Public transport is usually evaluated in terms of mobility: faster commutes, lower travel costs, and better access to jobs, schools, and services. But in cities, transport infrastructure can also reshape the public spaces around it. A metro station does not simply move people from one place to another. It can change a neighbourhood by increasing footfall, attracting shops and street vendors, improving lighting, and making previously quiet stretches more active.These changes may matter especially for women. In Indian cities, women’s access to public space is often constrained by the risk of harassment. The problem is especially acute in Delhi – a city strongly associated with incidents such as the 2012 Nirbhaya case. Women may avoid certain routes or turn down employment and education opportunities because the journey itself feels unsafe (Borker 2021). Public transport may therefore shape not only how people move, but also whether women can use the city more freely.In recent research, I study whether the opening of metro stations in Delhi affected reported sexual harassment crimes against women in nearby areas. The Delhi Metro has expanded steadily since the early 2000s, and my analysis focuses on the wave of stations that opened between 2016 and 2019. This staggered roll-out allows me to compare areas before and after a station opened, using areas that had not yet received one as a comparison group. The main finding: reported sexual harassment rates – crime reports per capita in an area – fall by roughly 32% after a metro station opens in a police jurisdiction compared to a police jurisdiction where a metro station has not opened yet. The effect is concentrated within one kilometre of the station. Areas between one and two kilometres show some evidence of a small increase, suggesting highly localised effects and possible spatial spillovers.Studying the staggered expansion of the Delhi Metro A central challenge in studying urban infrastructure is that new projects are not randomly placed. Metro stations may be built in areas that are already changing economically, demographically, or politically. A simple comparison of places with and without stations may therefore be misleading.To address this, I use the staggered timing of metro station openings between 2016 and 2019 to compare changes in per capita reported sexual harassment crimes around newly opened stations to changes in areas where stations had not yet opened, using the Callaway and Sant’Anna event-study estimator. This method is built for settings where different units are treated (subjected to intervention) at different times and avoids the biases that arise when conventional approaches use already-treated areas as controls (Callaway and Sant’Anna 2021). Intuitively, the method asks: after a metro station opens in a locality, does per capita reported sexual harassment change relative to comparable localities where stations had not yet opened? This helps separate the station effect from broader city-wide changes in crime, policing, reporting, or mobility.The dataset combines three sources. Reported harassment comes from Delhi Police FIR (first information report) records, covering offences against women that range from sexual harassment and assault intended to outrage a woman’s modesty or disrobe her, to acid attacks. To capture pre-opening commercial activity, I use satellite measures of night-time light radiance, a standard proxy for local economic density. To track changes in the street environment itself, I draw on a crowd-sourced safety app, Safetipin.The results show a clear decline within one kilometre of newly opened metro stations, weakening with distance. Between one and two kilometres, I find a slight increase, indicating that the benefits of station-led activity may be spatially concentrated rather than evenly spread across the wider neighbourhood.Metro stations as economic activity hubs Why might metro stations reduce harassment nearby? The evidence points to a station- induced economic activity mechanism. Metro stations not only bring people, but also complementary forms of urban activity. As commuters pass through an area, shops open, hawkers gather, lighting improves, and the street grows busier. These changes can increase informal surveillance – the everyday monitoring that happens simply because more people are around.This mechanism echoes Jane Jacobs’ famous idea of “eyes on the street”. In her seminal work, Jacobs (1961) argued that lively streets, with a mix of users and activities, can generate informal public order: safety comes not only from formal policing but from streets that are active, visible, and socially monitored. A deserted stretch offers few witnesses; a busy market street raises the odds that a potential offender is seen, interrupted, or identified.My study provides several pieces of evidence consistent with this interpretation.First, the decline in per capita reported sexual harassment crimes is concentrated in areas where market activity was low before the metro station opened. If stations reduce harassment by creating active public spaces, the effect should be larger in places that were initially quieter or less commercially developed – and that is what the baseline night-lights data show, with the fall driven almost entirely by below-median stations.Second, the timing supports the market-activity channel. The decline is concentrated between 4 pm and 8 pm. This is not simply a peak commuting result. Delhi also has substantial commuting between 8 am and 12 noon, yet mornings show no comparable fall. The evening window is also when local market activity around station areas is especially high: shops, vendors, pedestrians, and commuters all overlap in public space. The decline during this evening market-activity window, but not during the morning commute, suggests that the effect is not driven only by metro use or changes in reporting behaviour. It is more consistent with an actual reduction in harassment risk when station areas become active commercial and pedestrian hubs.Third, independent data on the street environment line up with the crime results: drawing on a crowd-sourced safety app, Safetipin, I find that within one kilometre of new stations, users report improvements in lighting, the presence of hawkers, market activity, a more balanced presence of women alongside men in public space, and a stronger sense of security after opening.Reported crime and actual safety Since the study uses reported sexual harassment crimes, the results need to be interpreted carefully. Reported crime can change either because actual incidents change or because reporting changes. This distinction is especially important for crimes against women, which are often underreported.However, several patterns suggest that the results are not merely about reporting. If metro stations only made reporting easier, one might expect reported crime to increase, not decline. If the effect were driven only by commuter flows, one might expect similar patterns during the morning commute. Instead, the decline is concentrated during the evening period when market activity is highest. The effect is also strongest in areas with low pre-existing market activity and is accompanied by crowd-sourced reports of better lighting, more hawkers, and busier streets nearby. Crucially, that safety-app evidence comes from a source entirely independent of police records, yet points the same way.Taken together, the evidence suggests that metro stations reduce per capita reported sexual harassment crimes by changing the local public environment, not merely how it is recorded.This does not mean that the police data capture every dimension of women’s safety. Many incidents go unreported, and women may keep avoiding places even as recorded crime falls. But the results show that transport infrastructure can meaningfully reshape the geography of reported harassment.Policy implications The findings have several implications for urban policy.First, metro planning should be integrated with street-level planning. The area around a station is part of the transport system: sidewalks, lighting, street vending, last-mile connectivity, and safe walking routes determine whether people – especially women – can use the metro comfortably.Second, cities should recognise the safety role of legitimate street activity. Hawkers, small shops, and evening markets are often treated mainly as sources of congestion or disorder. But they can also create active, watched public spaces. The policy goal should not be to remove informal activity from station areas, but to manage it well – clear pedestrian paths, adequate lighting, sanitation, and rules that keep public spaces accessible and active.Third, spatial spillovers need attention. The slight increase in per capita reported sexual harassment crimes between one and two kilometres from stations suggests that improvements should not stop at the immediate station precinct. Safer, well-lit, active walking routes need to extend into surrounding neighbourhoods.Finally, women’s safety should be part of how cities evaluate transport investments. Metro projects are usually assessed based on ridership, time savings, reduced congestion, and environmental benefits. If stations can make nearby public spaces safer for women, that too is part of their urban value.Beyond mobility The Delhi Metro has transformed mobility across the city. These findings suggest that its effects extend beyond movement. By creating dense pockets of footfall, lighting, vending, and market activity, metro stations can reshape nearby public spaces and reduce per capita reported sexual harassment crimes against women.The broader lesson is that safer cities are not produced by policing alone. They are also produced through everyday urban life – people walking, vendors selling, shops staying open, streets being lit and used. This is precisely the kind of urban vitality that Jacobs saw as central to safe streets.For women, the right to the city depends not only on the availability of transport but also on whether the spaces around it are safe. Delhi’s experience suggests that when metro stations become active market hubs, they can make nearby public spaces safer. Urban policy should build on this potential.This article first appeared on Ideas for India and has been republished under an arrangement.