Automation Was Supposed to Liberate Women. What Happened?
Feminists thought automation would break down gender stereotypes in the workplace—but it could also entrench them even deeper.
In 1967's SCUM Manifesto, radical feminist Valerie Solanas set out what she felt were the duties of every "civic-minded, responsible, thrill-seeking female." Her aims were fourfold: overthrow the government, eliminate money, institute complete automation, and destroy the male sex.
In the 50 years since the manifesto's publication, none of these aims have really come to fruition. Only one has come close: "complete automation." Solanas's vision of women freed from the sexist shackles of capitalism through automatic voting machines and the replacement of the entire male sex with machines didn't quite hit the mark, but automation has developed at startling speed. Developments in artificial intelligence and machine learning capabilities have led to more and more processes being automated—particularly in the workplace.
Earlier this year, UK MP Tom Watson made a speech at the EEF Manufacturing Conference in London to discuss "the real impacts robots are having on business." In some places, humanoid automatons have replaced human service workers altogether: Pepper, a robot powered by IBM's intelligent cognitive system Watson, mans SoftBank stores in Japan, helping customers buy consumer electronics. Elsewhere in Japan, cyborgs, automated vehicles and emotive technologies all work together to serve visitors at the Henn-na Hotel. Automation is creeping into retail, service work, large-scale manufacturing, mining and more.
This utopian vision of full automation may sound appealing, but it has one obvious negative consequence—the loss of human jobs. And, according to the World Economic Forum, it's going to be women losing out most. It's not quite the radical feminist dream that Solanas had in mind.
In a report entitled The Future of Jobs, the WEF warned that two industries about to be revolutionised by automation—"office and administrative" and "manufacturing and production"— are often populated by women; women who could lose their jobs. "As disruptive change is coming to business models, there is also a risk that these trends and drivers of change might sustain or worsen existing inequalities," the report writes.
Four million men are set to face job losses in the next ten years due to automation compared to 1.4 million gains; one job gained for every three lost. Women, on the other hand, will lose fewer jobs—3 million—but will gain just 0.55 million. That's five jobs lost for every job gained.
Others, however, suggest that automation could still help to narrow gender gaps in the workplace. "At best, automation can increase productivity and save workers from menial, low-paying tasks and jobs," Giorgina Paiella, a researcher on feminism and automation at the University of Connecticut, told me.
Sonia Liff, an occupational psychologist, suggests in a paper on the effect of IT on office workers that many occupations are gendered not because of their content but because of our cultural understandings of what they mean. Administrative jobs, for example, have no tasks or responsibilities that are inherently gendered, but rather reflect wider societal expectations about the roles women should, or do, play in the workplace. It's these expectations that underpin workplace dynamics and give power to our ideas about gender. But automation, Liff says, disrupts these dynamics by shifting relationships between both workers and managers and men and women—which could ultimately be liberating.
"Feminist perspectives of the woman-machine relationship have long oscillated between pessimistic fatalism and utopian optimism"
But Katherine Turk, writing on secretaries in America from the mid 90s to today, argues that automation has so far been a double-edged sword. Though automation had promised to "liberate women from the drudgery assigned to them by outdated stereotypes," the glass ceiling persisted, and automation "continued to sharpen the class divide in office work."
Automation itself can be gendered, too. It may seem laughable to ascribe gender to a self-service checkout, but as humanoid robots become more developed, it's hard to avoid sexed robots. Robot receptionists, for example, are being built and designed to perform all the tasks a regular receptionist would. More often than not, they're female. Even unembodied personal assistant systems—Siri being the most obvious—are gendered. So isn't automation simply reinforcing gender roles, as well as replacing women for real?
"Gynoids tend to be sexualised and racialized," Paiella told me. "And as technological innovation continues to progress, I think it's a real possibility for these robots to become commonplace." More attention should be paid to the design and function of robots, she said, to ensure that they're not being unnecessarily gendered, and this design process should involve a critical analysis of our expectations and assumptions.
So what's the real key to a healthy woman-machine relationship? Feminist theory, according to Judy Wajcman, who has written extensively on automation since the late 80s. Wajcman argues that "technofeminism," which "fuses the visionary insights of feminism with a materialist analysis of the sexual politics of technology," is the way forward for women worried about the impact technology will have on their careers and lives.
It's a tempered approach that combines two schools of thought—that technology could free women from the gendered expectations we've come to expect in the workforce, but that it could also, given that it often adheres to those expectations by both purpose and design, reinforce them.
"Feminist perspectives of the woman-machine relationship have long oscillated between pessimistic fatalism and utopian optimism," Wajcman writes. "Social science needs to continually engage with the process of technological change. It is a key aspect of gender relations."
Paiella agrees that a feminist approach to automation is key. This could both help women specifically losing their jobs to automation, because gender divides in the workplace would be less obvious, but could also affect gender equality more broadly by allowing us to re-evaluate our expectations of what makes labour gendered.
"Specifically in the case of female labour? I believe automation not only can free women from undesirable tasks, but can also help break down the gendered associations ascribed to certain jobs," she said.
Silicon Divide is a series about gender inequality in tech and science. Follow along here.