Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
The writer is an associate professor at the University College London Social Research Institute
When Sheryl Sandberg’s book Lean In came out in 2013, it struck a nerve with many women. In it, she posited alongside structural barrier, women are “holding themselves back” with their low confidence.
My research supports Sandberg’s hypothesis. But it also tells us that the problem is far more complex. Women’s underconfidence manifests long before they enter the labour market.
Using data from a Centre for Longitudinal Studies survey of individuals born in the same week in Britain in 1970, my colleague Anna Adamecz and I constructed a measure of over and underconfidence. We looked at the study’s participants’ performance on objective tests in mathematics, English, and IQ, as well as self-assessed ability questions (for example: how good are you at mathematics?), which were administered at ages five, 10 and 16.
Individuals who rated themselves as higher than their position in the actual skills distribution are classified as overconfident. Those who rank themselves lower are categorised as underconfident.
Perhaps unsurprisingly, women are more likely to be classified as underconfident. Thanks to the power of longitudinal data, we can follow these individuals into the labour market and track any gender gaps in their earnings when they are mid-career, at the age of 42.
Being underconfident is not the only reason for the gender wage gap, but does play a meaningful role. It is not simply the case that men are being rewarded with higher pay for being overconfident (and so are the few women who fall in that category), but that underconfident people, primarily women, face a pay penalty.
Why is that? Several factors are at play. Underconfident people (both men and women) perform worse on their school leaving exams, are less likely to study high return subjects such science and law at university and end up in less lucrative occupations.
Since women are more likely to be underconfident, they are worse affected by this penalty — despite outperforming men at school and university. Introducing confidence building training at work is therefore too little, too late. The underconfidence penalty already manifests in decisions about what to study at university, and which professions to enter.
It makes sense, then, to think about when and why underconfidence emerges. Using data on all twins born in England and Wales from 1994-96, we have started to unpack those drivers. In this dataset, boys and girls were asked to self-assess their mathematical ability at age nine. We also observed their actual grades in the subject.
If we take a boy and girl with the same level of mathematical performance, the boy self-reports as being much better at the subject than the girl. This difference is even more pronounced when we compare twins, and account for their shared genes and environment. Parents were also asked to assess the mathematical ability of their children: they too overestimated their sons’ ability and underestimated that of their daughters. Within families, the boy tends to be seen as “the maths person” and the girl is not, even if she is better at maths.
Economists are good at modelling many things, but trying to fully capture the gender norms and stereotypes that feed into female underconfidence is challenging. Home, school and the workplace are all environments where boys may receive overly positive reinforcement, especially when it comes to high-return subjects such as science and technology, and girls might not.
This state of affairs has labour market implications — including the gender wage gap. Instead of putting the onus on women to “lean in” in the workplace, we as parents, teachers, and employers can do better by ending the transmission of these norms to the next generation.
Read the full article here