“A gender-conforming man in South Africa has an expected employment rate of 48%. Gender non-conforming men have an 11% lower probability of being employed, gender non-conforming women 16% and gender-conforming women 12%. So all of these groups face a lowered probability of employment compared to gender-conforming men,” said Debra Shepherd.
“Gender-conforming men are also more likely to have tertiary education, more likely to be in urban metropolitan areas and have higher income.”
These are some preliminary findings from a detailed data analysis of a nationally representative dataset of 5600 South African adults undertaken by STIAS Iso Lomso Fellow Debra Shepherd of the Department of Economics at Stellenbosch University.
“I’ve used data from the South African Social Attitudes Survey which, for the first time, introduced questions around self-reported sexual identity and gender expression in 2015/16,” said Shepherd. “I’m looking at what interesting questions can be investigated using these data.”
Shepherd pointed out that although gender discrimination has been widely researched in economics gender identity has remained under the radar of economists.
“In economics we can be reductive about social identities. Economics hasn’t ignored identity but it has ignored it in a more complex, sociologically informed way,” she explained. “Most gender-discrimination research in economics takes a ‘dummy’ variable view of gender that upholds hegemonic gender categorisations. Identity is seen as a category to which you belong, and sex and gender are often thought to be one and same. There is also no interplay between the choices individuals make for their identity and the choices imposed by society.”
She described gender framing as a cultural frame which is both descriptive and prescriptive.
“Shared cultural beliefs about categories lead to stereotypes which can be descriptive – beliefs about the typical characteristics of men and women. Research has shown these to be consistent across culture, time and context. We also have prescriptive stereotypes – ideas about how men and women should be.”
“In the workplace these descriptors can promote negative expectations about an individual’s performance. For example, a stereotype in the workplace is that men are more competent whereas women are more nurturing.”
“Gender frames drive gender stereotypes through focusing on difference, inclusion and exclusion,” she continued. “This can lead to inequality. Furthermore, they mould the interests of men and women such that behaviour becomes actively gendered and essentialist expressions of gender difference are enacted, providing for prescriptions (of behaviour) that can be enforced through stigmatisation and threat of sanction.”
“How we frame gender is based on heteronormative ideas. Heterosexuality is seen as the norm and heteronomic constructions of masculinity and femininity are seen as natural. This supports traditional ideas of how masculinity and femininity are aligned in a ‘natural’ hierarchy.”
“Within these constructions sexuality and gender tend to be perfectly aligned, which, of course, they are not,” she added.
But context is important
Shepherd also pointed out that gender framing operates in context and can come to the foreground or move to the background depending on, amongst other things, the salience of sex differences and stigma consciousness.
“If everyone accepts the social norms for gender framing then it gets pushed to the background – this is how men and women are so we don’t have to think about it.”
“But if you are a woman entering a male-dominated space, then your gender identity can be pushed to the front.”
“And if you challenge the accepted notions of masculinity or femininity you may be seen as a ‘token man’ or labelled ‘a bitch’,” she said.
“Generally women are more affected by gender stereotypes,” continued Shepherd. “They also worry their behaviours will be interpreted as stereotypical. They are more conscious of the gender game.”
Interestingly, she noted that in wealthier settings we often go back to primary cultural frames to denote difference.
“If access to material goods are equalised you don’t have to consider those differences anymore but the more basic differences like gender and race may become more prominent.”
And the stereotyping starts young. Shepherd highlighted data from South Africa looking at self-reporting on maths abilities by Grade 9 learners which shows that in wealthy, private-school settings girls have the lowest sense of their abilities in maths and are the least motivated to do maths.
“There is no difference in poorer schools which would indicate that in wealthy settings their gender is being pushed to the front. Gender framing clearly has some impact here.”
She indicated that she found this surprising. “Everyone believes that the wealthy school system works for everybody and that it is a progressive space in which our daughters will thrive. But clearly there is something in the framing that is creating self-doubt and demotivation for women.”
Fluid and changing
Shepherd described gender as something we do and pointed out that the dominant definitions of what is typically masculine and feminine can change over time. In surveys she has done with her students she has found that many of the men define themselves as falling into the indeterminate category – not strongly masculine or feminine – so maybe the definitions are shifting.
“There is also rich variation in people’s experiences,” she said. “Variation that has generally been ignored in the literature.”
She pointed to the challenges of limited sample sizes and the sensitivity of the data which probably leads to under capturing – “the data are a snapshot – who that person is on that day”.
“On paper South Africa is a progressive society but when you ask people questions in the privacy of their homes a different reality is often exposed,” she said.
“Of the respondents, 98% self-described as heterosexual and only 2% as LGB. Roughly 80% gender conform either ‘strongly’ or ‘mostly’ to the male-masculine and female-feminine norms.”
“A self-defined LGB woman who is gender non-conforming has a 28% lower probability of employment. LGB women had the highest expected unemployment rate – they are significantly more likely to be unemployed.”
“The hurdle to get into paid employment is greater for LGB women; however, once in, they earn at the same level as gender-conforming men.”
Other research on sexual orientation and wages has found that overall women still earn less than men and that gay men earn less than heterosexual men.
The data also show some interesting racial dimensions. “Gender conformity amongst men is highest amongst white males,” said Shepherd. “For women, both white and black African females indicate higher representation of non-conformity. There seems to be very little gender non conformity amongst the coloured population.” However, she cautioned that the sample sizes are very small.
“This is the first time these types of questions have appeared in a dataset of this size,” she said. “Most of the other data available come from small case-specific studies. So I’m trying to say something that has never before been shown using these kinds of data.”
“I’m unsure exactly where this study fits in in terms of discipline. My goal is to show a relationship that hasn’t been shown and open the arena for more questions,” she continued. “I’d like to have some impact on the way gender discrimination is analysed within my discipline but I also want my research to be activist. I’m specifically working with students, as I see my immediate context as the place where my research can have impact. My long-term goal is to transform the space in which I exist.”
Michelle Galloway: Part-time media officer at STIAS
Photograph: Christoff Pauw