An elephant in the bus:
violence against women

The inclusive city is one of the pillars of the UN Sustainable Development Goals: discrimination based on gender should be eradicated. But cities are not gender-neutral — gender significantly shapes the experience of moving around a city. Although women represent roughly half the world's population, they do not enjoy the same freedom of movement as men, which limits their participation in economic and social life.

Public transport (PT) is the backbone of sustainable urban mobility, yet it often fails to serve women equally. Safety is one of the main barriers: women feel less safe than men at nearly every stage of a PT trip — on board, and while waiting at the station — and that feeling of insecurity shapes short- and long-term travel decisions, from daily mode choice to where to live and work.

This insecurity is deeply linked to gender-based violence, particularly sexual harassment, which is the most common and visible form of violence encountered in transit settings. This study uses survey data from Cairo, Egypt — the largest metropolitan area in the MENA region, home to roughly 23 million people — to understand the dimensions of GBV in transit and how it shapes women's travel decisions, and to surface what women themselves propose to improve safety.

The MENA region is one of the least-studied areas for GBV in public transport in the current literature — despite violence against women in public spaces being widely observed and rarely discussed. That contradiction gives the paper its title.

Women's travel needs differ from men's, shaped by their disproportionate role as primary caregivers. Their trips are more complex — chaining home, work, childcare, and shopping together — rather than the direct commute more typical of men. Across cities as different as Auckland, Dublin, Hanoi, Helsinki, and Manila, women make shorter trips, avoid peak-hour and nighttime travel, and have less access to private cars — a constraint observed in both developing contexts (India, China) and wealthy ones (the US, Germany, Sweden, Finland).

Yet women are not a homogeneous group. Low-income women in Dhaka face a distinct mix of infrastructural gaps and conservative social taboos; young ethnic-minority women in Auckland navigate "stranger danger" by pre-planning night trips and performing confidence; riders wearing visible religious dress report heightened discrimination. Treating "women" as a single vulnerable category obscures these intersecting burdens.

GBV is any act of violence against someone because of their gender — physical, sexual, psychological, economic, or emotional. On transit, it ranges across a continuum, from sexist jokes and staring to groping and, in rare but real cases, assault. Researchers typically group harassment acts into three families:

The pattern repeats worldwide, at different intensities: up to 97% of women in São Paulo report harassment; rates sit between 50–80% in Mexico City, Bogotá, and Santiago; 78% in Karachi and over 85% in Nairobi. Most women respond by staying quiet or walking away rather than confronting the harasser — and the incidents are massively under-reported everywhere, from 32% saying nothing in London to 93% not reporting at all in Manizales, Colombia.

The table below distills 18 studies from around the world — the same one this paper builds on to show that Cairo's numbers, while stark, are not an outlier.

Existing research is geographically skewed toward major cities outside the MENA region, and GBV in transit is usually studied in isolation from broader urban planning. No prior study brings together timing, perpetrator identity, victim response, emotional aftermath, and the survivor's own proposed fix in one unified framework — which is exactly the gap this paper sets out to close.

Cairo is the capital of Egypt and the largest city in the MENA region — about 10.3 million residents within the city, 48% of them women, inside a wider metropolitan area of 23.1 million. Its transport system leans heavily on informal paratransit: privately-owned 14-seat microbuses that carry over half of all PT trips, alongside Africa's oldest metro system and a public bus fleet that is, per capita, among the smallest in the region.

The research team ran a five-part online survey (via LimeSurvey, in Arabic and English) from January to April 2021, distributed through Facebook, LinkedIn, and WhatsApp. Only women, 18 or older, residing in Cairo and using PT, were eligible.

Because the survey was long, attrition was high — but a dedicated attrition check found dropouts' early answers tracked closely with those who finished, so each section's respondents were analyzed as-is (the Retained Sample) rather than discarded. The final complete sample (N = 227) skews young — 78% are 18–24 and 88% single — and more educated than the general population (65% hold a bachelor's degree or higher), which the authors are transparent about as a limitation.

An open-ended question let respondents describe GBV in their own words. Those answers were clustered using Latent Dirichlet Allocation (LDA) topic modeling, tuned via coherence, log-likelihood, and perplexity checks to four topics.

Respondents were regular riders — about 60% use PT at least twice a week. Most are captive riders: two-thirds cite "no other option" as their reason for using PT, rather than cost or convenience. Only about 1 in 5 owns a car.

Of 11 possible barriers, safety at the station and safety on board ranked first and second — ahead of vehicle condition, transfers, and trip duration. Stations, often less crowded than the vehicle itself, felt less safe than being on board — a pattern also documented in the UK, Australia, Sweden, Italy, Japan, and France.

Trip purpose matters: mandatory trips like commuting or school were rarely canceled over safety concerns, but leisure and shopping trips were likely to be cancelled by nearly 60% of respondents — a quieter, cumulative form of exclusion from everyday life.

Only 5% of respondents had never experienced harassment on PT. About 1 in 7 face it daily.

The five most common unwanted behaviors, each reported by at least half of respondents, were all verbal or nonverbal: staring, catcalling, sexual comments, suggestive facial expressions, and offensive language. Physical acts — blocking the way, invasion of privacy — were reported less often but still real. Most incidents cluster late at night and in the early morning (9pm–7am), when demand — and the presence of other passengers — is lowest.

Faced with these acts, women mostly did not confront the person responsible: across almost every behavior, not reacting or walking away are the two most common responses, together accounting for roughly two-thirds to three-quarters of answers. Verbal retaliation is the most common form of pushback — highest for the everyday, high-frequency acts like offensive language (32.2%) and catcalling (31.4%) — while asking a fellow passenger or the driver for help is rare almost everywhere. That's less a sign of passivity than a coping strategy shaped by fear, distrust of authorities, and a lack of visible support.

Asked to describe their single worst experience, 42% of respondents pointed to a physical act, 35% to nonverbal harassment, and 14% to verbal abuse — a reversal from day-to-day harassment, where verbal and nonverbal acts dominate. Most incidents (71%) happened in an averagely crowded setting, and in 16% of cases the harasser was a transit worker — a driver or conductor — rather than a fellow passenger.

The reasons cluster into three layers: practical (the incident was over too fast to act, or there was no witness), psychological (fear of not being believed, shock, fear of being blamed), and systemic (distrust that authorities would investigate, no clear way to report, weak legal protection).

The most severe incidents skew strikingly young: the average age at the time of the incident was 18, with the distribution concentrated in the teens and early twenties and a long tail stretching out to women in their thirties — meaning many of these "most severe" experiences happened while respondents were still girls, not adult women.

Asked how the most severe incident made them feel, respondents could select multiple emotions at once — and the top responses were unambiguous. Outrage (59.8%) and terror (55%) led, followed closely by confusion (45.4%) and feeling abused (43.6%). Fewer respondents described dissociation or minimization, suggesting most women recognized these incidents as serious violations in the moment, rather than downplaying them internally — even as nearly all stayed silent about it afterward.

Given 12 possible interventions, support was high across the board — but visible, responsive infrastructure led the list.

An open-ended question let respondents describe GBV freely. Clustering the answers with topic modeling surfaced four recurring themes:

GBV on transit is a universal phenomenon with a strikingly consistent shape from Cairo to São Paulo to London: most women face it, most stay quiet, and most keep riding anyway because they have no real alternative. That combination — high exposure, low reporting, low exit options — is what turns a safety problem into a driver of social exclusion: women skip leisure and shopping trips, withdraw from public life, and absorb the cost in time, money, and mental health.

The authors frame their recommendations in two tiers — what can move quickly, and what has to be sustained over years.

What travels across borders less well is the specific fix. Cairo respondents strongly favored surveillance cameras — a measure Australian riders view with skepticism. French riders lean toward personal tech, like smartphone alarms. Bangladeshi respondents pushed harder for systemic change: stricter enforcement, more vehicles, women-only carriages. The problem is universal; the right intervention is local.

The survey's length drove significant drop-off — but an attrition check found dropouts' early answers tracked the completers closely, so the final pattern of results is unlikely to be an artifact of who stuck around. The complete sample (N = 227) is modest and skews young and well-educated, in line with comparable studies elsewhere (Dhaka N = 225, Manizales N = 296, Lucknow N = 200). Because recruitment ran entirely online, older and less digitally-connected women are under-represented; face-to-face interviews were judged likely to introduce their own bias, given how private these incidents are. The analysis is intentionally descriptive rather than causal — appropriate for establishing a first baseline in a region this under-studied, but a clear invitation for future modeling work.

GBV on public transport is close to universal — nearly all women experience it, yet almost no one intervenes seriously to stop it. What stands out in this data is how similar the pattern looks whether the city is rich or poor, dense or sprawling: cities everywhere are still not designed around the needs of half their population.

That similarity is also the opportunity. A shared underlying pattern of need means a shared policy foundation is possible — one that treats safety and dignity as non-negotiable parts of public transport design, built from what women who actually ride the system say they need, rather than imposed from the top down.