New Delhi: Over 70% of women in Mumbai and Navi Mumbai are not visible on the cities’ streets, a study on gender disparities in public spaces has found. The study, conducted by researchers Varun Karekurve-Ramachandra and Gaurav Sood, is based on a visual audit of over 900 kms of roads in Mumbai and Navi Mumbai region.The study found that women accounted for 16.4% of visible people in Mumbai and 14.7% in Navi Mumbai. In terms of sex ratios, women accounted for 239 (Mumbai) and 223 (Navi Mumbai) pedestrians per 1,000 men. These numbers are far below women’s population share in the two cities. The residential census ratio of women is 838 and 910 per 1,000 men in Mumbai and Navi Mumbai respectively.The study relied on a visual audit of the roads based on 4000+ street images that were manually annotated. Trained human annotators – who reviewed photos and assigned necessary labels in the images – classified over 23,000 visible individual observations, recording gender*, transport mode, and streetscape features such as vendor presence, street markings, and road surface quality, the study said.The pattern was consistent irrespective of road types – primary, secondary, tertiary, and residential – with women’s share being low on all of them, the study found.The study also made a distinction between women’s share in pedestrian and vehicular traffic. When restricted to pedestrians, women constitute 19.3% of people on Mumbai roads and 18.2% in Navi Mumbai. On the other hand, women constituted just 8.4% of two-wheeler users in Mumbai and 5.7% in Navi Mumbai – significantly lower than their pedestrian counterparts.The researchers, while admitting the limitations of the study, said that women travelling inside enclosed vehicles like cars and public transport like buses and urban rail could not be counted as they weren’t visible in the gathered images. What this data meansWhile the study does not comment on why we lack female presence on roads, the researchers noted the implications of this observation. “When women cannot move freely in a city, the consequences can ripple through labor markets, civic life, and household bargaining. Urban streets are where these constraints become physically visible: a space can be crowded and still exclusionary if the crowd is overwhelmingly male,” the study noted.“Cultural and societal norms that treat unaccompanied female presence in public as suspicious or improper are well documented in Indian cities. Fear of harassment constrains women’s route choices, timing, and willingness to travel alone; in Delhi, female college students accept longer commutes to avoid streets perceived as unsafe. Infrastructural gaps – inadequate public sanitation, poor lighting, missing or obstructed sidewalks – may raise the cost of public presence disproportionately for women. Moreover, unequal distribution of household labour limits discretionary travel: women who spend more hours on domestic work have fewer hours available for the street. These forces likely interact,” it added.The study also found that women’s presence was elevated near street vendors and bus stations – locations anchored by practical errands and mixed-use activity. This finding was consistent with other studies that found that women were “permitted to traverse public space for work or errands but not to linger without purpose,” the study said.How women make trips around the worldIn her book, Invisible Women: Exposing Data Bias in a World Designed for Men, Caroline Criado-Perez wrote that despite the absence of sex-disaggregated data from every country, it was clear that women were invariably more likely than men to walk and take public transport while men around the world were more likely to drive. A key difference was also observed in the commute pattern of men and women. “Men are most likely to have a fairly simple travel pattern: a twice-daily commute in and out of town. But women’s travel patterns tend to be more complicated. Women do 75% of the world’s unpaid care work and this affects their travel needs. A typical female travel pattern involves, for example, dropping children off at school before going to work; taking an elderly relative to the doctor and doing the grocery shopping on the way home. This is called ‘trip-chaining’, a travel pattern of several small interconnected trips that has been observed in women around the world,” Perez wrote.When urban planning is centred around private cars instead of prioritising pedestrians and public transportation, it implicitly benefits men more than women. Perez attributes this to male bias as transport as a sector is highly male-dominated. In the case of India, infrastructural limitations like these are compounded by societal norms that work in tandem to keep women at home. Why put a number on this?“In 1990, Amartya Sen estimated that over 100 million women were missing from the world’s population, women who would have been alive but for sex-selective abortion and neglect. Sen counted lives cut short. Our missing women are alive (conditional on the sex ratio). What we are counting is a different loss: lives lived largely indoors, narrowed by fewer opportunities outside the home and heavier burdens within it,” Sood told The Wire.“Knowing that women are outnumbered five to one, in the country’s most cosmopolitan city, in broad daylight, is hard to shrug off. It makes us introspect about what kind of society we want to build,” he added.Even more surprising is the fact that Delhi has reported similar numbers in a pilot study, despite being considered far more unsafe than Mumbai, he said.A possible explanation for the lack of women on roads could be that data from enclosed vehicles was not captured in the study. However, Ramachandra pointed out something entirely different. “In India this argument doesn’t work: car ownership is low, the two-wheeler is the private vehicle and two-wheelers are fully visible in our footage. Women’s share there is not higher than on foot but far lower, roughly one in five pedestrians against fewer than one in ten riders. Women grow scarcer as the mode grows more independent. Who in a household gets the vehicle, who gets licensed, what riding alone costs a woman socially – all these are questions transport departments own and have rarely been asked,” he said.What can governments do?“Measure,” said Ramachandra. He said that while cities spend heavily on lighting, policing and other infrastructure for women’s safety, it rarely ever checks if an intervention has worked. “These audits cost little and repeat easily. Run one before an intervention and one after, and the spending becomes evaluated policy instead of hopeful policy. The data also comes with a built-in priority list,” he said. The data collected by Ramachandra and Sood showed corridors with female presence ranging between 7-30%. “A city wondering where to begin already knows,” he added.Ramachandra pointed to four other areas where the government could focus its data collection efforts. First, measuring the skew at night and in enclosed spaces. “Our cameras ran in daylight, and everything qualitative research tells us says the skew worsens after dark, which is exactly where the policy debate about women’s safety lives,” he said.Second, the perceived safety of specific locations. He explained that the area of focus needs to shrink from city-level perception to specific streets and hours of the day. “That matters because perception may feed on itself. A woman judges a street partly by whether other women are on it; if none are, she stays away, and the street stays empty of women. If that loop is real, it changes what an intervention has to do. It doesn’t just have to make a street safe, it has to break an equilibrium,” Ramachandra pointed out. For instance, a service road connecting a metro station to a primary road would be considered safe to use in the day by women, but deemed unsafe if it isn’t well-lit at night. “Our presence estimates, matched with location-level perception data, would let someone test this directly: Do women avoid a street because something bad happens there, or because no other women are there?” Ramachandra said.The third area is time-use and trip-purpose data along with locations. “Some of the gap has nothing to do with the street. A woman with no job outside the home and six hours of housework has neither reason nor time to be out, and no streetlight fixes that. India’s time-use surveys already document the hours; what’s missing is the geography,” Ramachandra said.The last area, and perhaps the most challenging, according to Ramachandra, is the societal expectation that governs how women occupy public spaces. This includes questions such as whether women are expected to have a reason to be out, if they are expected to be accompanied at all times, which neighbourhoods are considered safe to visit, and under what circumstances.This inevitably brings us to the question of leisure. Are women only occupying the streets to travel to work or run errands? In their book, Why Loiter?: Women and Risk on Mumbai Streets, Shilpa Phadke, Sameera Khan, and Shilpa Ranade, answer this question and raise many more. “Our data can’t test this directly, a photograph doesn’t reveal why anyone is standing where they are, but what we see is consistent with it. Women are scarce everywhere on these streets, and slightly less scarce in the places anchored by errands,” Ramachandra said. *Gender was inferred from visible appearance – clothing, hair, and build as they present in street-level photographs. This necessarily imposes a binary classification and cannot capture the full complexity of gender identity.