Social Sciences & Humanities Open 7 (2023) 100442 Contents lists available at ScienceDirect Social Sciences & Humanities Open ELSEVIER journal homepage: www.sciencedirect.com/journallsocial-sciences-and-humanities-open Regular Article Lessons learnt through gender-based travel data collection and related sexual harassment in Sub-Saharan Africa Marianne Vanderschuren a,*, Heather Allen b, Philip Krause c, Tanya Lane-Visser a a Centre for Transport Srudies, University of Cape Town. 1 Madiba Circle, Upper Campus, Rondebosch, 7701, South Africa b Independent Gender & Srntainable Transport Consultant., Median SCL, Rue du Moulin d'hoUers B, 1495, Villers la Ville, Belgium c GoAscendal. 54 Oxford St., DurbanviUe, Cape Town, 7550, South Africa ARTICLE INFO Keywords: Gender Personal safety Sexual harassment Public transport Africa 1. Introduction ABSTRACT Background and objective: Key motivations for the study described in this paper were to gain insights into the differences in travel behaviour between males and females, as well as gender-based sexual harassment exposure in public transport in sub-Saharan Africa. The mobility and sexual harassment results, based on data collection in two African cities (Blantyre in Malawi and Lagos in Nigeria), is presented in this paper, to document the true extent of sexual harassment and ways in which sexual harassment practices affected women's mobility on the continent. The analysis aims to illuminate gender-based mobility differences and sexual harassment experiences, as well as provide some guiding principles towards addressing sexual harassment challenges. Method: and Data: This study developed a standardised gender-based travel and sexual harassment perception questionnaire that was applied via pen-and-paper, as well as technology (offline tablets) in the cities of Lagos (Nigeria) and Blantyre (Malawi). Overall, 1478 respondents participated. Approximately 58% of the sample in both cities were women. The data collected was analysed to highlight differences between male and female travel behaviour and differences in relation to exposure to sexual harassment whilst in the transport system. Results: Data revealed that women make more, shorter trips, while they carry a significantly higher sexual harassment burden. All parts of public transport trips are affected, including travelling to/ from the system, while waiting for the vehicle and in the vehicle. The need for transfers increases the risk of crime and sexual harassment. This paper concludes that there is a need for improved policy frameworks, educational campaigns, for men and women, as well as the creation of safe reporting structures when sexual harassment does occur. Actions are needed in the two case study cities and beyond. Since the early 2000s, the topic of 'genderl-based development' has gathered momentum. Many of the Sustainable Development Goals look to embed gender equity and female empowerment into the International Agenda 2030. UNDP's Human Development Report (1 995) highlighted that 70% of the 1.2 billion people living in poverty worldwide are women. This figure has remained stubbornly high ever since (Azcona et aI., 2020). In practice, the sheer pace and volume of urbanisation compounds the under representation and difficulties of addressing gender issues. This is especially pronounced in Africa, where almost 900 million new urban dwellers are expected by 2030 Chttps://www.moib rahimfoundation.org). Today, half of Africa's urban dwellers live in precarious conditions in informal settlements. Some 75% of new urban inhabitants are younger than 35 years old, with a high share of them being female Chttps://www.citiesalliance.org). Across the world, women face real and perceived threats of genderbased violence, assault, and harassment, while negotiating their daily travel in public spaces. In developing countries, limited access to, and the safety of, transport is estimated to reduce female labour force participation by 16.5% (SuM4All, 2019). This paper focuses on the individuals' perception of their own security, which is defined as their * Corresponding author. E-mail address: marianne.vanderschuren@uct.ac.za (M. Vanderschuren). 1 Sex is the biological classification of males and females, determined at birth, while gender is a culturally defined set of economic, social, and political roles, responsibilities, rights, entitlements, and obligations associated with being female, male and LGBTQIA+ . In this paper the term gender is used interchangeably with women. https://doi.org/10.1016/j.ssaho.2023.100442 Received 27 June 2022; Accepted 7 February 2023 Available online 9 February 2023 2590-291l/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses!bync- nd/ 4.0/). M. Vanderschuren et aL perception of risk or threat from an intentional personal act, attack, or aggression. Women living in informal settlements primarily rely on public transport for their daily travel needs. Besides common concerns related to the use of (informal) public transport, such as affordability and road safety, women have the added concerns related to personal security (Jackson et aI., 2013; Vanderschuren et aI., 2019). Sexual harassment of women in public space is widespread. A largescale survey of street harassment, in 42 cities in 2015, based on a sample of 16 600 female respondents spread across the world, revealed that approximately 84% had experienced street harassment for the first time before they were 17 years old (Hollaback! and The ILR School, 2015). Allen et al. (2018), who interviewed female and male public transport users in Santiago (Chile), Buenos Aires (Argentina) and Quito (Ecuador), found both sexes pointed to personal safety as a major concern, but women (between 61 % and 73%) were more worried about personal safety while using public transport than men (between 58% and 59%). Orozco-Fontalyo et a1. (2019) found similar results (more than 60%) regarding the exposure to sexual harassment related to the use of Bus Rapid Transit (BRT) use in the city of Barranquilla. Feelings of insecurity were even higher for mothers with children under 15 years old (78%). Some 60% of the women who claim to have felt insecure were between 18 and 40 years old. Furthermore, nearly 90% of women in Buenos Aires had experienced some level of sexual harassment, at least once while using public transport in the previous 12 months. The study found that 36% of those who felt higher levels of insecurity belong to households with the lowest level of income (less than approximately US$85 per month), compared to 14% of women who live in households with the highest income in the sample (more than approximately US$190 per month), confirming that the sense of insecurity decreases as income increases. In India, 91 % of women felt that public transport was very unsafe, unsafe, or somewhat unsafe (Shah & Raman, 2019). In Loukaitou and Ceccato's recent book (2020), the Lagos (Nigeria) case study found that female college students were 2.2 times more likely to experience non-verbal sexual violence than males. Only 5% of students reported always feeling safe walking to or waiting at the bus, keke napep (tricycle), or motorcycle taxi stops after dark. About a third (31 %) of the students had never felt safe walking to the transport stops at night, and about one-fifth (21%) rarely felt safe (Loukaitou-Sideris & Ceccato, 2020). Capasso da Silva and Rodrigues da Silva (2020) established that over 20% of the respondents from a Brazilian University campus have been victims of violence on their trips to or from the campus. Concerns over personal security perception can lead to trip avoidance (Jeong et al., 2017; Piscitelli & Perrella, 2017) or a change in mode choice (Delbosc & Currie, 2012; Verma et aI., 2017). Fears about female personal safety and security are well documented in both the global north and south, but there is little evidence from subSaharan African cities currently available in the literature. In Africa, transport data is generally poor, resulting in disaggregated information on urban transport, which is rarely robust or widely available (Gwilliam, 2013). 2. Research Aim Key motivations for the study described in this paper were to gain insights into the differences in travel behaviour between males and females, as well as gender-based sexual harassment exposure in public transport in sub-Saharan Africa. The work included the development of a survey instrument for data collection, as well as a methodology which can be replicated in other cities of Africa, and beyond. This paper sets out to provide insight into travel behaviour differences between men and women, and their exposure to gender-based sexual harassment, based on data collection in the two above mentioned African cities, to document the true extent of sexual harassment, and ways in which sexual harassment practices affect women's 2 Social Sciences & Humanities Open 7 (2023) 100442 mobility on the continent. The analysis aims to illuminate gender-based mobility differences and sexual harassment experiences, as well as provide some guiding principles towards addressing gender-based mobility and sexual harassment challenges. The research approach and the findings are being used as part of a wider project, that develops a user-friendly online decision-support tool for urban transport planners and related stakeholders, with a focus on addressing sexual harassment in public transport in Africa. 3. Data Collection methodology This study focussed on collecting data from urban public transport users and key stakeholders in two African cities with distinctly different transport profiles - Lagos (Nigeria) and Blantyre (Malawi). Lagos is a large mega city (over 10 million inhabitants) with a wide variety of public transport modes, including formal (Bus Rapid Transit and ferries) and informal (minibus taxis, motorcycle and cycle taxis) modes; Blantyre has just over 1 million inhabitants and the modes available are predominantly minibus taxis and sedan taxis (no formal public transport services are available). Data was collected in both cities using the same questionnaires, slightly adapted for the local context. Both technology- (offline tablets) and paper-based survey collection methods were used to randomly intercept people at main transport hubs, while the survey findings were validated in focus groups in each city. Interestingly, the response rates were influenced by the data collection method, where digital collection rendered better response rates. Before data collection commenced, ethics clearance was obtained via the University of Cape Town. This was explained to questionnaire and focus group participants. The consent description that was shared with participants first, elaborated on the fact that participation was voluntary and that opting out was possible at any time. Participants provided consent in the very first questions answered. In Blantyre, 686 respondents participated in the survey while the response rate in Lagos was 792 participants. The data was collected in the period between 5 March and March 12, 2021, between 9h30 and 18hOO daily. For both cities, the response rate falls comfortably within the 95% reliability sample size margin. For Blantyre, based on a statistical Confidence Interval (CI) calculation, there is a 96.26% certainty that the sample is representative for the inhabitants in the city. In Lagos, the statistical reliability that sample participants represent the urban population in the city is even higher with 96.52% reliability. By design, in both cities, more females were targeted and, therefore, participated in the survey. In Blantyre 58.1 % of respondents were female, while this was 57.9% in Lagos. Nonetheless the total numbers of respondents (male/female) fall within one standard deviation, representing a lack of bias. Although the sample size of respondents in both cities is considered high, in Blantyre, there was answer fatigue towards the latter part of the questionnaire, which affects the sample size for some analysis conducted. The useable sample size is indicated throughout the analysis in this paper. A T-test comparison between the sample groups was carried out. Where statistically significant differences were found, this is indicated in the paper. 4. Respondent profiles Most respondents in Blantyre (55%) and Lagos (56%) were between the ages of 26-46 years. The second largest respondent age category was 19- 25 years (28% of respondents in Blantyre and 30% in Lagos - see Fig. 1). In Lagos, the average age for males and females was approximately 28 years, while the Blantyre females were on average 27 years old, and males a little older (average 30 years old). These averages are calculated, assuming a normally distributed sample. Surveyors also documented M. Vanderschuren et aL lagos, male lagos, female Blantyre, male Blantyre, female 20% 40% 60% SOo. 100% . 19·25 years _ 26·46 yea rs .46·55 yea rs . 56 years and over Fig. 1. Age Distribution of Respondents (Blantyre n = 686; Lagos n = 792). those who refused to participate in the survey and found that more women than men were interested in participating once they had been made aware of the subject matter. Furthermore, more males tended to omit answering some questions, affecting the response rates of individual questions presented in this paper. Respondents in Lagos have a particularly high education level. Over 80% attended college or university, while this was only the case for 35% of Blantyre's respondents. The high education level in Lagos was expected. Based on previous surveys in or around the BRT system in Lagos, as verbally indicated by the Lagos Metropolitan Area Transport Authority (LAMATA), responses were in line with the typical passenger profile of the BRT. The percentage males with college or university level education in Lagos was higher than for females (see Fig. 2). These high average education levels indicate that the BRT was used mainly for commuter trips when the surveys were done. Blantyre respondents overall had a lower college/ university education rate than in Lagos. However, the sample showed that there is a noteworthy higher proportion of male college/ university graduates (42%) than females (29%) in Blantyre. Overall (see Fig. 2), between 71 % (female) and 81% (male) of respondents in Blantyre have at least completed high school, while the proportion of female high school only graduates (41%) is higher than males (39%). In Lagos, the average ed· ucation level is substantially higher, i.e., between 98% (females) and 99% (males) of respondents have completed at least high school leveL Most respondents were employed, business owners or self-employed. In both Lagos and Blantyre, male respondents occupy more fonnal employment posts than female respondents. The occupation category in Lagos with the largest share was found to be 'fonnal employment', especially for males (49% vs. 42%). In Blantyre, the occupation category with the largest share was found to be 'self-employed', with 49% of females selecting this category. Remarkably, a substantially lower proportion (33%) of males selected this category. This may be caused by the inaccessibility of fonnal jobs by females, while the education levels may also play a role. Students comprised a slightly higher proportion of survey respondents in Blantyre than in Lagos: males (13% vs. 11%) and females (19% vs. 17%) respectively (see Fig. 3). Regarding those who answered 'other' as their occupation, the most cited occupation for lagos, male 15% 1 Lagos, female 17% 2 Blantyre, male •••••••• I:==::i39E%;:===='~9~%=] Blantyre, female 41% 29% 20% 60% 80% 100% _ Yes, College/University • Yes, Secondary/High school • Less than high school Fig. 2. Education Level of Respondents (Blantyre n = 686; Lagos n = 790). 3 Social Sciences & Hwnanicies Open 7 (2023) 100442 Lagos, male Lagos, female Blantyre, male Blantyre, female _ Employed _ Business owner _ Self·employed _ Student _ Other Fig. 3. Employment Profiles of Respondents (Blantyre n = 315; Lagos n = 792). women was 'homemaker', 'housewife' or 'staying at home', while men indicated that they were ' looking for work' or 'doing nothing'. It is noteworthy that Blantyre had a high number of respondents not willing to answer this question. The two cities, thus, vary greatly in terms of average education level and employment status of commuters and, consequently, in tenns of gender equality. 5. Trip profiles Most respondents in both cities indicated that they were making a typical trip at the time they were surveyed (64% of trips in Blantyre and 65% in Lagos). In both case study cities, more men were undertaking typical trips than the female respondents. In Blantyre, 68% of men and 61 % of women were undertaking typical trips. In Lagos, 63% of women were undertaking typical trips, slightly more often than their counterparts in Blantyre, while it was the same for men (68%) in both cities. This is in line with the occupation statuses; as more men in our sample have regular employment it is, therefore, more likely that they will be making more typical (often commuter) trips (see Fig. 4). In Lagos, typical trips were mainly for work (62%). This is to be expected based on the high education levels and employment profiles in Lagos. There are notable differences between males and females regarding typical work trips, however. Some 68% of trips by males were work related, while this was only 58% for females. Shopping trips in Lagos were profoundly higher amongst females, 15% of typical trips, verses 5% for men. Males in Lagos made more social trips (11%) than their female (8%) counterparts (see Fig. 5). Looking at Blantyre, 29% of typical trips by men were to work, while for women, just 10% of typical trips were to work. The highest proportion of trip purposes in Blantyre was reported to be shopping (26%). Compared to Lagos, the proportion of shopping trips for males (30%) was much higher than that for females (23%). Conversely, in Blantyre females made more social trips (15% vs. 9% for their male counterparts), while 'other' trip purposes were a major portion of trips, i.e., 33% of Lagos, male lagos, female Blantyre, male Blantyre, female "'" 40% 80% l(X'" • Tvpiul _ Non·typiGilI Fig. 4. Typical Trip Profile (Blantyre n = 685; Lagos n = 786). M. Vanderschuren et aL 8CM 100% . Wor\: • EduU'ion .$/loppinS • Social • H"alth vi";. for 5Omei>ody "I"" • H .. alth ,,;'