Missing Women
In Delhi, women made up about one in five pedestrians in our street photographs. Why so few? We counted people on the street, then used time-use surveys to see how often women and men left home.
The name comes from Amartya Sen’s 1990 essay. Sen asked how many more women would have been alive had they received the same care as men. My work starts with a related question about public life. Given how many women live in a city, how many do we see on its streets?
Who is out on the street?
In 2016 and 2017, I asked people to photograph 196 locations in Delhi. Women accounted for 18.5% of the people counted in the photographs. The photographers departed from the sampling plan, which made it hard to say how well those streets represented Delhi.
Varun Karekurve-Ramachandra and I later sampled roads in four Indian cities and sent motorcyclists along them with chest-mounted cameras. Would we get the same answer from photographs taken by someone else? We could check in three cities where we also had Google Street View images.
| City | Bike cameras | Street View |
|---|---|---|
| Delhi | 19.8% | 20.0% |
| Mumbai | 19.1% | 20.5% |
| Navi Mumbai | 18.0% | 24.8% |
The two sources agree closely in Delhi and Mumbai. In Navi Mumbai, the answer depends more on which photographs we use. India’s time-use survey offers a clue to the wider pattern. Women were less likely than men to leave home on the survey day. Even among those who went out, women made fewer trips.
How does the picture change across countries?
I tried two ways of extending the counts. Starting with 28,794 photographs contributed by the public in 101 cities, I found that women made up 36.0% of pedestrian sightings a vision model could classify. But contributors choose where to take pictures. In a second study, I sampled streets and walking routes in 98 cities before looking for images. Google returned no panoramas in 16 cities. In the images obtained, women accounted for 28.8% of classified pedestrian sightings. The model could not classify 66% of all sightings, including riders.
Sometimes the question is what people are doing once they are out. In 30 Delhi street-food videos, women accounted for 13.5% of apparent diner sightings that could be classified, giving each video equal weight. That describes whom the filmmakers showed eating.
Who appears in the record?
A different way to count is to ask whose names people write down. In India, electoral rolls let us compare the numbers of registered women and men who share a surname. Members of Parliament’s biographies let us count the sons and daughters they report. In each case, the people omitted from a record matter to its interpretation.
Sometimes the choice of whom to name is itself the question. Walter Guillioli and I searched US business registries for the words “son” or “sons” and “daughter” or “daughters.” Across the states we searched, the median ratio was twelve to one. Epic Children asks whom a family story preserves by counting the sons and daughters attributed to figures in epic and mythological traditions.
Studies
Streets and public life
- Women and Men on Urban Streets: Evidence from Crowdsourced Photographs in 101 Cities
- Measuring Women’s and Men’s Presence on Urban Streets: Evidence from Google Street View in 98 Cities
- Missing women on Indian streets
- Who Eats in Delhi Street-Food Videos?
- Women and Men on Delhi’s Streets: Evidence from Field Photographs, 2016–2017