Tuesday, March 19, 2013

Fun With Christina Hoff Sommers

She is best known as an anti-feminist who wrote a book attributing the world-wide phenomenon of boys doing worse at school than girls to American feminists, pretty much.  Now she has joined the Lean-In debate, centered around Sheryl Sandberg's book (which I still haven't bought).

Guess what her message might be?  It's not very hard if you know that she is an anti-feminist.

Yup.  She argues for innate differences between men and women as the cause why women don't really want to lean in.  That's a rather weird response, given that nobody is forcing women to lean in.  The approach is aimed at women who want to advance at work.  It's not some new feminazi law or anything similar.

But Hoff Sommers does have a deeper goal here, I think, and that is to argue that there is no actual need for businesses to change, because the reason women CEOs are almost as rare as hen's teeth is that women don't want to be CEOs.  And of course most women don't want to be CEOs, at least after they figure how unlikely that outcome might be.  But then that's true of most men, too,  I would guess.

Never mind.  Us girls get a separate treatment, because not giving us such a separate treatment is ignoring the fact that men are innately and immutably different.  With which I agree, of course.  Only women can give birth, for example.   The other possibly innate characteristics are much trickier to analyze.  The state of the arts right now seems to be that everyone and their aunt Agatha decides on an opinion and then believes that it is a fact.  I prefer to stay skeptical.

Anyway, Hoff Sommers uses two pieces of evidence for her arguments.  One is the Pew survey I discussed in an earlier post. Hoff Sommers:
In a 2013 national poll on modern parenthood, the Pew Research Center asked mothers and fathers to identify their "ideal" working arrangement. Fifty percent of mothers said they would prefer to work part-time and 11 percent said they would prefer not to work at all. Fathers answered differently: 75 percent preferred full-time work. And the higher the socio-economic status of women, the more likely they were to reject full-time employment. Among women with annual family incomes of $50,000 or higher, only 25 percent identified full-time work as their ideal. Sandberg regards such attitudes as evidence of women's fear of success, double standards, gender bias, sexual harassment, and glass ceilings. But what if they are the triumph of prosperity and opportunity?
Or what if they are traditional gender role requirements?  I think*  that she makes a mistake here.  The parents were not asked what they themselves would prefer to do about work.  They were asked what they thought was in general better for mothers and fathers.  The distinction is subtle but it does make a difference, because the way the question was framed means that general gender role norms could enter the answer.  That is less likely if people were asked what they themselves wanted to do.

But the major piece of evidence Hoff Sommers uses is a 2008 international study about gender differences:
Sandberg's goal is to liberate her fellow Americans from the stereotypes of gender. But is that truly liberating? In a 2008 study in the Journal of Personality and Social Psychology, a group of international researchers compared data on gender and personality across 55 nations. Throughout the world, women tend to be more nurturing, risk averse and emotionally expressive, while men are usually more competitive, risk taking, and emotionally flat. But the most fascinating finding is this: Personality differences between men and women are the largest and most robust in the more prosperous, egalitarian, and educated societies. According to the authors, "Higher levels of human development—including long and healthy life, equal access to knowledge and education, and economic wealth—were the main nation-level predictors of sex difference variation across cultures." New York Times science columnist John Tierney summarized the study this way: "It looks as if personality differences between men and women are smaller in traditional cultures like India's or Zimbabwe's than in the Netherlands or the United States. A husband and a stay-at-home wife in a patriarchal Botswanan clan seem to be more alike than a working couple in Denmark or France."

Why should that be? The authors of the study hypothesize that prosperity and equality bring greater opportunities for self-actualization. Wealth, freedom, and education empower men and women to be who they are. It is conspicuously the case that gay liberation is a feature of advanced, prosperous societies: but such societies also afford heterosexuals more opportunities to embrace their gender identities. This cross-cultural research is far from conclusive, but it is intriguing and has great explanatory power. Just think: What if gender difference turns out to be a phenomenon not of oppression, but rather of social well-being?
I got the study** and skimmed it, fairly quickly.  Then I had to go back and check that my reference was correct, because the study doesn't look at nurturing, risk aversion or competitiveness.  It looks at four characteristics:  extraversion, agreeableness, conscientiousness and neuroticism, and finds that men and women differ along those dimensions, on average.

The measure of difference, the d-value, ranges overall from 0.4 for neuroticism to 0.1 for extraversion.  These are fairly small gender differences.  The d-value for the difference between the average heights of adult men and women is 2.6, for example.  A value of zero denotes no average gender difference.

The researchers argue that more egalitarian societies (which are economically more developed societies)  exhibit larger differences than traditional and more hierarchical societies.  BUT the result is completely driven by differences between men in the two types of societies,  NOT by differences between women.

So assuming that all calculations in that study are correct, what do the results tell us?  That women and men in traditional cultures are closer to each other in extraversion, agreeableness, conscientiousness and neuroticism than women in more egalitarian societies, and that this result is driven by differences in men.  How does this relate to Hoff Sommers' argument that women are freer to be what they are innately meant to be in the more egalitarian societies?

I can't think of any actual relevance of the study to her argument, unless she argues that men in more egalitarian societies are freer to assume the male breadwinner role?  But that is rubbish.  It is the hierarchical societies where men are more likely to assume that role.

Neither can I see what relevance the traits of extraversion, agreeableness, conscientiousness and neuroticism have to do with women's desire to lean in (unless neuroticism makes work impossible, of course) .

In short, that particular evo-psycho study does not support Hoff Sommers' argument.  Even if the characteristics it studies were directly relevant for labor market participation rates of men and women, the average difference values are pretty small.  But I do admire the web she weaves from rather unlikely bits and pieces!

Sigh.  None of what I say here will have any influence on the debate.

The first footnote is added later, to clarify that point:
*I wrote "I think that Hoff Sommers made a mistake" because I couldn't match her data.  The closest I got was with the questions about what is ideal for mothers and fathers in general.  But a second search unearthed the most likely bit she refers to.  Only it doesn't apply to all  mothers but to mothers who work either part-time or full-time.  The Pew report states this:
 The recent shift toward a preference for full-time work has been more pronounced among working mothers themselves than among those who are not employed. Fully 37% of today’s working mothers say their ideal situation would be to work full time, up from 21% of working mothers in 2007. (Among non-working mothers, the increase from 16% to 22% is not statistically significant.)
Only 11% of working mothers say their ideal situation would be not to work at all, down from 19% in 2007. Part-time work remains the most appealing option for working mothers; 50% now say working part time would be ideal for them, down marginally from 60% in 2007.
Among mothers who do not work outside the home, in 2007, roughly half (48%) said not working was their ideal situation. Today only 36% of these mothers say the same. The share saying they would prefer to work either full or part time has increased slightly over the same period (from 49% in 2007 to 63% now).
For their part, fathers prefer full-time work. Fully 75% of fathers with children under age 18 say working full time is ideal for them. Some 15% say working part time would be ideal, and
10% say they would prefer not to work at all. In general, fathers’ views about what is ideal for them have not changed significantly in recent years.

**To get the pdf file, Google Why Can't A Man Be More Like A Woman. Sex Differences in Big Five Personality Traits Across 55 Cultures.