more likely to be published, the scientific literature becomes skewed towards the expected conclusion.

  As it happens, there’s a way of presenting data, called the funnel plot, that indicates whether or not the scientific literature is biased in this way.15 (If statistics don’t excite you, feel free to skip straight to the probably unsurprising conclusion in the last sentence of this paragraph.) You plot the data points from all your studies according to the effect sizes, running along the horizontal axis, and the sample size (roughly)16 running up the vertical axis. Why do this? The results from very large studies, being more “precise,” should tend to cluster close to the “true” size of the effect. Smaller studies by contrast, being subject to more random error because of their small, idiosyncratic samples, will be scattered over a wider range of effect sizes. Some small studies will greatly overestimate a difference; others will greatly underestimate it (or even “flip” it in the wrong direction). The next part is simple but brilliant. If there isn’t publication bias towards reports of greater male risk taking, these over- and underestimates of the sex difference should be symmetrical around the “true” value indicated by the very large studies. This, with quite a bit of imagination, will make the plot of the data look like an upside-down funnel. (Personally, my vote would have been to call it the candlestick plot, but I wasn’t consulted.) But if there is bias, then there will be an empty area in the plot where the smaller samples that underestimated the difference, found no differences, or yielded greater female risk taking should be. In other words, the overestimates of male risk taking get published, but various kinds of “underestimates” do not. When Nelson plotted the data she’d been examining, this is exactly what she found: “Confirmation bias is strongly indicated.”17

  This bias makes it misleading to draw conclusions from the entire literature, which inflates the size of the sex difference overall. So what differences in size did Nelson see when she looked only at the more precise results from the eight largest studies?18 These included a large-scale newspaper survey that asked tens of thousands of respondents a lottery question, and two large-scale analyses of actual investments (thousands of retirement portfolios, and more than thirty-five thousand stock investment accounts).19 Nelson’s best estimate based on these eight most precise studies was an effect size of about 0.13. This translates into about a 54 per cent probability of a man picked at random being more financially risk taking than a woman picked at random. When you consider the conclusions of Chapter 5 regarding the often-gendered factors that are so important in explaining differences in people’s risk-taking behaviour—like knowledge, familiarity, past experience, and gender norms that associate risk taking with masculinity—it begins to seem almost surprising that the difference is so small. Consider, too, that a person’s wealth, likely future wealth, and financial security understandably affect the kinds of financial risks they’re likely to be willing to make.20 Although researchers can and do take account of men’s and women’s current wealth and earnings when comparing their risk taking, in a society in which men are more likely to be promoted, and women are more likely to suffer the career costs of caring for children and elderly parents, the financial trajectories and expectations of even initially comparable women and men are unlikely to be the same.

  Sex differences in financial risk taking aren’t just small, they’re also conditional, showing themselves in some tasks, samples, and contexts, but not in others. One study, for instance, found no differences when the typical abstract lottery was expressed in the real-world context of investment
decisions.21 Similarly, Warwick Business School’s Ivo Vlaev and colleagues found no sex differences when lotteries were put into real-world contexts. In among the traditional abstract lotteries, they presented equivalent ones in terms of pension, salary, mortgage, and insurance contexts. For example, the student participants were asked to choose between, say, a job offering £30 daily payment for sure, versus a riskier but potentially more lucrative job offering a 50 per cent chance of earning £100, and 50 per cent chance of earning nothing. Overall, they found no differences between the sexes.22

  Another complication for the claim that men and women are fundamentally different in their approach to finances is that, as with other domains of risk taking (as we saw in Chapter 5) “this sample of females” and “this sample of males” shouldn’t be taken to stand in for all females and males. Evolutionary scientists Joseph Henrich and Richard McElreath used a traditional lottery task to compare financial risk taking in three groups far removed from the Western students typically favoured in these kinds of experiments.23 These were communities of small-scale farmers in Chile and Tanzania (the Mapuche and the Sangu, respectively), as well as a nearby Chilean non-farming community (the Huinca). Although cultural background made a difference—on average, the Mapuche and Sangu were risk prone, while the Huinca were risk averse—in none of the groups was being male per se linked with risk-taking propensity. Likewise, sex didn’t explain any variation in financial risk taking in a study of more than four hundred Chinese participants sampled from the general population.24 Meanwhile, another cross-cultural study comparing the matrilineal and patriarchal societies of the Khasi in India and the Maasai in Tanzania, respectively, failed to find any sex differences at all in risk taking on standard lottery and investment games.25