Human Diversity

by Charles Murray

  It should be pointed out that these hypotheses do not argue that the wider sex differences in personality in the more egalitarian countries are illusions. They de facto acknowledge the reality of the large differences in advanced countries; it’s just that the differences are masked in countries with strong sex roles. But the Schmitt study argues that in fact the hypotheses cannot be sustained in the face of the patterns in the data. Instead, the authors introduced an important new empirical perspective to explain the phenomenon of widening sex differences: Perhaps we’re looking at a general phenomenon that goes far beyond personality traits. For example, the Schmitt study points out, sexual dimorphism in height increases with a country’s wealth. So too with sexual dimorphism in blood pressure. So too with competitiveness in sports—as opportunities and incentives increase for women to compete in sports, sex differences in performance increase as well. So too with differences between advantaged and disadvantaged groups in health and education when new opportunities are made available to all. Two years after the Schmitt study made these points, another study led by Richard Lippa found that sexual dimorphism in visuospatial abilities also increased with gender equality.40

  Another surprise from the Schmitt study was its finding that men do most of the changing, in both the physiological and personality traits. When sexual dimorphism in height increases, for example, it is primarily due to greater height among males. In the case of personality, the Schmitt study found that the wider sex gap in emotional stability in advanced countries is not the result of women becoming less emotionally stable, but of men self-reporting higher levels of emotional stability, and also lower levels of agreeableness and conscientiousness, than men in less advanced countries.

  Whatever the explanation turns out to be, the evidence about personality profiles around the world needs to be taken on board by orthodox academics. In 2016, David Schmitt and colleagues returned to the body of evidence that had accumulated since the turn of the twenty-first century, summarizing it this way:

  Psychological sex differences—in Big Five traits, Dark Triad traits, self-esteem, subjective well-being, depression, and values—are demonstrably the largest in cultures with the lowest levels of bifurcated gender role socialization or sociopolitical patriarchy. Ultimately, the view that men and women start from a blank slate simply does not jibe with the current findings, and scholars who continue to assert gender invariably starts from a psychological blank slate should find these recurring cross-cultural patterns challenging to their foundational assumptions.41 [Emphasis in the original.]

  I would add that the international story of sex differences in personality is challenging not only to advocates of the sex-is-a-social-construct position. I know of no ideological perspective that would have predicted greater sex differences in personality in Scandinavia than in Africa or Asia.

  Recapitulation

  The core message of this chapter is that the personality profiles of males and females are different in ways that break along the People-Things dimension worldwide.

  Many of the differences conform to stereotypes of masculine and feminine characteristics, which in turn prompts me to remind you once again that we are talking about overlapping distributions. Many males are closer to the female profile than are many females, and vice versa. But neither is it appropriate to minimize those differences. Sometimes the effect size for a single aspect of the personality inventories is huge, as in the effect size of +2.29 for the “sensitive, aesthetic, sentimental” factor in the 16PF standardization sample. Sometimes separate but conceptually related facets point to a large aggregated difference even when such aggregates have not been calculated, as in the case of the facets in the FFM that are related to the People-Things dimension.

  The traits that differ along the People-Things dimension are not shocking. The technical studies tell us that women are, on average, warmer, more sympathetic, more altruistic, and more sensitive to others’ feelings than men are. I suggest that these technical findings are face-valid. They match up with common human experience.

  Why do many of these differences apparently become more pronounced in the most gender-egalitarian nations? I will not try to adjudicate among the explanations that others have advanced, but I will disclose my own, admitting that it is completely ex post facto: The deprivations of freedom that women still suffer in traditional societies sometimes suppress the expression of inborn personality traits. For example, whatever genetic tendencies toward extraversion that women in a strict Muslim culture may have, they are under enormous cultural pressure to modify their expression of those tendencies to meet cultural norms. Perhaps such conditions also warp the expression of male personality. Under this hypothesis, genetically-grounded personality differences widen in the most gender-egalitarian societies for the simplest of reasons: Both sexes become freer to do what comes naturally.

  3

  Sex Differences in Neurocognitive Functioning

  Proposition #2: On average, females have advantages in verbal ability and social cognition while males have advantages in visuospatial abilities and the extremes of mathematical ability.

  As groups, men and women have different cognitive profiles. Those differences manifest themselves many different ways, leaving us with a lot of ground to cover in this chapter. I begin with summaries of the state of knowledge about a variety of specific abilities, then turn to cross-national data, and conclude with two syntheses that help tie the pieces together.

  Specific Skills and Aptitudes

  For some of these summaries, I make use of Diane Halpern’s fourth edition of Sex Differences in Cognitive Abilities (2012).1 I do not try to interweave the dozens of references Halpern cites, giving instead a single endnote at the outset with the pages for her book. I add endnotes for material that she did not use or that postdate the fourth edition.

  On Average, Females Have Better Sensory Perception Than Males2

  When it comes to the five senses—taste, touch, smell, sound, vision—the story is mostly one of small female advantages.

  Females tend to be better than males at detecting pure tones.

  Adult females tend to have more sensitive hearing for high frequencies than males.

  Females tend to have better auditory perception of binaural beats and otoacoustic emissions.3

  Females tend to detect faint smells better than males.

  Females tend to identify smells more accurately than males.4

  Males under 40 tend to detect small movements in their visual field better than females.

  Age-related loss of vision tends to occur about ten years earlier for females than for males.

  Males are many times more likely to be color-blind than females (the ratio varies by ethnic group).

  The balance of evidence indicates that females are more accurate than males in recognizing the basic tastes (sweet, sour, salty, bitter), though some studies find no difference.

  Females tend to be better than males at perceiving fine surface details by touch. This holds true for blind people as well as sighted ones.5

  Most of these differences have small individual effect sizes consistent with the stereotype that “women are more sensitive than men,” For taste, touch, smell, and sound, women are (on average) more sensitive instruments than men are. They are not a lot more sensitive on any single sense, but somewhat more sensitive on all of them. Even when it comes to vision, the male advantage in detecting movement is counterbalanced by a greater male disadvantage in color blindness.6

  Women are also more sensitive to pain than men. It’s not that women aren’t as tough as men; rather, the neurological experience of pain is more intense for women.7 Evidence for greater female sensitivity to pain is found even in infancy.8 Another aspect of this greater sensitivity is a pronounced sex difference in “disgust,” whether in response to exposure to pathogens (e.g., reactions to seeing a rat or to an oozing wound) or with regard to sexual activity that involves risk of disease or decreases reproductive fitness (e.g., inces
t). Effect sizes for sexual disgust can be large, ranging from +0.60 to +1.54.9

  On Average, Females Have Better Perceptual and Fine Motor Skills Than Males10

  Females have an advantage on certain perceptual-motor tasks. On digit-symbol coding, for example, where each symbol corresponds to a number (e.g., “substitute 2 for #”), women code faster than men do. Sometimes these differences involve large effect sizes, with d = +0.86 in one major study.

  Females have an even larger advantage in a variety of fine motor skills involving hand-eye coordination. Evidence for this advantage goes back to infancy.11 Diane Halpern observes dryly that while such tests of fine motor skills “are sometimes labeled ‘clerical skills tests’… I note here that fine motor skills are also needed in a variety of other professions such as brain surgery, dentistry, and the repair of small engines.”12 In tests of motor skills, it sometimes happens that men are faster but women are more accurate.

  On Average, Males Have Better Throwing Skills Than Females13

  Men have a substantial advantage in many large motor skills, but few of them have much to do with cognition. The major exception is males’ pronounced advantage on tasks that involve throwing objects accurately at stationary or moving targets, because that accuracy is highly dependent on visuospatial processing in the brain. Effect sizes have sometimes exceeded 1.0 and persist when right-handed subjects are throwing with their left hands.14

  On Average, Females Have Better Memory on Several but Not All Types of Memory15

  Memory comes in many forms—long-term and short-term, “autobiographical,” “episodic,” and “semantic,” among others. Here are the main themes of the research, mostly drawn from Diane Halpern:

  Females tend to be better than males at remembering faces and names.

  Females tend to be better than males at recognizing facial emotions.

  Females tend to be better at remembering the minutiae of an event (labeled peripheral detail), while males tend to be better at remembering the core events (labeled gist).16

  Females tend to remember speech they have heard better than males, particularly when it relates to emotionally laden events in their past.

  Females tend to retain memories from earlier childhood better than males do.

  Females tend to have better short-term memory than males (e.g., given a list of single-digit numbers, they remember longer lists than males do).

  Females tend to have better verbal working memory (e.g., remembering a list of numbers while answering questions about an unrelated topic).

  Females tend to have better memory for locations of objects (e.g., remembering where the car keys were left).

  Males tend to have better visuospatial memory (e.g., navigating on the basis of a combination of landscape features).17

  On Average, Females Have Better Verbal Ability Than Males in the Normal Range18

  There are no official definitions of “normal range” of ability versus “gifted.” The most common decision rule for placement in gifted programs is an IQ score of 130 or above, putting someone in the top 2 percent of the population. My purposes in the discussion below don’t require a hard-and-fast cutoff. When I discuss scores in the normal range, I report effect sizes and male-female ratios for entire samples. When I discuss scores at the high end, I focus on the top five percentiles. By “extreme high end” I mean the top two percentiles at most.

  On tests with nationally representative samples, females can be expected to consistently outperform males on a variety of verbal tasks, with a small advantage in reading and a more substantial advantage in writing. The note describes the details for the Cognitive Abilities Test (CogAT) and the National Assessment of Educational Progress (NAEP), which have produced effect sizes ranging from near zero to +0.20 for verbal reasoning, around +0.20 to +0.30 for reading, and in the +0.40 to +0.60 range in writing.[19]

  Reading disabilities. Boys experience dyslexia more commonly than girls. In the past, the size of the discrepancy has been clouded by referral bias, but epidemiological samples have established that the male-female ratio is in the range of 1.5 to 3.3, depending on the criteria for severity of the problem and the minimum IQ used for diagnosis.20

  EXPRESSING RATIOS

  I’ve just given the first of dozens of ratios that are reported in the book. Unless I specify otherwise, all the ratios I report are produced by dividing the male value by the female value, which means the base for all the numbers is 1, as in, for example, 3.3:1. I omit the base and report the ratio as 3.3, which is easier to read.

  Females Probably Retain an Advantage at the Extreme High End of Verbal Ability Before Puberty and Probably Lose It Subsequently

  At the beginning of adolescence, girls still have some advantage at the extreme high end of verbal ability, but it does not seem to persist through high school for American students. These are provisional conclusions awaiting further evidence.

  Gifted 7th graders. Since 1981, Duke University has sponsored the Duke University Talent Identification Program (Duke TIP), which now operates in a 16-state region of the South and Midwest. Students are invited to participate if they have previously scored in the top five percentiles for their grade level on the composite score of a standardized test or a relevant subtest (usually math or verbal). It attracts more than 100,000 students a year, accumulating more than 2.8 million participants since 1980. Upon entering TIP, participants take either the SAT or the ACT. Because they are taking college entrance tests designed for 17-year-olds, SAT scores obtained in their seventh year of schooling discriminate levels of ability among students in fractions of the top percentile.

  The female advantage in the top percentile persisted to the highest levels of verbal ability. For the top 0.01 percent in the SAT (the largest female advantage), there were 1.4 girls for every boy.[21]

  Gifted students as of grade 12. I’m still talking about verbal ability as measured by the SAT and ACT, but no longer about 13-year-olds. Rather, I refer to students who take it at the normal time. They are self-selected for academic ability. Stated conservatively, the test-taking populations for both the SAT and ACT are still concentrated in the upper half of the ability distribution.[22]

  For the test pools as a whole, the young women who take the SAT have consistently had fractionally lower scores than the young men. The opposite is true of the ACT—the females have fractionally higher scores. The note gives the details.[23]

  Sex ratios in the top few percentiles. What goes on in the top few percentiles of verbal skills for adolescents ages 17–18? The only source I have been able to find that casts light on the answer is the SAT data broken down not only by gender but by score intervals (the ACT does not publish such information). The SAT data are internally consistent in showing that the female advantage disappeared in the verbal reasoning test and was trivially small in the writing test at the top levels of ability, but it’s a single set of test batteries. The pattern needs to be replicated in other large databases before making much of it.

  Sex Differences in Math Ability in the Normal Range Are Inconsistent and Small24

  Now we turn to mathematics, which has gotten most of the attention in the debate about sex differences in test scores. It’s one of the rare cases in which the data are plentiful and the story doesn’t vary, at least within the United States:

  Females get higher classroom grades than males in math at all K–12 grade levels—but, for that matter, females get higher grades than males in just about everything during the K–12 years. On standardized tests, sex differences in mean scores on mathematics tests usually favor males, but the effect sizes are quite small for representative samples of students. A meta-analysis of the NAEP mathematics test from 1990 to 2011 found effect sizes of –0.07, –0.04, and –0.10 (favoring boys) in grades 4, 8, and 12 respectively.25 In the most recent test for grades 4 and 8, 2017, the effect sizes were –0.06 and –0.03 respectively. The most recent math test for grade 12 was in 2015, when the effect size was –0.09.26

  H
alpern’s review of meta-analyses of differences in math scores includes many other standardized tests showing similarly small effect sizes. To the question, “Is the typical male better at math than the typical female?” the answer is close to settled: “If yes, not enough to be noticeable,” with an open possibility that a small gap will close altogether.

  Males Have a Persisting Advantage in Math at the Extreme High End

  “Sex differences in mathematics become progressively larger as the sample becomes more selective and the type of math skill becomes more advanced,” writes Halpern, and herein lies a major issue in the study of cognitive sex differences.27 The literature has been extensive and aroused contentious reactions, but the dust has settled (as much as any dispute about sex differences is allowed to settle) on a few basic points.28 We once again have two major sources of data: scores of 12–13-year-olds who are tested in 7th grade using a college entrance exam (SAT or ACT), and for 17–18-year-olds who are tested in their senior year.