A Labour Senator has questioned what impact artificial intelligence might have on the “gender pay gap” in Ireland.
Speaking during a meeting of the Joint Committee on Artificial Intelligence last week, Senator Laura Harmon asked witnesses how AI might “help or hinder” the existing gender pay gap and sectors where women are disproportionately affected, and whether particular sectors might be affected more than others.
“I want to ask if any of them have any comment to make on how it would affect different demographics but specifically in terms of gender,” Harmon said.
“I acknowledge that all of the witnesses in the room today are women, by coincidence. I know there was a recent report published by TUD in the Irish Journal of Academic Practice, which shows the gender pay gap in higher education is still quite stark.
“We already have a gender pay gap. We already have sectors where women are disproportionately affected. I would like to hear how the witnesses think AI might help or hinder with that or if they think there are particular sectors that might be affected over others.”
Responding to the Senator, civil servant Fiona O’Byrne, assistant secretary at the Department of Further and Higher Education, Research, Innovation and Science, pointed to a “very significant gap” between men and women in the use of advanced digital skills.
“When it comes to basic digital skills, we do not see much of a gender gap,” O’Byrne said.
“When it comes to advanced digital skills, there is a very significant gap. A total of 44% of men in Ireland use advanced digital skills in their jobs compared with just 18% of women.
“That is a very substantial gap and is driven by relatively high advanced digital skill use among men rather than low use among women and by the makeup of our workforce overall. Women in Ireland are broadly comparable to women elsewhere in Europe. This is explained because women are under-represented in the most digitally intensive jobs.”
O’Byrne added that her department is monitoring data on exposed sectors while leading interventions to increase female participation in STEM fields.
When Harmon followed up to ask how further education and universities could improve digital specialisation – from a gender perspective or broadly, referencing a figure of 6.2% of digital specialists mentioned earlier by O’Byrne – the assistant secretary pointed to deep societal issues around preference and decisions made very early in life about what people see as being for them.
“These are deep societal issues around preference,” she said.
“A lot of the evidence shows the decisions are being made very early in individuals’ lives and what people see as being for them. That is where we would work with colleagues in the Department of Education on campaigns and on promoting STEM, in particular, for all students.”
Dr. Kara McGann, head of skills and social policy at IBEC, echoed those concerns and warned that children are already narrowing their career choices by the age of 12.
“To add to that, it is absolutely correct that gender, geography and generations are the three we have concerns about being left behind,” McGann said.
“From an upskilling and reskilling perspective that needs to be front of mind when we are looking at it. I absolutely echo what Ms. O’Byrne says about the occupational segregation that we see and the need for career guidance to really address that at a very early stage.
“We see that by the age of 12 they are narrowing career choices early on. That has a knock-on effect both on the gender pay gap and decisions around this, so it will only become more important as we move forward.”
Economic and Social Research Institute (ESRI) senior research officer Dr. Tara McIndoe-Calder told the committee that while a group of typically female-dominated occupations has low exposure to AI, the gender gap is widening in favour of men in roles facing the “highest risk of displacement.”
“Just to point out, there is a category which has low exposure to AI,” McIndoe-Calder said.
“There is a whole group of occupations that are quite typically female dominated; there is a high female proportion there. We see the only place the gender gap is in favour of men and is growing is in the most highly exposed, least complementary occupations. Those are the ones that are of highest risk of displacement.”
Harmon asked the researcher to specify which jobs fell into that high-risk category.
“Which ones would they be?” Harmon asked.
“Will Dr. McIndoe-Calder give examples?”
McIndoe-Calder listed several customer-facing and administrative roles as being among the most highly exposed and least complementary, while also noting that gender does not stand out as a factor limiting workers from moving between occupations once other characteristics such as age and education are accounted for.
“Those would be sales workers, clerks, clerical support workers and customer service workers; whereas the least exposed roles are skilled agricultural workers, trades workers, and in manufacturing and crafts, handicrafts and assemblers. Those roles have a higher share of women,” McIndoe-Calder said.
“When we look at the factors associated with whether people can successfully move between occupations, we do not really find an effect for gender.
“Once we have accounted for other characteristics such as age and education, gender does not stand out as something that limits people from moving occupations. It is a nuanced picture because there is this advanced digital skills divide and there is a gender pay gap.”
The exchange took place as part of the committee’s ongoing discussion on AI’s impact on the employment market and the broader economy.