Women’s Issues in Transportation: Summary of the 4th International Conference, Volume 2: Technical Papers (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

105 Youth Transport, Mobility, and Security in Sub-Saharan Africa The Gendered Journey to School Gina Porter and Kate Hampshire, Durham University, United Kingdom Albert Abane, University of Cape Coast, Ghana Alister Munthali and elsbeth Robson, University of Malawi, Malawi Mac Mashiri, CSIR, Pretoria, South Africa Augustine Tanle, University of Cape Coast, Ghana This paper draws on empirical data from a three-country (Ghana, Malawi, and South Africa) study of young peo- ple’s mobility to explore the gendered nature of children’s journeys to school in sub-Saharan Africa. Gender differ- ences in school enrollment and attendance in Africa are well established: education statistics in many countries indicate that girls’ participation in formal education is often substantially lower than boys’, especially at the sec- ondary school level. Transport and mobility issues com- monly form an important component of this story, though the precise patterning of the transportation and mobility constraints experienced by girls and the ways in which transport factors interact with other constraints vary from region to region. In some contexts, the journey to school represents a particularly hazardous enterprise for girls because they face a serious threat of rape. In other cases, girls’ journeys to school and school attendance are ham- pered by Africa’s transport gap and by cultural conventions that require females to be responsible for pedestrian head loading (transporting loads such as food crops or fuel on the head) and other work before leaving for, or instead of attending, school. evidence comes from a diverse range of sources, but the data used here are principally drawn from a survey questionnaire conducted with approximately 1,000 children ages 7 to 18 years across eight sites in each country. The aim of this study is to draw attention to the diversity of gendered travel experiences across geographi- cal locations (paying attention to associated patterns of transport provision); to explore the implications of these findings for access to education; and to suggest areas in which policy intervention could be beneficial. This paper examines the gendered journey to school in sub-Saharan Africa. Gender differences in school enrollment and attendance in Africa are well estab- lished: education statistics in many countries indicate that girls’ participation in formal education is often substan- tially lower than boys’, especially at the secondary school level. Transport and mobility issues commonly form an important component of this story, though the precise patterning of the transportation and mobility constraints experienced by girls, and the ways in which transport factors interact with other constraints, vary from region to region. On the basis of a large set of empirical data gathered by the authors in a research project on chil- dren’s mobility in Ghana, Malawi, and South Africa, this paper explores the diversity of gendered travel experi- ences across geographical locations, paying attention to associated patterns of transport provision. The research covers 24 urban and rural sites across Africa. Issues of security form a key theme in some of the locations where data were collected, especially where girls are concerned: they may face a very real threat of rape. In other cases girls’ journeys to school and school attendance are ham- pered by Africa’s transport gap and cultural conventions that require females to be responsible for pedestrian head loading (transporting loads such as food crops or fuel on the head) and other work before leaving for (or instead of attending) school. There is insufficient room in this paper to examine in depth the role that girls’ household duties associated with Africa’s transport gap play in shaping gendered patterns of travel to school, but it is important to note their presence and contextual significance for the daily travel patterns described below.

106 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 Following a short review of background literature and methods, the authors present and comment on the compar- ative survey data for the locations in which they worked. The implications of the findings are examined for gendered patterns of access to education, and areas in which policy intervention could be beneficial are suggested. BackgRound Primary education enrollment figures have been boosted substantially across sub-Saharan Africa by the emphasis on “free” universal primary education in the Millennium Development Goals (MDGs), but attendance figures often remain substantially below enrollment. Children’s time spent at school has to be balanced against the opportunity costs of its alternative use in contribution to household production and reproduction. Additionally, although fees have been abolished, parents are still expected to fund other items such as uniforms, books, furniture, parent– teacher associations, and so forth. Other potential con- straints include poor school quality and lack of access to credit, though these vary in importance between urban and rural areas (ersado 2005). The significance of the time, effort, and costs of transport incurred in getting to school has been little considered in the literature. Once children have completed primary school or junior secondary school (where this latter middle stage occurs), a move to senior secondary education may be contemplated but is often not achieved. The barriers to enrollment are more substantial at this level since fees are imposed. Additionally, secondary schools are fewer in number than primaries and tend to be located in major service centers (Porter 2007). Consequently, in addition to fees and the related school-based costs of uniforms, books, and so forth, there are likely to be substantial travel costs, a very long daily walk, or costs of accom- modation at or close to the school. In more remote rural areas even primary school enrollment and attendance may be affected by travel distance, since schools cannot normally be provided in every settlement, but at the sec- ondary level the barriers imposed by distance are usually much increased. Studies directly concerned with travel to school in Africa are rare, especially outside of South Africa. early work in Uganda (Gould 1973) showed how poor trans- port services forced most children to walk to primary school; secondary school children usually had to live away from home due to the distances involved. This situ- ation is still common across Africa. In Ghana, Avotri et al. (1999) found that the closer the secondary school, the more likely it is that children will be sent to primary school, because continuity of the child’s education is feasible. Their work also shows (94–95, 165) how long walks to school due to lack of or high cost of transport and associated lateness problems encourage late “over- age” enrollment (especially of girls), truancy, and early drop out. In South Africa, a number of studies indicate the scale of the problem of travel to school. A 1998 sur- vey of rural KwaZulu Natal schools found that 75% of secondary school walkers walked over 3 km to school, while 43% of primary school walkers walked over 3 km: at least 280,000 children in this region walked for over 1 h, one way, to school. An associated survey of school principals found 60% of pupils were often late, 58% sometimes absent, and 70% of pupils were often tired at school, due to long walking distances to reach school each day (Mahapa 2003). The 2003 South Africa National Household Travel Survey adds further support to this picture: 76% of “learners” were reported to walk to their educational destination and almost 3 million out of the 16 million total (especially those located in more rural provinces) spent more than an hour a day walking to and from educational institutions (RSA Department of Transport 2005). Long journeys to school, when cou- pled with required contributions to household labor, are likely to have a strong impact on attendance. However, Filmer (2007) suggests from his analysis of data for rural areas of 21 poor countries (including some of the poor- est in Africa) that simply building schools will not bring large increases in school participation rates: a combina- tion of interventions may often be essential to improved participation. Children’s school attendance and performance is further impeded by the widespread household trans- port gap, notably lack of piped water and electricity for cooking. Girls, in particular, are often required to carry water, wood fuel, and other loads, as well as perform- ing other household duties, both before and after school. This delays the time when they leave home, not only increasing the likelihood of late arrival at school (with consequent punishment from their teachers), but also leaving them exhausted during lesson time. Poor roads and unreliable, costly, or nonexistent transport services may add to their problems on the journey to school. In some countries the school transport situation is further complicated by the substantial expansion of private education. Perceived deficiencies in the quality of education provided in state schools may persuade even relatively poor families to send their children to a more distant private school rather than the local state school. As Lewin and Sayed (2005) observe, the limited availability of places at state secondary schools for the growing numbers of children graduating from the free primary education promoted by the MDG priorities may also encourage this trend. Since girls’ school enrollment rates are considerably lower than boys’ across much of sub-Saharan Africa, transport impacts on girls’ education are of particular interest. Low enrollment and attendance among girls

107YOUTH TRANSPORT, MOBILITY, AND SeCURITY IN SUB-SAHARAN AFRICA are in part related to their particularly heavy household duties, but also associated with cultural perceptions regarding the (limited) value of girls’ education, and per- ceived dangers for girls who have to travel a long dis- tance to school or board away from home. A study in Niger using DHS surveys indicates that there are only 41 girls per 100 boys at school in rural areas, compared with 80:100 in town, and that distance of home from school is a key factor (Department for International Development 2001, 7). Improvements to road access and transport availability can probably make a significant impact on girls’ attendance in some contexts, as research in Morocco indicates (Khandker et al. 1994; Levy and voyadzis 1996 cited by African Union et al. 2005). methodology This paper draws on a survey of approximately 1,000 children aged 7 to 18 years in each of the research coun- tries (n = 2,967). In the discussion, statistics are com- pared from the quantitative survey for four locational types in each country: 1. Remote rural (RR) with no school and few, if any, basic services. Poor access along unpaved road with very limited or no transport services. 2. Rural with services (RS), that is, with at least a pri- mary school and some kind of health service, if only as an occasional visiting facility. Poor access along unpaved road with some (fairly limited) transport services. 3. Periurban (PU) or small town within daily reach of an urban center, with some residents traveling daily into the urban center. 4. Poor, high-density urban (U) neighborhoods. each category includes two sites per country from dif- ferent agroecological zones. A random sample of approximately 125 children per settlement was obtained (see www.dur.ac.uk/child .mobility). The parent or caregiver was asked for basic information about the selected child. The subsequent interview with the child was conducted wherever pos- sible out of hearing (but within sight) of the parent and other family members. The survey was preceded and accompanied by substantial qualitative research not only by academic researchers but also by 70 child researchers, who interviewed their peers, made photo diaries, and in some cases showed the authors their routes to school. The child and adult qualitative research played a crucial role in shaping the survey design, specific questions, and the interpretation of findings. findings Gender and Schooling The vast majority of the children and young people interviewed in the survey were enrolled in school, which suggests an encouraging trend, probably at least in part a reflection of efforts associated with the MDGs. Gen- der patterns of enrollment were as one might expect for Ghana (82.4% of girls enrolled compared with 92.6% of boys across the full-country sample of children) and South Africa (90.3% of girls compared to 95.4% of boys), but in Malawi, where overall enrollment figures in rural areas are lower than in Ghana and South Africa, a smaller pro- portion of boys were enrolled than girls among the total Malawi sample (82.2 % girls compared with 74.2% of boys). However, it is probably more useful to focus on enrollment of children between the ages of 9 and 15 years, rather than the total sample when looking at enrollment statistics, since it is common for younger children to be enrolled fairly late, especially in more remote areas, while children over 15 are beyond the age of compulsory educa- tion. A breakdown of this data for 9- to 15-year-olds, by settlement type, provides useful detail (Table 1). Overall, it would appear that enrollment is directly linked to settlement accessibility only in Malawi, where for both for girls and boys there is a clear pattern of increased enrollment from remote rural to urban settle- ment types. The significantly higher figures for girls than for boys in Malawi remote rural settlements can be related to the fact that some boys are involved in herding, char- coal burning, and other livelihood activities in the remote rural areas. In rural Ghana and South Africa, enrollment figures are substantially higher than in Malawi. Overall, in both these countries there is less distinction in acces- sibility terms than in Malawi. In Ghana girls’ enrollment TABLE 1 Percentage of Children Ages 9–15 Years (Inclusive, n = 2,180) Enrolled in School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 96.3 98.7 88.9 90.7 75.8* 58.5 RS 98.9 98.7 90.4 95.8 76.9 87.6 PU 98.8 100 89.5* 97.1 90.8 89.2 U 95.0 98.2 94.9 93.3 96.6 93.3 *P (chi-square) < .05.

108 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 is significantly lower than boys in the periurban loca- tions surveyed, perhaps because in this settlement type girls’ household contributions through retail and other employment in a nearby urban economy are highest. School attendance figures are actually much more important than enrollment figures in terms of children’s access to education: very many children who are enrolled spend many days out of school. Unfortunately, atten- dance data for the last week across all three countries could not be satisfactorily compared because in some locations, particularly in South Africa and Malawi, the survey had to be conducted during the school holidays. Travel to and from School Travel Patterns by Gender: Attending the Nearest School In the following discussion the focus is on all children in the survey who travel to school (i.e., not only those between 9 and 15 years). The vast majority (95% or over) of both boys and girls interviewed in the survey travel daily to school in all three countries. The majority interviewed also attend the school (relevant to their edu- cational level) closest to home, but there are important distinctions according to location in Ghana and Malawi, where both boys and girls were particularly likely to attend the closest school in rural settlements with ser- vices. This is because such settlements all have at least pri- mary and usually also junior secondary school provision, and are sufficiently far from other centers to disincline parents from choosing to send their children to a school elsewhere, since this is likely to involve substantial travel costs. The picture is less clear in South Africa. In all three countries, but especially Ghana, urban children are less likely than others to attend the nearest school: there are opportunities to select other schools which, although not the nearest for their particular level, are clearly within reasonable travel distance, and can usually be reached on foot. They are selected for a variety of reasons, includ- ing factors such as family members’ having attended the school, religion, proximity to the family business, or government allocation (in the case of secondary school). Urban figures regarding attendance at the nearest school are highest in Malawi, presumably because this is the poorest, least developed of the three countries, and thus pressures to educate as close to home as possible and at minimum cost are highest, even in urban areas. Gender distinctions in each country regarding atten- dance at school closest to home are relatively small in urban areas and rural areas with services (where schools will be close), but in remote rural areas in Ghana and particularly in Malawi, a considerably higher propor- tion of girls than boys attend the closest school. This presumably reflects poverty levels in remote areas, girls’ workloads, and concerns about girls’ safety during the journey to school, an issue that is considered later in this discussion. In South Africa, although there are serious safety concerns regarding girls’ travel to school, there is relatively more availability of scholar transport in remote rural areas (see below) (Table 2). Daily Travel Times Children were asked to estimate the time it took to travel to school on the most recent school day. In Ghana, just over half of all girls who responded to this question (n = 1,998) estimated they reached school in 15 min or less (50.6%). Figures for girls in Malawi and South Africa were lower (34.3% and 28.3%, respectively). Corre- sponding percentages for boys were somewhat lower than for girls in Ghana (44.6%) but very similar in Malawi and South Africa (35.3% and 28.5%, respectively). In all three countries roughly 20% of boys and girls esti- mated their journey at between 46 and 90 min. Journeys over 90 min were few in Ghana and Malawi (under 3% of girls and boys in Ghana, under 5% in Malawi), but reported by 7.8% girls and 8.4% boys in South Africa. A review of differences across different settlement types in each country adds significantly to the broad pic- ture. As one would expect, the survey data for estimated daily travel time to school in Malawi shows the longest journeys in remote rural areas, where a majority of chil- dren, both boys and girls, estimate that their journey takes over three-quarters of an hour. By contrast, jour- neys in rural centers with services tend to be the shortest; TABLE 2 Percentage of Children Attending the School Closest to Home South Africa (n = 879) Ghana (n = 865) Malawi (n = 784) Location Female Male Female Male Female Male RR (n = 569) 88.7 90.9 90.7 86.7 92.0* 80.9 RS (n = 650) 91.1 89.8 97.3 97.4 98.9 98.1 PU (n = 629) 92.8 87.8 95.2 90.6 95.9 86.9 U (n = 680) 66.2 67.9 56.6 59.1 72.3 70.7 Note: All children ages 9–18 years currently attending school, n = 2,528. *P (chi-square) < .05.

109YOUTH TRANSPORT, MOBILITY, AND SeCURITY IN SUB-SAHARAN AFRICA not surprising given that this is the settlement type in which the most children are likely to attend the nearest school. Journey time is more varied in both urban and periurban locations, reflecting proximity to schools in urban areas, but also the tendency to attend a school other than the closest, especially in urban locations. Dif- ferences in gendered patterns are relatively small, though more boys (especially in remote rural Malawi) travel distances requiring long journeys (over one and a half hours) compared to girls (Table 3). Daily travel times for South Africa are similar to Malawi in terms of distribution across geographical locational categories, with the majority of the longer journeys in remote rural locations and the majority of short journeys in rural settlements with services. Gen- der distinctions appear relatively small. The data accord with the 2003 South Africa National Household Travel Survey conclusion that 3 million out of the 16 million total learners (especially those located in more rural provinces) spent more than an hour a day walking to and from educational institutions (RSA Department of Transport 2003) (Table 4). Daily travel times for Ghana as a whole are relatively short, particularly when compared to South Africa. The preponderance of short journey times in Ghana is in peri- urban sites, even more than in rural settlements with ser- vices. This probably reflects the high density of schools in the periurban areas concerned, including private schools. No particular gender pattern is discernable (Table 5). Travel Mode to School walking dominates as the mode of transport to school in all three countries, in all types of settlement, and across both genders. In Ghana and Malawi it is almost the exclusive mode of transport to school in all settlement types. In Ghana 98.6% of girls and 97.4% of boys had walked to school the previous school day, and 99.3% of girls and 99.1% of boys in Malawi. Bicycle use to school (either as cyclist or riding pillion) was reported by only one boy and not even one girl in the Malawi sample, and by three boys and just one girl in Ghana. This may seem surprisingly low, especially in view of the fact that when asked whether they knew how to ride a bicycle, so many children said they could (Ghana: 58% girls, 87% boys; Malawi: 44% girls, 72% boys; South Africa: 48% girls, 82% boys). However, many children learn to ride a cycle through borrowing a family member’s bicycle or by hiring a cycle for a few minutes at a time (called kobo- kobo in southern Ghana). If the family has a bicycle, it is unlikely to be available to be parked at school all day TABLE 3 Estimated Daily Travel Time to School on Most Recent School Day: Malawi 46 min to 1 h 31 min to Under 15 min (%) 16–45 min (%) 1 h 30 min (%) 2 h 30 min (%) Over 2 h 30 min (%) Location Female Male Female Male Female Male Female Male Female Male RR 8.8 5.4 17.6 19.6 63.2 57.1 10.3 17.9 0 0 RS 67.7 59.8 29 36.6 1.6 1.2 1.6 1.2 0 1.2 PU 42.2 46.8 37.3 39.2 20.6 11.4 0 2.5 0 0 U 25.2 18.1 57.5 66.7 16.5 13.9 0.8 1.4 0 0 TABLE 4 Estimated Daily Travel Time to School on Most Recent School Day: South Africa 46 min to 1 h 31 min to Under 15 min (%) 16–45 min (%) 1 h 30 min (%) 2 h 30 min (%) Over 2 h 30 min (%) Location Female Male Female Male Female Male Female Male Female Male RR 7.7 4.6 21.2 29.6 57.7 50.0 13.5 15.8 0 0 RS 60.4 53.3 28.8 33.7 8.8 9.0 2.1 3.1 3.1 1.7 PU 49.0 47.8 34.6 37.3 15.2 13.8 1.1 1.1 0 0 U 37.2 38.1 48.8 51.4 12.8 10.0 0.5 0 0.3 0 TABLE 5 Estimated Daily Travel Time to School on Most Recent School Day: Ghana 46 min to 1 h 31 min to Under 15 min (%) 16–45 min (%) 1 h 30 min (%) 2 h 30 min (%) Over 2 h 30 min (%) Location Female Male Female Male Female Male Female Male Female Male RR 9.5 4.8 26.2 45.2 58.3 45.2 6.0 4.8 0 0 RS 62.8 49.1 17.7 27.2 16.8 18.4 2.7 5.3 0 0 PU 72.5 58.8 22.5 32.4 3.9 8.1 1.0 0.7 0 0 U 49.2 54.3 40.6 35.9 9.4 9.8 0.8 0 0 0

110 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 while the child is in lessons. Fears regarding the vulner- ability of bicycles to theft at school or on the journey to and from school were also expressed. Motorized transport use was similarly remarkably low: in Ghana three girls and three boys had travelled to school by motor taxi, while in Malawi only one boy and two girls had taken a bus or minibus to school the previous school day. The Ghana data accord closely with earlier findings in Ghana’s coastal region (Porter and Blaufuss 2003). In South Africa the travel to school pic- ture is a little more varied, however, and worth examin- ing in detail since survey data can be compared with data from the national travel survey briefly reported above (Table 6). Although walking dominates among both genders in South Africa (as in Ghana and Malawi), in remote rural areas of South Africa, the availability of motorized trans- port services for part of the school journey has clearly proved advantageous for some children, both boys and girls. Cycle use is remarkably low, however. There have been sporadic efforts to expand the use of bicycles among school pupils in poorer areas of South Africa under the Shova Kalula program ((Mashiri et al. 2001; Mahapa 2003), but this has clearly had no influence in the study settlements. According to the 2003 South Africa National Household Travel Survey, 76% of “learners” were reported to walk to their educational destination. This figure is still far exceeded in the (overwhelmingly poor) study areas where the present project focused. Dangers Faced on the Journey to and from School In the survey children were asked about problems faced on the journey to and from school. They were first asked to identify the principal danger (if any) they faced as they travelled to school. In South Africa the three prin- cipal dangers identified were (in rank order) risk of attack from thieves or thugs (8.4%); rivers and streams to cross (8.2%); and rough terrain travelling to school (5.1%). In Ghana snakes were ranked first by children as the biggest danger (14.3%), followed by rough terrain (5.1%) and dangerous taxis (4.9 %). In Malawi, rough terrain ranked first (5.2%), followed by crossing rivers and streams (4.4%) and harassment or verbal abuse by drunkards (4.3%). In each country, a significant propor- tion of children said they did not experience any major dangers on the journey to and from school: 56.3% in Ghana, 57.9% in Malawi, and 46.1% in South Africa. These figures accord with the qualitative data in that they also suggest that children in South Africa report facing greater dangers on their journey to school in comparison with children in Malawi and Ghana, though as shown below, the degree to which children perceive themselves to face overall danger varies between rural and urban environments in all three countries, and to a smaller extent by gender. The figures above indicate that the children who feel safest as they travel to and from school in South Africa reside in rural locations with services, where girls feel almost as safe as boys. These are the settlements in South Africa where children’s travel time to school is shortest. In urban areas slightly more boys than girls feel safe, per- haps related to girls’ perceived (and actual) risk of rape in urban locations (see below). In Ghana the children who appear to feel safest as they travel to and from school are those resident in urban and periurban locations. A greater proportion of Ghanaian girls than boys feel safe in urban locations but a greater proportion of boys feel safe than girls in rural locations. In Malawi the urban children reported the least dangers on the way to school, while those in remote rural communities experienced the most hazards; in all Malawi locations, a greater propor- tion of boys than girls feel safe (Table 7). The types of danger experienced also vary considerably between locational types and in some cases by gender. The respondents were asked whether they were exposed to particular specific risks: rough terrain, streams that are difficult to cross, risk of attack from people, risk of harassment (verbal abuse), risk of rape, fear of animals, dangerous vehicles, and supernatural risks. As might be anticipated, there is substantially less difference in gen- der perceptions of danger for some of these risks than for others. Dangers Associated with Topography and Rivers Rough terrain, unsurprisingly, is considered a more fre- quent hazard in remote rural areas than other locations across all three countries, with the highest perception TABLE 6 Major Modes of Transport to School on Most Recent School Day: South Africa walked Only (%) Bus, Minibus, or Combi (%) Bicycle (%) Location Female Male Female Male Female Male RR 84.7 82.8 15.3 15.5 0 0 RS 80.3 88.5 16.6 11.5 0 0 PU 93.2 88.5 1.7 3.8 1.7 0 U 87.4 85.7 7.4 10.2 0 0

111YOUTH TRANSPORT, MOBILITY, AND SeCURITY IN SUB-SAHARAN AFRICA of danger reported in remote rural South Africa (Table 8). In one of the two remote rural South African settle- ments, in eastern Cape, children must cross a mountain even to reach primary school. Gender patterns for the high-risk areas relating to rough terrain—remote rural sites—suggest that a slightly greater proportion of boys perceive topography as an issue than girls in all three countries. This may be because for girls other dangers seem much more important and by contrast rough ter- rain is considered a less significant impediment to travel, or possibly because girls are used to rough terrain as they are exposed to it when undertaking tasks such as firewood collection. It is also likely to be related to the statistics (see above) indicating that in remote rural areas of Malawi and Ghana, a higher proportion of girls than boys attend the closest school. Boys from remote rural areas who are sent to more distant schools rather than the school closest to home are very likely to encounter additional hazards of rough terrain along their longer route. The significant difference between girls’ and boys’ perceptions of terrain hazards in Malawi rural with ser- vices settlements will need further investigation, but may be related to boys’ play or work diversions on the way to school. Stream- or river-crossing dangers present a similar pat- tern to rough terrain. In South African remote rural sites they present a major danger to many children. In Malawi remote rural sites they are also a considerable problem for some children, but in the Ghana sites, stream crossings mostly present a relatively insignificant issue for children on their way to and from school. However, the difference between girls and boys in Ghanaian rural with service set- tlements is significant and may reflect some deviation from the most direct route to school by boys associated with play (as suggested also in the case above concerning terrain and Malawi boys in rural areas with services). The otherwise relatively low reportage of dangers associated with rough terrain and rivers in Ghana is not surprising given that the two Ghana project study regions, Cape Coast and Sunyani (and indeed, the country as a whole) have much less rug- ged topography than Malawi’s Shire Highlands or South Africa’s eastern Cape study regions (Table 9). Traffic Hazards Traffic on the journey to and from school mentioned by children as a danger included tipper trucks, taxis, mini- TABLE 7 Percentage of Children Who Experience No Dangers on the Journey to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 20.6 24.4 31.2 43.8 41.9 44.9 RS 72.6 75.4 59.5 49.5 57.0 66.0 PU 39.5 38.1 62.0 58.0 52.6 56.1 U 44.3 47.8 70.3 64.2 67.9 69.7 TABLE 8 Percentage of Children Reporting Rough Terrain as a Danger in Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 22.7 26.4 13.1 16.7 20.9 21.7 RS 6.6 13.0 10.6 15.5 5.1* 18.2 PU 11.7 13.0 13.7 13.2 7.9 11.0 U 3.6 4.2 5.4 3.3 1.4 5.3 *P (chi-square) < .05. TABLE 9 Percentage of Children Reporting Crossing Rivers or Streams as a Danger in Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 24.8 22.1 0 1.2 18.6 15.9 RS 3.5 6.5 3.5* 10.3 2.6 3.0 PU 1.5 1.9 1.0 0 1.8 6.1 U 0.5 1.9 0.8 2.2 0.7 1.3 *P (chi-square) < .05.

112 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 buses and combis, even bicycles (the latter especially in Malawi). Children refer to traffic dangers more frequently in urban and periurban areas than in rural areas across all three countries, as might be expected, given associ- ated traffic density patterns (Table 10). In South Africa and Ghana, even children living in remote rural areas are likely to encounter traffic dangers on the journey to school because their schools are located in centers where some vehicular traffic exists. Malawian rural children do not record motorized traffic as a danger, even in the two rural centers with services, because these centers are located off-road in a country where poverty is such that rural vehicular traffic is limited. Malawian figures for urban areas are also relatively low compared to Ghana and South Africa. So far as gender patterns are concerned, girls are significantly more aware of traffic dangers than boys in rural areas in Ghana, which accords with the fact that boys tend to have more traffic-related accidents than girls (see Porter and Blaufuss 2003 on Ghana). In urban areas, however, across all three countries, the data sug- gest that boys are slightly (South Africa) or considerably (Ghana and Malawi) more aware of traffic danger than girls. However, this does not appear to have translated into reduced traffic accident statistics for boys. Animal Hazards The child researchers first alerted the authors to the sig- nificance of animals and reptiles (e.g., biting dogs and snakes) as a serious potential hazard for some children, and consequently questions about animals were incorpo- rated in the survey questionnaire. In the survey children included dogs, snakes, cattle, wild pigs, worms, lions, rats, and donkeys among the animal dangers they might encounter on the journey to and from school. while no child is likely in reality to meet a lion, the imagined pros- pect of such a meeting can be a very significant issue (Porter et al. in press). The data below show that animal- related dangers were reported in remote rural areas in Ghana and South Africa; in Ghana, in particular, snakes were a major concern. Snakes were reported to be a par- ticular problem for children passing through less habited areas, notably fields or uncleared bush areas on the way to school. The lower percentages for remote rural Malawi are more difficult to explain. There is no clear gender pattern in the data (Table 11). Attacks from People, Including Rape, Other Forms of Attack, and Verbal Harassment Hazards presented by people were recorded by many children on their journey to school. Among these “peo- ple,” children sometimes specified bandits, older boys, murderers, kidnappers, thieves, fighting with friends, men who chase, thugs, unruly pupils, and children who throw stones. The highest proportion of children observing general danger of attack from people was in rural areas. In urban areas perceived risk of attack is highest in South Africa and lowest in Ghana. In Ghana and Malawi relatively few children, particularly in urban and periurban areas, perceived danger from people. However, gender patterns show a significant difference in perception of dangers in Ghana periurban areas, where boys are more concerned than girls, and in Malawi rural with service settlements, where girls are more concerned than boys (Table 12). TABLE 10 Percentage of Children Reporting Dangerous Vehicles as a Hazard on Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 6.2 2.4 13.1* 2.4 0 0 RS 5.4 4.7 3.5 1.7 0 0 PU 20.9 21.6 16.7 22.8 20.2 13.4 U 20.9 23.6 22.5 32.6 11.3 18.4 *P (chi-square) < .05. TABLE 11 Percentage of Children Reporting Dangerous Animals as a Hazard on Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 36.8 40.9 53.6 56.0 15.1 17.4 RS 22.7 26.3 27.7 37.9 22.8 22.0 PU 15.4 13.8 5.9 11.8 13.3 8.5 U 3.3 4.7 6.2 7.6 0.7 1.3

113YOUTH TRANSPORT, MOBILITY, AND SeCURITY IN SUB-SAHARAN AFRICA However, the data are perhaps more usefully examined below in terms of specific dangers from rape. Figures for Ghana in the generalized category of risk from attack by people were substantially higher than for rape or for ver- bal harassment (below): children in Ghana often appear to have included risks like fights with friends and bully- ing from older boys in this general category. The specific danger of rape on the journey to school was indicated by a much higher proportion of South African children (12.9% of the total sample) than either Ghanaians (0.4%) or Malawians (0.2%). However, the sentiments expressed by a fathers’ group in Kanyola, Malawi, were common in qualitative interviews across the southern Africa sites: “we fear girls will be cheated on or get raped on way to school.” The figures for Ghana are particularly low and are supported by qualitative evi- dence, which suggests parental perceptions of children’s travel dangers are associated more with getting lost than with “stranger danger” (see Porter and Blaufuss 2003). In southern Africa, rape presents a particularly great danger given the high incidence of HIv and AIDS in the region. In South Africa, boys are possibly more exposed to dangers of rape than elsewhere in Africa, due to long- standing cultural patterns of male co-residence in male mine-labor communities. Nonetheless, significant gender differences regarding perceived danger of rape are found in South African remote rural, periurban, and urban sites (and in Malawi urban sites) (Table 13). Risk of harassment and verbal abuse on the journey to school were reported to come from diverse people in various contexts: drunkards, strangers, cannabis users, and minibus conductors were all singled out as potential sources of verbal abuse. Such harassment is reported par- ticularly widely as a hazard by boys and girls in Malawi. Like fear of rape, it is rarely reported in Ghana, and partic- ularly rarely by Ghanaian boys. In both Malawi and South Africa, a higher proportion of girls in urban and periur- ban areas report harassment or verbal abuse than boys: in Malawi’s periurban areas this gender pattern is statistically significant (Table 14). Despite all the dangers and difficulties the children reported, remarkably few children, girls or boys, are accompanied by adults on the journey to school, what- ever their age. Most children travel to and from school in TABLE 12 Percentage of Children Reporting Risk of Attack from People on Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 22.3 21.2 25.0 17.9 17.4 21.7 RS 13.2 11.9 13.3 20.7 19.0* 7.0 PU 10.3 14.1 2.0* 8.8 7.0 6.1 U 11.6 13.0 2.3 5.4 6.4 6.6 *P (chi-square) < .05. TABLE 13 Percentage of Children Reporting Rape as a Danger Experienced on Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 10.3* 1.4 0 0 2.3 0 RS 1.2 0 0.9 0 1.3 0 PU 8.1* 1.1 0 0 2.6 0 U 7.2* 0.5 0.8 1.1 6.4* 0 *P (chi-square) < .05. TABLE 14 Percentage of Children Reporting Harassment or Verbal Abuse as a Danger or Difficulty on Journeys to and from School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 7.0 6.8 2.4 1.2 12.8 15.9 RS 6.2 6.8 1.8 0.9 13.9 18.0 PU 13.6 4.5 2.0 0 22.8* 11.0 U 11.6 4.2 3.1 0 17.1 9.2 *P (chi-square) < .05.

114 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 groups with their siblings and friends—their parents are at work, on the farm, or otherwise occupied (Table 15). conclusion: significance of findings foR theoRy and puBlic policy Analysis of the empirical data presented in this paper pro- vides a comparative cross-country and locational perspec- tive on boys’ and girls’ travel to and from school hitherto unavailable. The three countries present rather differ- ent contexts, which are reflected in the findings. Some of the selected study areas of Malawi (Blantyre region’s Shire Highlands) and South Africa (eastern Cape region) exhibit considerably more rugged topography than the sites in Ghana, for instance, which are all relatively flat: consequently the journey to school for children in remote rural areas in South Africa and Malawi tends to be more hazardous than for children in Ghana in terms of terrain and flooded rivers. This confirms the rather predictable conclusion that topography presents variable challenges for children on their school journeys, but it is important to take account of the varied terrain of African land- scapes in making public policy. with reference to diverse levels of economic development, only South Africa has the resources to provide some degree of motorized trans- port provision in rural areas, including dedicated school buses. Malawi and South Africa are both experiencing a high prevalence of HIv and AIDS, which not surprisingly translates into concern about exposure to attack from people and, in particular, fear of rape. Some basic characteristics of the journey to and from school are remarkably similar across the three countries. walking is the dominant mode of travel throughout all regions and in the remote rural areas of all three coun- tries children walk particularly long distances to and from school each day. The authors have accompanied children on their journeys from school and can attest not only to the very real physical difficulties and dangers they face but also to the less concrete fears they express on these journeys (Porter et al. in press). Passing by a lonely graveyard or sacred grove where witches, lions, and robbers are reputed to lie in wait is a fearsome task if the child is delayed and has to walk home alone. As one pupil in her early teens observed, as she accompanied the authors on her long walk home from school in a remote area of the Shire Highlands region in Malawi during one of the pilot studies, “There are so many problems [here]. You meet a very long person and I suspect it’s a witch, and there are dogs that chase us and fierce animals— lions and hyenas—and they bite us. Three children were once bitten—one by a lion and she died, and two boys were bitten by hyenas, but they are still alive.” Children in rural centers with services appear to be best served in terms of short length of school journey, but this may also reflect a lack of available alternative schools in the vicinity. In urban and periurban areas, although the number of schools is much higher and journeys to school might be expected to be short, more parents choose to send their children further than the closest school appropriate to their level, or are required to do so by school allocation systems. Such urban jour- neys may impose additional stresses, especially on girls in South Africa and Malawi, who report high levels of harassment and, to a lesser extent, fear of rape. Again, the findings are supported by qualitative research with children and their parents and teachers, including the child research collaborators’ own findings. Cecilia observes: “we have to walk in groups because there are boys who are not schooling who take our money and mobile phones” (Cecilia, 18 years, Umtata, South Africa). For Susan, a 17-year-old secondary school pupil in winterveld in North west province, South Africa, the dangers are even greater. She walks daily about 5 km across an uninhabited bushland area to school: “I fear people who hide in the bush [waiting] for us. They wait for us in the bush and as we walk, especially when you are alone, they grab your school bag and all the belong- ings that you have with you . . . . The bush is bad, you cannot see people hiding or seeking you . . . . It is even worse to cross the bush at night. There are so many rapists there at night and a lot of drunk people.” The authors accompanied Susan on one of her journeys; the next day a girl from the same school was raped along this route. Some boys also fear attack, as the data have illustrated, but for girls the perceived risk appears con- siderably higher, with likely impacts on broader pat- terns of mobility. Several possibilities exist to improve safe school access, especially for girls. These include TABLE 15 Percentage of Children Accompanied by a Parent or Other Adult on the Journey to School South Africa Ghana Malawi Location Female Male Female Male Female Male RR 0 0 0 0 0 0 RS 1.6 0.4 3.5 0.9 0 0 PU 1.1 0.7 0 0.7 1.8 1.2 U 1.4 1.4 0.8 1.1 0.7 1.3

115YOUTH TRANSPORT, MOBILITY, AND SeCURITY IN SUB-SAHARAN AFRICA 1. expansion of girls’ boarding house provision and general boarding provision, especially at primary schools. Boarding is more often available at secondary schools, but provision is inadequate (especially in Malawi). Boarding is also needed at the primary level, particularly for girls who live in remote rural areas. The threat of rape is such that parents delay sending their daughters to school and curtail their attendance. Boarding may reduce children’s domestic work burdens during the school week, and the more likely availability of electricity (commonly absent in remote rural settlements) would enable them to com- plete their homework. The loss of child contributions to domestic production and reproduction would inevita- bly affect families, particularly in rural areas. A holis- tic approach will be needed that incorporates not only expansion of girls’ school boarding provision, but also pays attention to labor-saving interventions, especially improved water provision, which will reduce domestic work demands in home villages. 2. The walking bus is used in western contexts to help children to walk safely to school in regions where traffic levels are high and child obesity is a problem (e.g., Kingham and Ussher 2007). However, the walking bus may also offer a route to reducing the dangers of rape and harassment. This suggestion was received with inter- est in the periurban site at winterveld, after an author- accompanied school walk with Susan (described above) and other children emphasized the very real fear of rape among children schooling in that community. Funding is now in place to support a walking bus pilot in collabora- tion with a local NGO. 3. Cycle hire centers based at schools might help to overcome the shortage of cycles available to girls, in particular, for traveling between home and school. This could be particularly effective if accompanied with secu- rity measures for guarding cycles at schools and train- ing the girls in cycle riding, maintenance, and repairs. Though many girls say they know how to cycle, being able to cycle over a long distance on poor roads requires associated knowledge in maintenance and repair. 4. various nontransport interventions aimed at reduc- ing girls’ time poverty, including improved availability of water supplies, community woodlots, and grinding mills. These interventions would reduce girls’ time poverty by freeing them from some of the tiring and time-consuming duties that interfere with their ability to attend school. 5. Sensitization of teachers and education authorities to transport- and distance-related lateness. Many teach- ers impose extremely harsh punishment on children who arrive late for school, such as corporal punishment, clean- ing the lavatories, sweeping the yard, grass-cutting, or carrying sand for school building. In some cases children reported being sent directly home by teachers because they had arrived at school late, requiring them to retrace their long walk to school, possibly alone: this is the type of punishment that encourages truancy. Teachers whose pupils collaborated in this research admitted that they had not adequately appreciated the problems that children face in getting to school on time, and there was a common response that they would be more sympathetic to lateness among pupils living at a distance from school in future. Above all, however, a stronger focus on gendered transport, mobility, and access issues in the develop- ment policy and practitioner community is essential. The failure to recognize the specific importance of mobility and transport and the ramifications of immobility in the MDGs is clear testament to this omission. In the con- text of the MDGs regarding universal primary educa- tion (MDG 2) and the promotion of gender equality and women’s empowerment through the elimination of gen- der disparity in all levels of education (MDG 3), a firmer and more specific recognition of the roles which distance, perceived and real travel hazards, transport availability, and other mobility factors play in allowing or barring access to school is urgently required. acknowledgments This research on which this paper is based was funded by the U.K. economic and Social Research Council and the U.K. Department for International Development. How- ever, these organizations can accept no responsibility for any information provided or views expressed. The authors extend their grateful thanks to the 70 child researchers, 20 research assistants, and the many respondents. RefeRences African Union. 2005. Transport and the Millennium Develop- ment Goals in Africa. Africa Union and U.N. economic Commission for Africa, with the collaboration of the African Development Bank, world Bank, and european Union. Avotri, R., L. Owusu-Darko, H. eghan, and S. Ocansey. 1999. Gender and Primary Schooling in Ghana. Institute of Development Studies, Sussex, United Kingdom. Department for International Development. 2001. Children Out of School. United Kingdom. ersado, L. 2005. Child Labor and Schooling Decisions in Urban and Rural Areas. World Development, vol. 33, No. 3, pp. 455–480. Filmer, D. 2007. If You Build It, will They Come? School Availability and School enrollment in 21 Poor Countries. Journal of Development Studies, vol. 43, No. 5, pp. 901– 928. Gould, w. T. S. 1973. Movements of School-Children and Pro- vision of Secondary Schools in Uganda. African Popula- tion Mobility Project working Paper No. 3. Department of Geography, University of Liverpool, United Kingdom.

116 wOMeN’S ISSUeS IN TRANSPORTATION, vOLUMe 2 Khandker, S. R., v. Lavy, and D. Filmer. 1994. Schooling and Cognitive Achievements of Children in Morocco. Discus- sion Paper No. 264. world Bank, washington, D.C. Kingham, S., and S. Ussher. 2007. An Assessment of the Ben- efits of the walking School Bus in Christchurch, New Zea- land. Transportation Research Part A, pp. 502–510. Levy, H., and C. voyadzis. 1996. Morocco Impact Evalua- tion Report: Socio-Economic Influence of Rural Roads. Report No. 15808-MOR. Operations evaluation Depart- ment, world Bank, washington, D.C. Lewin, K. M., and Y. Sayed. 2005. Non-Government Sec- ondary Schooling in Sub-Saharan Africa: Exploring the Evidence in South Africa and Malawi. Department for International Development educational Papers. www .sussex.ac.uk/education/1-4-7.html. Mahapa, S. 2003. Integrating Gender into World Bank– Financed Transport Programmes: Case Study South Africa: Shova Kalula. Mashiri, M., v. Lechwiti, and B. Mpondo. 2001. Shova Kalula (Ride Easy) National Bicycle Partnership Pro- gram: Assessment of Phase 1. South African Department of Transport. Porter, G. 2007. Transport Planning in Sub-Saharan Africa. Progress Report 1. Improving Access to Markets and Ser- vices. Progress in Development Studies, vol. 7, No. 3, pp. 251–257. Porter, G., and K. Blaufuss. 2003. Children, Transport and Traffic in Southern Ghana. Revised version of paper pre- pared for International workshop on Children and Traf- fic, Copenhagen, Denmark, 2002. www.dur.ac.uk/child .mobility. Porter, G., K. Hampshire, A. Abane, A. Munthali, e. Robson, M. Mashiri, and G. Maponya. where Dogs, Ghosts and Lions Roam: Learning from Mobile ethnographies on the Journey from School. Children’s Geographies 2010. In press. RSA (Republic of South Africa) Department of Transport. 2005. The First South African National Household Travel Survey 2003. Technical report.