Main

Bicycling is widely promoted as a strategy for reducing the climate impacts of transport and for improving population health owing to reduced emissions and physical activity benefits1. The Intergovernmental Panel on Climate Change (IPCC) promotes bicycling as a key strategy to mitigate transport emissions2 and the World Health Organization (WHO) lists bicycling infrastructure as a ‘best buy’ for tackling noncommunicable diseases3. Air pollution and physical inactivity are leading risk factors for noncommunicable diseases in low- and middle-income countries (LMICs)4,5,6. Nevertheless, automobiles continue to increase and remain aspirational in the cities of LMICs, with limited understanding of pathways to increase bicycling. Through mixed-methods research in four LMIC cities, we show that bicycling cultures persist in LMIC cities primarily among marginalized communities despite limited infrastructure or policy attention, and we examine the relational experiences of bicyclists that reveal the inequities underlying the expanding automobile-centric transport systems. Viewed through a kinopolitical lens, these bicycling cultures reflect deeper social, economic and political dynamics. The locally embedded experiences and relationships of bicyclists offer valuable insights and resources for building more equitable and sustainable mobility systems that prioritize accessibility, safety and social inclusion for all.

Mimi Sheller’s work on mobility justice emphasizes that equitable and sustainable transitions in mobility systems should attend to the relational and multiscalar kinopolitics within which the contemporary urban and climate crisis are embedded7. Some cities in high-income countries (HICs) have made large investments in bicycling infrastructures that have shown successful increased bicycling8,9, decreased car use, reduced vehicular emissions and improved air quality10,11,12,13. Following such examples, many LMICs are increasingly investing in bicycling infrastructures14,15. While HICs, including those with low cycling rates, negotiate transitions from peak car-centric automobile cultures, LMICs face highly mixed-traffic roads and expanding highway development, alongside pre-existing transportation equity policies and non-elite bicycling cultures.

Literature documents an exclusionary history of bicycling in the cities of HICs, where built environment, policy and cycling infrastructures have contributed to new forms of social exclusion on the roads7,16,17,18. Intersecting factors such as race and socioeconomic status19 or age and migration status20 further amplify these barriers to bicycling. However, it is notable that in LMICs, bicycles have also been used to equalize transportation through financial and manufacturing infrastructures, particularly to make bicycles affordable. For example, bicycle distribution schemes have aimed to improve the access of girls to bicycles as part of female empowerment efforts, with growing evidence of their success21,22 in a larger, highly gender unequal transportation landscape in India23.

Jonathan Anjaria argues that recent urban bicycling-oriented policies in LMICs are primarily using built infrastructure-focused logics from HICs, with limited attention to the experiences of their existing bicyclists and the local context24. Taking this critique seriously, we conducted intercept surveys, semi-structured interviews and ethnographic observations in New Delhi and Chennai (India), Dhaka (Bangladesh) and Accra (Ghana). Through mixed-methods research, we ask: “Who bicycles in the selected cities? How do they experience and perceive bicycling? What barriers to bicycling do they face?” We synthesize these findings to describe the infrastructures25, including physical, financial and human, that shape bicycling in LMIC cities. Through bicyclists’ relational experiences and anxieties, we trace infrastructures that engage in ‘legitimations of speed’ and those that challenge the ‘unjust power relations of uneven mobility’7, and identify pathways to foster increases in bicycling.

Context

Delhi and Chennai (India), Dhaka (Bangladesh) and Ghana (Accra) are rapidly densifying and expanding cities, with flat terrains and heterogeneous traffic. They share subtropical to tropical climates with very hot summer days and seasonal flooding. Globally, physical inactivity accounts for 7% of all deaths, with the bulk (69% of deaths) occurring in LMICs such as India, Bangladesh and Ghana4. Similarly, 12% of global mortality and 19% of cardiovascular mortality are attributable to air pollution5,6. The cities share a high burden on people’s health from transport-related pollution26, noncommunicable diseases5,27 and road traffic injuries (Extended Data Fig. 1). A large shift toward nonmotorized transport, especially bicycling, can help countries toward longer-term climate goals. For example, rapid growth in the motor vehicle fleet (600% in the past two decades) has made transport the fastest-growing contributor to greenhouse gas emissions in India, and transport now accounts for nearly 30% of emissions in the major cities (Extended Data Fig. 1). Among the transport sector options for meeting India’s Nationally Defined Contributions, the largest reductions will be from shifts to low-carbon transport such as walking and bicycling28.

Urban planning and strategy documents and discourses in these cities mention bicycling. However, infrastructures to increase bicycling (bicycle-sharing systems and separated lanes) and improve equitable mobility (bicycle distribution through government schemes and philanthropic efforts) are conducted with little data on the mode and its user experience. While household bicycle ownership data are available in India in general, limited data are available for the four cities on bicyclist trips, demographics and road use. In this context, we conducted intercept surveys with bicycle users on arterial roads, semi-structured interviews with bicyclists and stakeholders, and ethnographic field observations in diverse locations. Our research explores pathways to support increases in bicycling in such contexts.

Results

Who bicycles in Delhi, Chennai, Accra and Dhaka?

Socioeconomic status

Bicyclists surveyed on arterial roads in Delhi, Dhaka and Accra were largely poor, adult and male. Except for one woman in Dhaka, all survey respondents were men, predominantly young adults to middle aged, with a slightly older distribution in Delhi (Table 1). They were largely from lower-income households in the four cities and had jobs involving substantial physical activity (Table 2). Most respondents’ households owned no other vehicle (Extended Data Fig. 2). Interviewees had a strong tendency to associate bicycling with either poverty or childhood. They noted that once people can afford automobiles, bicycles are not a preferred mode of transportation unless one undertakes bicycling for physical exercise (Extended Data Fig. 3, Q1–2).

Table 1 Bicycle use data highlights from intercept surveys
Full size table
Table 2 Occupation (percentage) and bicycle use characteristics of intercept survey respondents
Full size table

Bicycle retailers and activists have emphasized the need for more automobile users to bicycle and for infrastructure to support this transition. Some bicyclists noted an increasing presence of elite bicyclists (Extended Data Fig. 3, Q3), although our surveys and observations showed that they currently comprise a very small proportion of the bicycling population. We observed high volumes of bicycling geographically concentrated in several corridors of the cities at selected times. Notably, policy actors who did not directly work on bicycling typically held the view that there was hardly any bicycling in the cities these days. This invisibility of bicyclists among decision-makers was repeatedly noted by bicyclists and advocates (Extended Data Fig. 3, Q4). The ‘common sense’ view that nobody bicycles was also evident in the limited discussion of bicycling in planning documents, where bicycling was typically mentioned as recreational infrastructure in urban beautification and development projects, and increasingly as active mobility. Sustainability policy documents largely focused on electric vehicles and public transit, with little attention or intentional budgetary allocation to bicycling infrastructure. The school bicycle distribution scheme was a notable exception, enabling adolescents from low-income families to access education. While this policy was implemented in Chennai and several other regions in India21,22, all four research cities had philanthropies distributing bicycles for socioeconomic access.

Age, gender and care work

Bicycle use varied strongly by age and gender. Children largely cycled within traffic-restricted residential pockets and buildings. The few children we observed on arterial roads were probably of adolescent age, male and were more common in Accra compared to Delhi, Chennai and Dhaka. Interviewees whose adolescent children commuted by bicycle explained that they were either caregivers or needed to travel independently owing to the lack of adult caregivers. Women on bicycles, often as pillion riders, were slightly more common on arterial roads than children. We saw more women cyclists on arterial roads in Chennai than in other cities. A few women cyclists were observed in Delhi and Dhaka on arterial roads, and none in Accra. Some male bicyclists in Delhi expressed reservations about women in their families bicycling. Alternatively, several interviewees in Delhi, Chennai and Dhaka mentioned that their daughters or granddaughters bicycled but not on larger roads.

We interviewed women bicyclists in residential neighborhoods where domestic workers’ bicycle use was particularly common. They described a hierarchy of vehicles between cycles and automobiles (Extended Data Fig. 3, Q6). Just as automobiles were preferred by higher socioeconomic classes, in households that had both bicycles and motorcycles, men used motorcycles and mentioned traveling longer distances compared to women who used bicycles.

Women interviewees mentioned avoiding arterials and bicycling shorter distances (~2 km) with multiple stops for work and other reasons. By contrast, men typically reported a long single ride to work, even if they trip-chained close to their destination (Table 2). In Chennai and Delhi, women’s travel to work schedules were later in the morning than those of men, who typically reported early morning travel. We observed several women caregivers and students using cycles to carry younger students or multiple school bags (Extended Data Fig. 3, Q7). Many female bicyclists with pillion children and bags dismounted and walked their bicycles when traversing arterial road stretches and resumed bicycling on smaller roads.

Migration status intersected with the gendered experience of urban bicycling. Women bicyclists in Delhi often described migrating from states with active school bicycle distribution schemes, such as West Bengal. A female bicyclist in Dhaka illustrated the complexity, noting that while women’s cycling continued to be a taboo in certain regions, people found workarounds. After moving to the city from her village, she had learned to bicycle as an adult from her younger sister, who had a particular passion for it (Extended Data Fig. 3, Q5). Migration also shaped men’s bicycling. Many survey respondents described learning to bicycle in locations with fewer and slower automobiles on the road compared to their current routes, which they considered risky for their children and younger kin.

Safety and comfort

Sharing roads with automobiles and experiencing risk

Bicyclists expressed that arterial roads were unsafe owing to fast-moving vehicular traffic, influencing who can and prefers to ride on such roads. While adult male bicyclists saw themselves as capable of traveling cautiously in high-speed traffic, irrespective of socioeconomic class, they tended to focus on safety risks when describing their younger kin or children riding bicycles (Extended Data Fig. 4, Q8). Many bicyclists and parents placed the responsibility to navigate risks on mixed-traffic roads on individuals, describing children, adolescents and the elderly as incapable of sufficient caution. They described near misses as a routine part of the cycling experience and recollecting minor crashes as a step in learning to bicycle. Crash-related expenditures to repair bicycles or damage to the goods they carried were common.

In describing the dangers associated with bicycling, some interviewees perceived bicycles as particularly unsuitable for persons who might face difficulty recovering from injuries. An interviewee expressed her concern about her diabetic teenage grandson’s safety owing to ‘crazy cycling’ by his peers, worried about his slower healing time if injured. Echoing parents’ anxieties, elderly interviewees—who also mentioned healing slowly—discussed safety and comfort as main reasons why their adult children dissuaded them from bicycling, even when they perceived that bicycling enhanced their well-being (Extended Data Fig. 4, Q10).

While elite and middle-class bicyclists and bicycling activists attributed crashes to automobile drivers or infrastructure, most others placed the onus of safety on the cyclists themselves (Extended Data Fig. 4, Q9). Many elite and middle-class bicyclists who participated in group rides also tended to describe bicycling as a social activity that offered relaxation, ideally outside urban traffic or during early mornings (Extended Data Fig. 4, Q11), while they continued to use motorcycles or cars within the city and to navigate peak automobile traffic.

Navigating mixed traffic and urban built environment

We often observed bicyclists interrupted by infrastructure primarily intended to support unimpeded vehicular flow (for example, foot overbridges, flyovers and medians). “[One] can’t cycle there. Now they have fully locked it. Earlier, [you] could easily ride through”, explained a mother who had walked her bike with her young son on the carrier across a busy arterial in Chennai and had climbed a newly built foot-high median to get to the other side.

Medians, U-turns, crossings, intersections and roundabouts, where bicyclists cannot avoid interactions with motorized vehicles by just ‘riding on the side’ or ‘staying in their lane’, were considered particularly risky by parents and adult bicyclists. Bicyclists either waited at pedestrian crossings or grouped to establish their presence before crossing the roads or navigating turns. Infrastructural interruptions increased both the effort and distance for bicyclists. Bicyclists, including adolescent school children, pedaled or walked their cycles up flyovers alongside buses, trucks, cars, motorized three-wheelers and motorcycles. Where pedestrian infrastructures such as overpasses existed, we saw bicyclists carry their bicycles up the stairs. It was also common for bicyclists to subvert such autocentric road rationales by moving counterflow to automobile traffic on the main carriageway, traversing median barriers with their bicycle or holding on to slow-moving automobiles such as autorickshaws to be towed up flyovers (Extended Data Fig. 4, Q14).

While such infrastructural features were a source of anxiety for bicyclists and their kin, they particularly added to the exertion of adult male bicyclists who cycled relatively longer stretches on the ‘main roads’ often designed and maintained by highway authorities for automobiles’ uninterrupted long-distance flow. An urban development official in Dhaka described the lack of clarity on which authority is responsible for supporting increasing bicycling (Extended Data Fig. 4, Q12). We observed this ambiguity across state and metropolitan planning guidelines and budgets.

Bicyclists undertook strategic modifications to make their bicycles more visible in such mixed-traffic roads. Most bicycles we encountered did not have lights and reflectors, and their manufacturer-provided bells got drowned out in the constant automobile honking common in these cities. Bicyclists in Delhi repurposed and used removable road reflectors, which were brighter and more robust than those sold for bicycles. Some bicyclists in Dhaka and Chennai affixed headlights and motorcycle side mirrors that doubled as reflectors. In Dhaka (but not in the other cities), battery-powered horns were used to circumvent this issue (Extended Data Fig. 4, Q13). Such modifications may be read as subversions of automobility through its tools and how bicyclists establish their belonging toward the roads and demonstrate their fluid position between pedestrians and automobiles.

Urban housing geography also shaped bicycling. Adult male survey respondents traveled by cycles to, from and during work for long distances and durations, with the longest commute times in Delhi (Table 1). Respondents often lived far from their work locations, probably because of cheaper housing in the urban periphery, where some had been relocated to resettlement colonies as part of government programs to remove slums near affluent areas29.

People bicycled despite these hostile conditions as bicycles were relatively easier to afford and access compared to motor vehicles, especially on routes with limited public transport. The status of bicycles as nonmotorized vehicles, exempt from registration and licensing requirements, reduces barriers to their use and offers unique opportunities for riders on mixed-traffic roads. In Delhi, Dhaka and Chennai, they were also not penalized for riding on footpaths, counterflow to traffic or turning at junctions in ways that are prohibited for motor vehicles. Bicycles also offered benefits in a congested city and were supported in distinctly local ways. Traffic police often took cognizance of bicyclists in Dhaka at signals and allowed them to move ahead of motorized traffic (Extended Data Fig. 4, Q15).

Comfort and bicycle lanes

Being exposed to weather conditions and broken road surface conditions were matters of everyday concern. The respondents and interviewees in all four cities described traveling, when possible, early during the day and late in the evening to avoid the harsh sun. Stakeholders saw escaping harsh weather through automobiles as a promise that urban class mobility offered (Extended Data Fig. 4, Q16). However, we witnessed bicyclists, albeit fewer, out during the rain in Delhi, Dhaka and Chennai, who used raincoats, hand-held umbrellas and single-use plastic bags as rain cover. We also encountered bicyclists using the road during extreme heat, including two heatwaves each in Delhi and Dhaka, and flooding in Dhaka during our fieldwork.

Interviewees described segregated pathways with shade as a desirable convenience for bicycling if they were maintained and regulated properly. A public sector technician in Delhi compared a bicycle track in a different city to a segregated pedestrian pathway or ‘footpath’ (Extended Data Fig. 4, Q17). The high number of motorcycles remains a challenge in Delhi, Dhaka and Chennai to bicycle tracks because of their tendency to use the tracks to avoid traffic congestion on the main carriageway. An urban planner and transportation researcher in Delhi shared her experience designing and iterating on a bicycle lane: while speed control measures work well, bicycle lanes pose design challenges because bicycles and two-wheelers are similar in form. Barriers such as bollards that block two-wheelers also impede bicycles (Extended Data Fig. 4, Q18).

We observed, and many interviewees highlighted, that during periods of congestion on the main carriageway, separated bicycle tracks were flooded by motorized two-wheelers, posing a safety hazard and a travel impediment to bicyclists. Numerous news stories echo interviewee responses and document encroachment by motorcycles and street vendors and the disappearance of protective barriers, often within months of installation (Extended Data Fig. 4, N1–N2). Furthermore, it was common for bicycle lanes and tracks to be obstructed by parked cars, shop extensions, fallen trees during the monsoons, and construction debris and rubble. Consequently, we saw bicyclists prefer cruising on the main carriageway of arterials during long commutes, even in the presence of separated bicycle tracks or lanes.

In addition to these issues, experts and activists also emphasized the need for longer lanes or networks throughout the city, rather than providing them in small stretches (Extended Data Fig. 4, Q18–19). The separated bicycle tracks and painted lanes that we observed in Delhi and Dhaka were largely situated near high-income neighborhoods. These tended to be in disconnected and short stretches that would account for an insignificant portion of most bicyclist trips in the area.

Retail, repair and durability of bicycles

Over two-thirds of respondents in Delhi rode steel-frame fixed-gear roadsters with rod-actuated brakes, 1.25–1.5-inch-thick wheels of varying diameters and fitted with a back carrier. These were also common in the other cities, but fewer than in Delhi (‘Simple’ or ‘Desi cycles’ (Delhi), ‘Saada cycles’ (Chennai), ‘Busanga Volvo’ (Accra) and ‘Phoenix’ or ‘Bangla cycles’ (Dhaka)). Although typically used by poorer bicyclists, these were not the least expensive bicycles but were well-recognized for their durability, longevity and ease of access to parts and repair skills in the market. These roadsters are designed for slow movement and carrying goods or a pillion rider, often for longer distances. Government-distributed bicycles in Chennai targeted at school-going adolescents were these roadsters, which also circulated among the urban public through the second-hand market.

Bicycles tended to be used for a long time after purchase (Table 1), including models now phased out of manufacturing. Most respondents rode second-hand bicycles (Extended Data Fig. 5). Bicycle users noted that regular repair and maintenance contributed to their longevity. Across types of stakeholders, people registered the importance of repair shops and services for sustaining bicycling in cities. Repair shop owners tended to be highly knowledgeable about local bicycling conditions and tuned bicycles accordingly (Extended Data Fig. 6, Q20).

Despite their importance to bicycling in the city, repair service providers for bicycles were largely small shops or footpath stalls with a precarious existence. Many had downsized in the past decade and had to frequently negotiate with police and regulatory authorities to continue to occupy footpaths (Extended Data Fig. 6, Q21). The marginal place of the repair service providers further resonated with the marginalization of the low-income bicyclists in the city, as we heard repairers speak of themselves as poor bicycle riders. Larger bicycle retailers shared issues such as increasing taxes on the components and accessories of bicycles that impacted the repair service providers and buyers (Extended Data Fig. 6, Q22–23).

Electric bicycles and bicycle-sharing systems did not connect with such existing bicycle repair providers. We observed very few electric bicycles in the four cities and most repairers, who did not specialize in e-bicycles, expressed difficulties in handling their electrical components and finding parts. Further, existing bicycle-sharing systems in Delhi, Chennai and Dhaka were largely in disuse owing to a lack of maintenance.

Compared to automobiles, bicycle parking required little real estate, and many bicyclists reported parking inside their homes. Establishments offered bicycles and parking spaces to employees close to their place of work (Extended Data Fig. 6, Q24). New and proposed metro rail stations in Chennai, Delhi and Dhaka offer bicycle parking stations. Suburban railway stations in Delhi and Chennai also have paid bicycle parking services near them, enabling last-mile connectivity not just for trains but also for arterial roads. Such stations are actively used by cyclists.

Bicycle repair and parking service providers across the cities mentioned several shared issues, such as taxation, road quality, bicycle designs and increasing economic need to accommodate motorcycle parking and air-filling services. Such service providers, however, appeared to have little political representation as a group or network to lobby for issues concerning the bicycle service economy.

Discussion

Policy actors in LMICs and globally recognize the importance of increasing bicycling to support a global transition toward sustainable mobility. As we show, in many LMIC cities, supportive infrastructures and home-grown bicycling cultures persist despite increasing automobility and expanding demands to decongest urban roads and build expressways, that is, the portal to foster bicycle cultures has not yet been foreclosed. In contrast to the car-centric built environments in many HICs, LMIC cities navigate a highly evolving vehicle-use culture and infrastructures. The experiences and perspectives of current bicyclists and stakeholders surrounding bicycling cultures offer valuable insights into fostering large increases in bicycling in these environments23,24,30, and to decode and interrupt the societal and infrastructural pressures to motorize. Therefore, we used mixed-methods research to understand the current state of and the impediments to bicycling from the perspective of bicycle users and center issues concerning bicycling in four LMIC cities.

Mimi Sheller’s mobility justice framework builds on the idea that all forms of mobilities and their lack are enmeshed in power relations. Sheller uses ‘kinopolitics’ to theorize this dynamic, building on the ‘new mobilities paradigm’31, and argues that just mobility transitions require understanding embodied and everyday movement, urban and regional transport systems and global movements (including of knowledge) as interconnected7. Our analysis approaches bicycling experiences in Delhi, Chennai, Dhaka and Accra as windows to multiscalar urban kinopolitics—particularly of urban fragmentation—and explores research and policy pathways to mobility justice that the fragments, bicycles, can unveil.

Bicycling in a splintered city and relational ontologies

Sheller’s kinopolitics manifests in cities through fragmentation: infrastructural developments, such as high-speed road networks, intensify connections to centers of global capital while disconnecting them from local contexts and fracturing urban geographies32. Bicyclists in LMICs in our study navigate increasingly fragmented cities, whose geographies are ‘splintered’ along various physical, temporal and institutional axes, including socioeconomic class and migration7,33. We find that these bicyclists are predominantly poor, politically disempowered and invisible to most policymakers. In the face of automobile-centric urban growth34, bicyclists face systemic marginalization in LMIC cities35. Built infrastructural interventions for bicyclists and organized advocacy usually happen in elite parts of the cities aimed at converting car users36,37, attending less to the needs of a large fraction of current bicyclists, who tend to be poor, akin to other contexts in Global South38, and to the ubiquitous presence of motorized two-wheelers, which are closer to bicycles in cost, form and their use of road space and extant bicycle lanes. Several bicyclists were also domestic migrants to the city and had learnt to bicycle in times or areas with fewer automobiles. The varying bicycle demographics on arterial and residential roads further highlight a gender and age divide. Raised medians, U-turns and multilane arterials physically splinter roads and exacerbate unequal access to bicyclists. Such road features support uninterrupted automobile flow but interrupt and prolong bicycle trips. The few bicycle lanes that remain are, quite literally, in short fragments. The space-time of the city roads was further fragmented through the day and by weather. The fragmentation in and between the roads mirrored institutional splintering. The responsibility for bicycling infrastructure was peppered among different government bodies. Urban development, advocacy and sports affairs agencies that engaged in promoting bicycling amongst youth and automobile users for physical activity benefits and leisure followed choice-based models of physical activity39. Foley and colleagues40, in the African context, highlight that given the already high levels of walking, policies should focus on context-sensitive improvements to safety and reducing poor walking conditions, rather than simply promoting walking. Following this suggestion, while increasing cycling should remain a goal in LMICs, improving safety for current cyclists, many of whom possess limited financial and political clout, should not be undermined by a focus on converting automobile users. Current, and particularly low-income, bicyclists we met, including the recipients of bicycles distributed by education agencies and philanthropy focused on empowerment through access, faced the consequences of unequal budgetary allocations for roadways, emphasizing unfettered automobility.

While urban automobility splinters cities, studying bicycles as urban fragments, as Colin McFarlane suggests33, through the experiences of people who use, repair and retail them, reveals their relations with other elements along urban faultlines. We see that while relationships such as repair services, parking facilities, police assistance and second-hand bicycle access support bicycling, those that deter include dangerous automobile interactions, exertions from physical built environments, safety concerns raised by caregivers and financial barriers to bicycling. Our respondents, even when they individualized the responsibility of risks to the bicyclists, consistently identified traffic safety as a primary concern influencing their choice to bicycle, as has also been reported in HICs16,41,42,43. It is well documented that bicyclists in general face a high risk of injuries in LMICs44,45, including owing to car-oriented infrastructure that increases exposure for bicyclists46,47. However, perceptions and individualized attributions of risk are socially constructed and experienced along the contextually situated axes of power and inequality48, here of socioeconomic class. Our findings additionally revealed that relational dynamics in risk perceptions shape the perspectives of adult kin in discouraging children and the elderly from bicycling, extending prior work on age and gender variation in stated preferences against navigating mixed traffic49. The broader culture of individualizing such risks has political implications, such that lobbying and public discourse around safer bicycling is largely limited to elite and middle-class-led bicycling organizations. We register a parallel disenfranchisement and lack of organization amongst bicycle repair and parking service providers. However, both elite advocacy for bicycle safety and widespread kin concerns regarding young or elderly bicycling emerged from a framework of care for the potential bicyclists’ health and safety.

Bicyclists’ careful relations and interfaces with other urban elements are important for designing interventions that increase bicycling in ways that resist fragmentation and reconfigure integrated and just mobility systems. Ironically, high-speed automobility creates an environment that is so hazardous for bicycling that the most prominent demand of bicycling advocates and researchers is for bicycle infrastructure that separates them from automobiles. Bicycle lanes and tracks are increasing, albeit slowly, in LMIC cities. However, the utility and value of such physical infrastructure investments will greatly benefit from better geographic understanding of bicyclists’ needs, routes and safety concerns. Bicyclists’ subversions of current road logic particularly offer insights into these, resonating with prior research highlighting the need to audit automobile-enabling road-design features for bicycle friendliness47,50.

Beyond hard infrastructure, there is a need to support the ecosystem of material resources, services and stakeholders that enable cycling in LMICs. These are critical relational and collective dimensions of bicycle infrastructures. We observed that existing bicyclists are served by long-standing maintenance infrastructures, including bicycle repair service providers, and resale and reuse cultures thrived across the cities, caring and supporting the longevity of bicycles51. By contrast, large bicycling-sharing projects are failing typically because of inadequate maintenance.

Thus, we argue that bicycling policy must move beyond a fragmented focus on physical infrastructure, especially cycle tracks, and empowerment through bicycle access. An integrated and relational approach, integrating financial, maintenance, planning and human infrastructures (Extended Data Fig. 7), is essential to support bicycling and advance mobility justice in LMIC cities.

Repairing bicycling infrastructure through an ethics of care

Mobility transitions require recognizing how “differential relations and capabilities for mobility are… built into cultural economies, urban forms, transportation infrastructure, and global mobility regimes, shaping differential access and exposure to harm” (Sheller, p. 436)31 and working with an ‘ethics of care’. Policy actors and bicycling stakeholders undertake or partake in caring infrastructures, for instance, through bicycle distribution schemes or letting bicyclists move ahead of vehicular traffic at traffic signals. Even care is far from apolitical. For instance, through an analysis of bicycle lanes in Ecuador, Julie Gamble provides a critical analysis of governmental expressions of care solely through built interventions and crisis infrastructure, which fail to attend to systemic and sociocultural experiences of risk52. Alternatively, kin dissuasions of children and the elderly bicycling and bicycle repair work emerge as complex expressions of care in the social and human infrastructures surrounding current and prospective bicyclists.

The limited governmental ‘care’ or scaffolding for bicycling in LMIC cities discussed emerges partly from the perception of vehicles, generally assumed to be automobiles and governed by transportation sectors, as sources of risk that need to be licensed and regulated. As bicycles inherently pose very little risk to their riders compared to motor vehicles, they have largely been ignored from policy perspectives in these cities as personal pursuits. Within the cities’ policy environments, facets of bicycling are governed by education, transportation or sport authorities. However, amid the kinopolitics of automobility, increases in bicycling support public health and sustainable development by reducing greenhouse gas emissions, pollution and traffic injuries, while increasing physical activity. We propose that fostering bicycles for these benefits requires policy advocacy from health and environmental agencies (for example, ministries of health or pollution control agencies), supported by research on the barriers faced by current and prospective bicyclists. While the specific departments implementing or investing in bicycle-related policies may be varied, a coordinating health agency’s care, stewardship and monitoring of enforceable policies can be used to track data-driven implementation to support investments in bicycling.

Learning from bicycle repairers’ care for bicycles and their essential role in supporting bicycle cultures, we adopt a repair-oriented approach to bicycling infrastructures and policy. As Anjaria demonstrates24, existing bicycling cultures in these cities challenge perceptions of LMICs as ‘not yet’ having sustainable transportation systems. The cities already engage with bicycling policies and practices, although fragmentarily negotiating heterogeneous and fast-changing transportation environments, as we have demonstrated in this paper. We propose that future policy and research help repair and strengthen existing bicycling infrastructures toward safe, sustainable and just mobility transitions responsive to local needs.

Methods

We conducted mixed-methods research with interdependent components—ethnographic field observations, intercept surveys and semi-structured interviews53. Data collection was led by local teams trained in common methods. The teams maintained ongoing communication to ensure consistency and integration across sites.

Our methodology was geared toward exploring what kinds of data the specific methods and their combinations yield for us. The research activities (Extended Data Fig. 8) were distributed across cities based on the availability of site-specific funding and background data, varying institutional timelines of the research. The research activities were not replicated in all four cities toward a comparative perspective. Instead, we explored what the different methods can offer toward answering the research question and how their insights may be synthesized.

Field observations were conducted to identify research locations (Supplementary Information 1: Ethnographic observation prompt list). Road and market observations undertaken in June and October 2022 in Delhi, Chennai and Dhaka were fluid and grounded in principles of ethnographic research54. Observations in June and August 2023 were open-ended, albeit guided by a tentative list of physical infrastructural features, characteristics of cyclist-vehicle interactions, sociodemographic features and any other distinct aspects relevant to bicycling.

Ethnographic research within this project was conducted with cognizance of the multiscalar politics of mobility. The extent of international and intercity engagement in the cities’ fieldwork was limited to the necessities of collaboration and sharing of tools. A mid-project review workshop (internal to the team) and a South Asia-specific stakeholder workshop (public event) were conducted in Delhi. One team member from Chennai participated in the scoping fieldwork in Dhaka. The team members were immersed in local bicycling contexts and general road use in the four cities in various ways throughout the research conception and implementation in 2022–2024. We also paid ongoing attention to emerging bicycling communities, policy changes and urban development processes, including budgetary plans55,56,57, built infrastructures and news stories emerging about cycle infrastructures. Two of the authors commuted to their workplace by bicycle regularly and one participated in city bicycling group events. Many of the authors and research assistants were nonbicyclists themselves but were kin of commuter and leisure bicyclists. Two of the authors bought new bicycles and safety-related accessories during this period, including using a monthly installment financing. In Delhi, Chennai and Dhaka, we made multiple trips to bicycle repair shops to seek services for our bicycles, including air-filling and fixing punctures, and used paid parking services near public transit. We accessed local neighborhood roads, enclaves and arterials, including highways, observing and experiencing bicycling.

Intercept surveys with bicyclists were conducted on arterial roads in Delhi (N = 171), Dhaka (N = 50) and Accra (N = 238), in the local language (Hindi, Bangla, English and Akan). Intercept surveys were conducted where respondents are engaging in the activity being researched58,59,60, allowing focused attention on bicycling sites. Survey sites were chosen to have a high bicyclist presence based on traffic counts in Delhi61, field observations, regular transit experiences and information received through interviews in the other cities. Survey respondents (‘respondents’, in this paper) were recruited by researchers on the roadsides of arterial roads. It is important to note that our intercept surveys represent a biased sample of the bicycling population in these cities because of the survey locations (primarily high bicycle volume arterial sites), time of survey (daytime), weather conditions and who agreed to participate. Notably, only some of the bicyclists who rode by the survey site stopped, and this group was probably biased by occupation (and related trip characteristics), gender, age, caste, ethnicity and ability of the bicyclists. Of the very few women that we saw on the arterial roads in Delhi and Dhaka during surveys and field observations, only one (Dhaka) participated in the survey.

Respondents were asked about their most recent trips, income, migration status, household and bicycling skills (Supplementary Information 2: Intercept survey questionnaire). Features of the respondents’ attire and bicycles were noted as a proxy for income, socioeconomic class and employment. Surveys were conducted between October 2022 and November 2023. Additionally, 50 nonbicyclists were surveyed in Dhaka at sites adjacent to the intercept survey sites to explore surveying nonbicyclists who encounter bicycles during road use and to recruit interviewees with varied socioeconomic class, gender and road-use statuses. The focus on bicyclists and stakeholders surrounding them was particular to the 2022–2024 phase of a larger ongoing research project exploring urban bicycling in LMICs, considering funding and project timeline limitations. Our research phase focus on bicyclists, particularly in the intercept surveys, is, by no means, to devalue research with current nonbicyclists in the city regarding bicycling. We see immense potential in such research to explore a range of concerns, including mode shift preferences, safety concerns and travel experiences. The interviews reflect a beginning toward this commitment.

Semi-structured interviews were conducted in Delhi (45), Chennai (32) and Dhaka (32) to explore cyclists’ experiences through in-depth conversation62. These were either 8–15-min-long with intercepted cyclists or 20–60-min-long with stakeholders (Supplementary Information 3: Semi-structured interview preliminary questions and probes). Stakeholders (Extended Data Fig. 9) included personnel who either interacted regularly with bicyclists, such as repair service providers and retailers, parking providers or people invested in bicycle infrastructure policy or design, including manufacturers, bicycle activists and officials. Interviews were predominantly conducted in the everyday environments of the users and stakeholders, allowing for conversations regarding their relationship with bicycles, how they use the vehicle and their thoughts on cycling demographics and infrastructure in the cities. Several of the interviewed stakeholders were not currently bicyclists. As indicated in Extended Data Fig. 9, over 50% of the interviewees in Chennai and Dhaka, and 28% in Delhi, were nonbicyclists. In Chennai, interviews were conducted in areas surrounding schools with parents and kin of bicyclists, and two doctors near a school zone. Additionally, we conducted targeted interviews with women bicyclists in Delhi and Chennai because there were very few women who participated in the intercept surveys. We prioritized relational narratives regarding bicycle use and perceptions over personal information, including age, income and other social identifiers.

Interview data and field observation notes were thematically coded using Dedoose in addition to subjective qualitative analysis through discussions with teams across cities. An initial limited list of codes was generated through a literature review. Following open coding of the interview data and ethnographic research, a preliminary list of themes, organized through the lens of infrastructures25 (Extended Data Fig. 7), guided data organization. These were further divided into broader themes of transportation equity, sustainability, safety and comfort toward the final synthesis of data. The research employed fluid and responsive constructivist methodologies, policy relevance and field-specific grounded and dialogical thinking among the research team. Policy practicalities and politics of public health-implicated infrastructures are imbricated in the sociopolitical process of producing empirical data and evidence, such as on (1) what counts as bicycle infrastructure and (2) bicycle infrastructure reach, use and limitations. We have attempted both to identify and integrate bicycle infrastructures and allied domains of policy influence through contextual research (Extended Data Fig. 7) and to analyze the findings through the lens of their implementation (Supplementary Information 4: Iterative implementation framework for bicycling in LMIC cities).

Qualitative findings emerged through analytic memo writing and triangulation of survey findings, observations and interview research. We have emphasized analytical depth and relevance of themes over quantitative measures such as frequency of themes or data saturation, in line with established ethnographic research methods54. Our decision to do so emerges from the inherently subjective quality of the methods. Further, the time-span of the interview, the interviewer’s understanding of saturation of a theme in a specific interview during the conversation, and across interviews can impact measures such as frequency.

Approaches akin to intersectional analysis were central to our research emphasis on equitable mobility. We studied how gender, class, age and migration intersected in various ways through urban geography in bicyclists’ experiences. One key limitation of our research is that race, ethnicity and caste were not used as frameworks of analysis and require a discussion on the specific geographies of each of the cities, which is beyond the scope of this article.

The study was approved by the Social and Behavioral Sciences Institutional Review Board, The University of Chicago (protocol number IRB22-1171-AM003).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.