The ability to move about freely and safely is a prerequisite for individuals' participation in civic and economic life outside the home. Yet many women in Pakistan face street harassment, pressure to conform to social norms against traveling alone, and don’t often have the ability to drive vehicles or ride bicycles. Over the past decade, policymakers like the Punjab Commission on the Status of Women, NGOs like the Salman Sufi Foundation, and even companies like Karachi Electric have tested “scooty” or motorcycle loan interventions to boost women’s ability to participate in the labour force. Then, why do we still not see women driving the most ubiquitous method of transportation in urban Pakistan?
The costs of restricted mobility for women
------------------------------------------
A wealth of evidence indicates that in South Asia in general and Pakistan in particular women have lower levels of mobility than men do. Women are much less likely to leave the home for any purpose than men, and then take more expensive, riskier, and more inconvenient modes of transport when they do (for example, expensive private cars, rickshaws that fail to come at their daily scheduled times, or walking, which provides little shield from urban street crime).
When discussing their experiences getting around Karachi, women frequently complained to me about the inconvenience of relying on their male family members for rides. They described instances of sexual assaults at night as well as midday, and catcalls from men emboldened by the boys’ club that is Karachi’s public spaces. It is not surprising that even families who disagree with the social norm of women remaining within _chador aur chaar-diwari_ (“the veil and four walls”) end up, in practice, restricting women’s mobility for legitimate safety reasons.
One can see how the immobile equilibrium sustains itself. Evidence from South Asia suggests that women's immobility leads to them developing fewer skills from school, opting out of the labour force or for lower-paying and less ideal jobs, and to sustaining smaller social networks. As households depend on men for income and decision-making, women struggle to gain bargaining power—reinforcing the cycle. Can interventions disrupt this pattern?
Designing effective mobility programmes for urban Pakistan
----------------------------------------------------------
Drawing on lessons from private- and public-sector scooty loan programmes in Pakistan, I aimed to design a mobility-improving treatment that could have broad take-up and lasting results. Rather than recruiting only among the most progressive or younger crowd or targeting women currently in the labour force or with ample funds to buy motorbikes outright, I worked with several local partners to expand my sample. For example, one recruitment partner was The Hunar Foundation, a large vocational training institute that works across the city. Kashf Foundation came on board to provide an in-kind loan product, a more scalable intervention than giving motorbikes outright.
Over several months, research assistants surveyed hundreds of women who responded to recruitment materials, conducting repeat surveys to gather financial details and interview male loan guarantors.
In the end, out of over 500 women who made credible and costly signals of interest, only 45 came to formally participate in our programme. Combining statistical analyses of dropout predictors and qualitative interviews with former participants, I suggest the following insights:
Figure 1: A random forest model showing the importance of different variables in the scooty programme take-up.
Notes: A random forest model, a machine learning method that identifies which variables are most predictive of an outcome—in this case, dropping out before randomisation. Income, age, and unemployment were strong predictors, even after excluding households earning less than PKR 60,000—a common loan eligibility threshold. Figure generated by the author.
**Cost remains a major barrier.** Despite intending to appeal to lower-to-middle-income women, our programme found that this group's economic crisis in Pakistan in 2022 and 2023 limited participation. Due to inflation, oil, and price hikes (particularly for scooters), our programme’s minimum monthly instalment nearly doubled from PKR 4,200 to over PKR 8,300 during recruitment. Figure 1 above shows the results of a random forest model, a machine learning method that identifies which variables are most predictive of an outcome—in this case, dropping out before randomisation. Income, age, and unemployment were strong predictors, even after excluding households earning less than PKR 60,000—a common loan eligibility threshold. Cost remains a primary constraint, even for interventions targeting financial inclusion.
**Male guarantors impose further constraints.** Requiring a male family member to serve as a loan guarantor, as is typical in Pakistan, excluded women from households with absent or migrant men and even capable breadwinners who did not meet bank criteria. Professional women sometimes had to rely on unemployed male relatives, like an 18-year-old son, adding yet another person to convince—one who may not agree with their participation. The State Bank of Pakistan (SBP), responsible for governing Microfinance Banks, would need to communicate clearly that it supports lending to financially-qualified women who apply independently.
**Existing skills are not the bottleneck.** A striking 69% of women in my sample (of 185) could already ride a bicycle, a skill closely related to motorbike operation. Cycling ability was not a significant predictor of dropping out before randomisation, and 80% of women who began the motorcycle training completed it (93% if excluding unrelated medical and family emergencies). This highlights that prior technical skills were not the main obstacle—structural and financial barriers were.
**Access to family-owned vehicles is not enough.** Two-thirds of initial respondents’ families already owned personal vehicles. Yet, this was not the primary driver of participation. While having a vehicle in the household might suggest opportunities for men to teach women to drive, in the subsequent experiment there is no meaningful association between a family’s prior motorbike ownership and women driving these vehicles, after controlling for treatment assignment.
Addressing social and economic barriers will be critical for scale-up
---------------------------------------------------------------------
These findings reveal that programmes to improve women’s mobility must go beyond skill provision. Results from the motorcycle training programme and the training-and-loan programme suggest that both meaningfully increased the likelihood that women had actually driven a motorcycle recently (unlike the results of other training-only interventions where women keep this skill in their back pocket). Looking forward, addressing economic and social barriers will be critical for scaling such programmes. The SBP must encourage and financially support more inclusive loan structures, while policymakers need to prioritise investments in affordable, women-friendly public transportation. For Pakistan to achieve inclusive growth, Pakistan must make improving women’s mobility a central focus of its development policy.