Invisible Women: Data Bias in a World Designed for Men

After highlighting the many ways in which maleness is the universal standard, Perez then demonstrates how women are consequentially negatively impacted.

In the narrative of human development and evolution, there is a persistent bias. Evolutionary theory over-emphasizes the importance of hunting, but for a long time assumed that women had never been warriors despite physical evidence to the contrary. Cave paintings in France and Spain were presumed to be done by men, but now we know that the majority were done by women.

Language itself presumes a male standard. The World Economic Forum found that countries with languages that are “gender-inflected” (6), such as French, are “the most unequal in terms of gender” (7). Yet countries with genderless languages are not the most equal. The most equality occurs in countries whose languages allow gender to be stipulated, but do not embed it in words themselves. The male bias is so strong in language that people tend to interpret gender-neutral words as male. Initially, even emojis were mainly male.

The presumption of maleness is understandable given the over-representation of men. In films and on television, males have more speaking parts and leading roles. This imbalance replicates itself in statues, banknotes, the news media, and school textbooks so much that any attempt to point out the male doesn’t have to be standard invites criticism: “It is because what is male is universal that when a professor […] named her literature course ‘White Male Writers,’ she hit the headlines, while the numerous courses on ‘female writers’ pass unremarked” (12). Women seem to be a niche, although they make up half of humanity. Their status and accomplishments are often forgotten.

Historically, women’s accomplishments have been hidden or men have claimed credit for them. For example, 19th-century German Jewish composer Felix Mendelssohn published six of his sister’s pieces under his name. Because of the inaccuracy of historical depictions of women, male bias affects the modern world. For instance, to be included on British currency, the Bank of England requires that a person have broad name recognition—a standard that ignores the systemic exclusion of women from power. Given their vastly unequal financial resources, men have been better able to preserve their writings and creations. As a result, there is a substantial data gap about half of humanity.

The false dichotomy between the public and private spheres is another way to discount the significance of women’s lives. Dismissing the importance of the private realm glosses over the history of women’s oppression. White men, who see themselves as universally representative, complain about identity politics, failing to recognize that whiteness and maleness are identities, just as Blackness and femaleness are.

After a Swedish official joked that snow-clearing at least was beyond the reach of gender-equality initiatives, it turned out that even snow-clearing was optimized for male life. The snow-clearing schedule had not intentionally been designed to benefit men, but because of a gender data gap, roads were cleared first, followed by walkways and bicycle paths. As it happens, men are more likely to drive and women more likely to take public transportation or walk. When this bias was corrected, there was a cost savings—the number of pedestrian injuries, which women disproportionately suffered, declined.

This same bias recurs in city planning and the design of transportation systems. Public transportation designs facilitate trips to work and educational facilities. For example, London’s metro, like those of most other cities, resembles a “spider’s web” that is “incredibly useful for commuters who just want to get in and out of the center of town” (36). It is much more cumbersome for trips between neighborhoods, which women are more likely to take when they drop children at school or take elderly relatives to doctors’ appointments. Male-oriented public transportation systems make it difficult for women to perform this unpaid work and get to a place of employment on time. Because the data collected on the purpose of trips is biased, with all paid work placed in one category and care work divided into several categories, the resulting designs prioritize paid work. Barcelona has made efforts to make its transportation system accommodate “trip-chaining” (37), with more flexible bus routes.

Zoning laws, which segregate cities into commercial, residential, and industrial areas, often harm women as well. Because women do “three times the amount of unpaid care work men do” (40) globally, the geographical separation of homes from workplaces complicates women’s lives. When combined with poor transportation systems, the difficulties are magnified.

Brazil’s public housing scheme, launched in 2009, did not take into account female patterns of work, and was therefore disastrous. Residential zones were far from areas with jobs and from transportation. What is more, they were designed for nuclear families, which is not standard for the culture. As a result, many women had to drop out of the paid workforce and try to make a living from home via improvised haircutting salons and other illegal services. In contrast, when officials used appropriate data, Austria built housing more accommodating to women’s needs. Residences were near schools, medical facilities, and public transportation; and floor plans were designed with care needs in mind.

In public places, such as theatres, the lines for women’s restrooms are always long. This happens because standard designs, in an effort to be gender-neutral, give equal space to men’s and women’s rest rooms, despite the fact that urinals require less room than toilet stalls, and despite the fact that women typically take longer, as they are often accompanied by children or have to change sanitary products. Once again, gender neutrality does not yield equality.

What is worse, one-third of the world’s population lack access to toilets. Even in wealthy countries, field workers do not always have such access. Without safe and accessible toilets, women often wait hours to relieve themselves because they justifiably fear sexual assault when relieving themselves. In one township in South Africa, researchers estimated that there were 635 assaults due to a paucity of toilets, costing the area approximately 40 million dollars annually.

According to surveys, women experience far more fear of violence and harassment in public places than men do. Such fears affect women’s mobility. Instead of addressing this problem by redesigning transit services and city layouts, officials blame women for their fears and cite statistics of low crime rates. However, data collected about sexual harassment is woefully incomplete; even serious offenses, such as rape, often go unreported because there is a lack of proper reporting procedures. For instance, when women complain to bus drivers, their complaints are not taken seriously. Airlines also fail to arrest passengers who sexually harass women. Transit services and city planners need to admit that there is a problem of bias and providing things like better lighting, digital timetables for buses, and more security officers, to alleviate women’s fears.

Recreational spaces, such as parks and gyms, are designed with the universal male model in mind. Women and girls are hesitant to enter large open spaces, such as weight rooms and large fields, alone because often males congregate at their entrances or create a hostile environment within them. In Sweden, towns redesigned parks to have subdivisions and multiple entrances—changes that encouraged female participation in activities, contributing to better health and therefore public savings.

On October 24, 1975, 90% of women in Iceland went on a one-day strike, refusing to do any paid or unpaid work. Icelandic men called it the Long Friday. The next year, Iceland passed the Gender Equality Act; five years later, the country elected the first female head of state.

Globally, women perform 75% of unpaid work. As a result, women, with few exceptions, work longer hours than men. The increased workload negatively impacts women’s health: Women experience higher rates of stress, anxiety, and depression. Swedish studies have found that while moderate paid overtime work has a positive effect on men’s health, paid overtime work increases women’s hospitalization and mortality rates. Women typically have care responsibilities that do not go away with increased hours of paid work.

Because women often have care responsibilities, they need flexibility in their paid work schedules. Lacking this flexibility, they opt for part-time employment, which is less lucrative than full-time work. Women’s segregation into such work is not a choice: They simply cannot abdicate their responsibility to care for children and others. Gender pay gaps result; for example, women in the UK comprise 61% of those making less than a living wage, while data from multiple countries show that women earn between 31% and 75% less than men over their lifetimes.

Pensions reflect this difference too, with women more likely to face “extreme poverty in their old age” (77). Women are effectively penalized for taking time off for unpaid care work, early retirement, and living longer. Better maternity leave policies would help to alleviate these biases. While all industrialized countries except the US have paid maternity leave, most do not offer enough payment or length of leave. The US, one of only four countries with no paid leave, provides unpaid leave to just 60% of its workforce.

American universities discriminate against female professors via their “gender-blind leave policies” (82): Given only seven years to complete the publications necessary to earn tenure or be fired, only 44% of women with children earn tenure—compared to 70% of men with children. In countries such as Japan, a culture of extremely long work days—in which time spent in the office is equated with effectiveness—also harms women, whose care responsibilities put them at a disadvantage.

Gender-specific policies, versus neutral ones, are the best means to address these biases. One solution is paid paternity leave, which positively impacts female employment. For instance, Sweden’s parental leave policy, which requires fathers to use the leave or have the couple forfeit the time, generates the most use. Men who take paternity leave are more likely to be involved in subsequent childcare. Men’s participation in childcare could change workplaces that are typically not childcare-friendly.

Meritocracy is a very popular myth, especially in the US. According to a 2002 survey, over 90% of US firms use performance evaluations and merit-based pay despite strong evidence that these approaches do not work. Ironically, a belief in one’s personal objectivity tends to make one less objective and “more likely to behave in a sexist way” (94).

Male biases are particularly evident at tech companies, where 40% of women leave before ten years compared to 17% of men, and in academia, especially in STEM fields. Moreover, female students and academics are less likely than male ones to receive funding and to get jobs. Though papers written by women are accepted for publication more often under double-blind review—a practice that anonymizes author and reviewer—but this system of review is not universal.

Women professors are also expected to perform more unpaid and less prestigious work: Fearing that they will be penalized for saying no, they absorb more of the administrative burden and more workplace housekeeping, such as taking notes at meetings. Often asked to teach more, women have less time for research that wins prestige. Teaching evaluations are notoriously biased, with gender-inflected words such as “mean” and “annoying” (99) frequently employed against women, in contrast to male professors, who are more likely to be labeled “brilliant” or “genius” (100). Increasingly, students write aggressive and violent comments about female professors.

This brilliance bias results from a data gap. Female geniuses are not remembered because they have been erased from history and their accomplishments sometimes claimed by men. Even letters of recommendation, an important part of the hiring process, reflect gender bias. Men are more likely to be described as remarkable and outstanding, while women are depicted as kind and nurturing or a better fit for teaching than research.

Originally, computer coding was considered women’s work. Once programming gained prestige, however, women were pushed aside. Nowadays, the world of computer science is male. Men’s hacker brilliance is associated with young men’s traits like their single-minded devotion and willingness to stay up all night. Women might have less time to express their love of computers on social media; however, algorithms searching for tech employees often comb the social data of applicants. One tech company, Gild, gave higher scores to those visiting a particular website. Women, who have a greater burden of unpaid work, have less time to frequent social coding sites. In this case, the algorithm that benefitted males was exposed. Problematically, most such algorithms are secret.

These problems also affect promotions. For example, when given data showing that male employees were much more likely to nominate themselves for promotion than female ones, Google tried to change women or make their behavior in line with the male model—an inappropriate response to bias. However, there is much organizations can do much to alleviate the biases data reveals. One tactic that can help is blind recruitment systems. In one instance, women’s inclusion in the New York Philharmonic Orchestra increased dramatically when blind auditions were implemented.

At the workplace, the male standard prevails. The standard office temperature is comfortable for men, but too cold for women. Close parking spaces for pregnant women are absent. What is worse, occupational research has mainly focused on male-dominated industries, which have become safer workplaces as a result. In contrast, little is known about the health risks in female-dominated industries, such as care work where women have to lift people and then “go home to a second unpaid shift where there is more lifting” (115).

Research on work-related cancers, which kill 8,000 people annually, is biased toward men. Chemical absorption is different in men and women, yet studies done on men are presumed to apply to women. Endocrine disrupting chemicals (EDCs) are particularly harmful, even at low concentrations, and they are found in plastics, cosmetics, and cleaning products. Women working in auto-plastics or cosmetology are at high risk of exposure. Yet chemicals with a known link to breast cancer are poorly regulated in most countries. The cumulative effects of multiple exposures to different chemicals are understudied; such studies would benefit women, who are exposed to cleaning products in the home and other chemicals in the workplace. As a result of this data gap in occupational health research, women are dying unnecessarily.

Women who work in male-dominated jobs, such as farming, the military, and policing, are compelled to use equipment and products designed for the male standard. Standard sizes are an uncomfortable for women and tools do not quite fit their smaller hands. Even the too-long marching stride in armies produces pelvic stress fractures in women. Military and police uniforms—overalls and protective equipment—often make urination impossible for women, especially if working outdoors.

More problematically still, personal protective equipment for police officers and other personnel is designed for the male body, resulting not only in women’s discomfort, but also in their inadequate protection. In one case, a British police officer was stabbed and killed entering an apartment—she had removed her body armor, which made it too difficult to operate the ram she needed to break down a door. Perez attributes many of these problems to what she dubs the Henry Higgins Effect, after a sexist professor in the musical My Fair Lady who quips, “Why can’t a woman be more like a man?” (123).

In 2008, a US federal health agency announced the serious health dangers, such as cancer, of interacting with big bisphenol A (BPA), a chemical found in millions of consumer products. If researchers cared about the fate of the mainly female workers producing these items in factories, the dangers of this chemical could have been detected earlier.

Women are often the most vulnerable workers. Nail salons are an example of a female-dominated workplace with few protections. The technicians operate as independent contractors, allowing employers to pass financial risks onto them. When one manicurist splashed nail polish remover on a customer’s shoe, the owner compensated the customer with money deducted from the manicurist’s pay and fired her. This dismissal prompted the manicurist to comment that she was “worth less than a shoe” (131-32).

Women have been disproportionately impacted by anti-labor trends. One trend is precarious labor with few protections: Examples of this include short-term contracts and subcontracting through agencies. Also problematic are the algorithms the retail sector often uses to create last-minute scheduling to accommodate likely customer demand. Having to be available at such short notice is extremely difficult for women, who often have to arrange for childcare in advance. Even in the field of higher education, there is greater reliance on part-time workers paid low wages. Since women are more likely to hold these part-time positions, the trend exacerbates the income gap between men and women.

Given the widespread prevalence of sexual harassment, there is woefully inadequate data on the problem. Women in part-time and precarious positions are at greater risk of sexual harassment and have less recourse. For women who interact with the general public, “harassment all too often seems to spill over into violence” (138). Nurses, for example, are subjected to more acts of violence than police officers or prison guards, reporting daily threats. The design of traditional hospitals, with long hallways, isolates nurses and lessens the chance that calls for help will be heard. Their safety could be improved with circular wards and simple steps like signage.

Perez concludes that the design, location, regulatory standards, and hours of the modern workplace serve women poorly.