Rural – urban connectivity strengthens agrarian peace: Evidence from a study of gender and motorcycle taxis in Sierra Leone

Rural-based insurgencies disrupted the forest margins of Upper West Africa in the 1990s. A subsequent return to peace was accompanied by strong growth in small-scale trade in foodstuffs and other agrarian produce in high demand in towns. Motor cycle taxis are a feature of this increased rural – urban market integration. It was a mode of transport pioneered by ex-combatants. Where rural women were once attacked by rural young men without job pros-pects press ganged into fighting for the rebels, bike taxi riders now carry them to rural periodic markets, many of which are new since the end of conflict. The study provides an analytical account of these developments, drawing on discussions with villagers in three heavily war-affected localities of Sierra Leone. The evidence indicates that communities divided by conflict have quietly built new cooperative links conducive to peace based on local agricultural production and petty trade.


| RESEARCH QUESTION
Around the year 2000, after a decade of civil wars, pessimism about the future of sub-Saharan Africa was widespread among international commentators. 1 An expanding population was failing to find enough productive work to generate social cohesion, and the region was succumbing to irregular youth militia armed with Cold War-surplus semi-automatic weapons deployed in pursuit of plunder not politics (Collier, 2000;Kaplan, 1994).
A recent article by Frankema and van Waijenburg (2018) draws a line under this imagined dystopia. The civil wars in the coastal Upper West African forest belt ended rather more quickly and completely than many anticipated, throwing a spotlight on new economic forces fostering social reintegration. Among domestic generators of peace, Frankema and Waijenburg cite the growing importance of stronger ties between town and countryside. Growth of provincial towns, in particular, created dynamic markets for agrarian produce fostering economic integration and generating rural jobs.
Rural-urban interconnections are, of course, not new in Africa as a whole. The requirement for the countryside to supply towns has long been an integrative force in and around some of the continent's larger precolonial cities (Guyer, 1987). But the process had not yet reached some of the more inaccessible parts of the African frontier, where towns were few and small and distances too great for head-loading of bulky, lowvalue goods. 2 The war-torn western margin of the Upper Guinean forest was one such region where the pull of towns has been historically weak. Tighter binding together of town and country in the aftermath of conflict in the 1990s was fostered both by displacement towards towns for humanitarian protection and by a major innovation in local transport supply-the introduction of the motorcycle taxi. 3 By this means of trade could now reach districts hitherto denied transport due to the prohibitive cost of building and maintaining roads and bridges for four-wheel motorized vehicles.
These isolated localities had been sought out by rebel movements for rear bases, because they offered protection from the vehicle-mounted weaponry accessible to better armed and supported counter-insurgency forces. Lack of roads had brought war. Now, by an ironic twist, demobilizing rebel fighters brought in motorcycles to improve connectivity in their former stamping grounds, as ceasefire zones were consolidated (Peters, 2007a). Frankema and van Waijenburg (2018) identify three major aspects of this improved rural-urban connectivity: it stimulated economic growth through enhanced consumer demand (and not primarily through supply-side changes in production); it improved productivity by increasing division of labour, sharing of resources and knowledge and providing better access to public goods such as education and health; and (thirdly) it strengthened linkages between agricultural, industrial and service sectors, thus reducing reliance on one sector (such as manufacturing) alone as the engine of economic change. Better domestic rural-urban market integration, therefore, is (in their view) a major reason why Afro-pessimism provided a poor guide to what actually happened in the African interior in the first two decades of the 21st century.
The motorcycle (taxi) is an especially significant technological innovation in this story because it is so often the only mode of motorized transport available in more inaccessible rural areas of sub-Saharan Africa (Ehebrecht et al., 2018). On some rural roads, motorcycle taxis provide 70% to 80% or more of annual passenger and goods transport (Starkey, 2016). The rapid spread of mobile phones also helped to reinforce the impact of motorcycle taxis on rural mobility, as transport could now be called when required, removing the need for rural people to waste time searching for a vehicle. Porter (2016) is probably justified to claim that the expansion of motorcycle taxi services is 1 Encapsulated, at the height of international confusion over UN peace-keeping operations in Sierra Leone, by an item in The Economist magazine, characterizing Africa as 'the hopeless continent'. 2 For a wide-ranging analysis of frontiers regions in African history, see Kopytoff (1987). 3 The low price of Chinese or Indian-made motorcycles was a key driver of their popularity in sub-Saharan Africa. In Sierra Leone, the deplorable condition of the road network due to lack of maintenance and the reduced availability of four-wheeled vehicles (destroyed or driven to safety across the border during the war) further added momentum to motorcycle taxi development. the most dramatic development in rural transport services since the introduction of motorized vehicles in the early twentieth century, because for the first time, many rural dwellers-even the very poor-are able to summon transport whenever they need it.
One of the most significant aspects of this new, much more broadly distributed connectivity, has been to bring rural women, previously tied down by domestic or farm activities, more fully within the scope of market relations of production. 4 Women in areas once cut off by distance and lack of communications are now increasingly present in the intensified petty commodity production associated with market integration in previously highly inaccessible regions (Jenkins et al., 2020a).
There are two question marks hanging over this development, however. The first is that the pioneering group of riders in the war-affected regions were ex-combatants-the highly stigmatized young 'barbarians' blamed for the appalling levels of atrocity and gender-based violence for which the wars of the 1990s became notorious-and some wondered whether war-time habits might die hard. 5 The second is that motorcycle taxis are considered extremely unsafe. Governments tend to tolerate the development of motorcycle taxi transport as an unavoidable evil. Few welcome them, and some try to ban them outright, as occurred in Ghana (Afukaar et al., 2019).
This brings us to the research question implicitly posed in our title-if rural connectivity supports peace and social integration, to what extent is this undermined by the threats to personal safety implicit in the motorcycle taxi as a mode of transportation ridden by putatively reckless male youth?
Our method to answer this question was to choose three areas in rural Sierra Leone that were especially badly affected by civil war and to ask focus groups comprising village women and men to discuss the perceived benefits and drawbacks of the motorcycle transport modality, allowing groups to reflect about issues of safety and gender violence relative to other concerns.
We discuss our methods further below. But an initial remark about focus groups and how we have chosen to present our results is in order. Although used very widely in development research, 6 the focus group is vulnerable to the posing of leading questions. Experience has taught us that the greater the poverty of a group, the greater is the tendency to provide answers thought likely to please a potential enquirer. In our own field practice, we are very clear that we seek cooperation but cannot offer benefits and that our main commitment is to offer to convey a representative account of what the groups choose to discuss. If it is judged that our presentation below simply lists what people said, this is intentional. It reflects an ethical commitment to allow people a clear voice on the topics under consideration, as free as possible of editorial manipulation.

| RURAL ISOLATION IN SIERRA LEONE
Our case study country is Sierra Leone, where a 12-year rebel war (1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002) was followed by a rapid period of post-conflict urban-rural integration of the kind analysed by Frankema and van Waijenburg (2018). Rural trade in the colonial period in Sierra Leone was oriented towards cash crops exported along a government railway (opened fully in 1910 and closed in the 1970s) supplemented by a rudimentary road system reflecting its original purpose (to feed the railway).
Some rural-urban agrarian integration had begun in this period. Riddell (1974), for example, showed that rural periodic markets (known in Krio, the commercial lingua franca, as luma) had begun to appear in those areas served by motor road connections linking diamond mining areas and the capital, Freetown. 4 This remark requires some regional qualification. Women traders have long been prominent in rural-urban trade relations in Lower West Africa. This is less true of the portions of the Upper West African coastal forests with which we are concerned in this paper. 5 Even where ex-combatant motorcycle taxi operators have long retired and have been replaced by a post-war generation of operators, the profession may still be associated in the public's mind with the category that pioneered the activity in the first place. 6 There is a substantial literature on use of focus groups. The approach was first developed by Merton and Kendall (1946). Morgan (1997) provides a practical introduction. Methodological issues are debated in Barbour and Kitzinger (1999) and Bloor et al. (2001). Relevance in African fieldwork is discussed in Jakobsen (2011)  Other areas, without motor road connections, remained devoid of periodic markets, however, and largely dependent on the commercial services of village mercantilists, often of Mande (Malian or Guinean) ethnic origin who used money lending activities to leverage 'distress sale' purchasing deals with farmers with few alternative outlets.
Economic development stalled in the immediate post-colonial period due to major mismanagement of the export economy. The mercantilists, distributed in ones or twos in the larger villages in interior regions without markets, enjoyed an extended life in more isolated areas. They absorbed bad debts, especially in the event of crop failure (Richards, 1990). On the other hand, peasant households were unable to profit when economic conditions were better, and this caused local resentment.
Colonial efforts to link up the country with modern transport connections were significantly incomplete at Independence in 1961. Economic recession in the 1980s, linked to international structural adjustment, resulted in scaledback transport and road maintenance, and poorly connected rural areas reverted to a former state of isolation. At times governments used isolation to punish recalcitrant groups (Richards, 1996).
Post-colonial Sierra Leone was thus a country internally divided into areas of considerable market integration (notably the mining districts), and other more remote areas with poor access characterized by high levels of economic deprivation and political resentment. The patchiness of this market integration is germane to any attempt to understand the dynamics of descent towards armed violence, not only in Sierra Leone but more widely in the forests joining Guinea, Liberia and Sierra Leone (Ellis, 1999;McGovern, 2011McGovern, , 2017Richards, 1996Richards, , 2005. A most dramatic instance of this politically imposed isolation was Kailahun District, in the east of the country, a stronghold of political opposition, where all national connectivity rested on a poorly maintained dirt road often impassable for long periods during the annual rainy season. Failure to maintain this road was a political decision stemming from State House. It is not accidental that the Revolutionary United Front (RUF)-the main protagonist in the 1990s decade-long civil war-chose to launch its rebellion in Kailahun, a populous district where impoverishment and political isolation went hand in hand. Leveraging isolation was thus a tactic of rebellion. The RUF operated through small groups of lightly armed fighters infiltrating along remote bush paths, and advertising their presence by outrage, if and when collaboration was refused. An army dependent on vehicles and heavy equipment found it hard to respond, given the decrepit state of the roads.
The consequences of low-intensity bush war were especially severe for youth and women (Coulter, 2009;Peters, 2011). Sexual slavery (for a young woman) or conscription (for male children and youth) became widespread, as a result of which the towns, more readily defended by conventionally trained government troops, overflowed with women and youth fleeing defenceless villages in the interior.
Peace from 2002 was consolidated by two factors in particular. First, ex-combatants were given resettlement grants, and some invested these in motorcycle taxis (Menzel, 2011), with which they re-entered their old fighting grounds, but now motivated by interest in commerce rather than loot (Peters, 2007a). Second, donors (notably the EU) invested heavily in a revamped national road improvement programme. 7 The motorcycle taxis served as feeders to these new motor roads, along which trade now flowed, driven by an internal post-war commercial boom.

| How we chose our study areas
Our study design aimed to represent districts affected by different phases of the war. This requires us to justify the choice of case studies areas, one each from the three provinces into which Sierra Leone is divided.
The civil war in Sierra Leone was fought in three phases. The Libyan-backed RUF, a movement of radical youth, invaded the country across the Liberian border in 1991, aiming to make a dash for power but instead the movement found itself tied down in Kailahun District towards the end of 1993, and decided on a bold guerrilla plan of campaign which unfolded over the next 4 years. The movement's stalwarts split up and built jungle training camps in a series of isolated, hilly forest reserves across the country, located well away from main roads (Peters, 2011).
Each camp controlled areas of adjacent isolated villages in roadless terrain and conducted its operations along bush paths. The surrounding villages were raided for both food and recruits. This constituted Phase 1. This phase of the war was ended when South African mercenaries, allegedly funded by a mineral concession, bombed key RUF camps.
The second phase (from 1996 to 1998) involved the RUF coming out of the bush to take part in a peace process, the failure of which resulted in the country being divided along urban-rural lines, with an elected government in tenuous control of the towns, and the RUF (who were excluded from the democratic process) battling a miscellany of locally organized civil defence militia in the countryside.
The countryside was a mix of warlord camps and sequestered civilian farming communities. Much of the rural population resided at this time in large peri-urban displaced camps. An army revolt on May 1997 then displaced the fragile democratic regime, and the RUF was invited to support it in a power sharing agreement, abruptly terminated by an invasion of Nigerian peace keepers in February 1998 which resulted in many RUF cadres regrouping in their old stamping grounds along the Liberian border in Kailahun. From there, they were re-equipped by a second group of South African mercenaries for a march on Freetown in January 1999 only narrowly averted by the Nigerian army.
RUF forces then retreated as far as Makeni, the provincial capital of the north, controlling access to the diamond fields in Kono District.
This started the third phase of the war in which the RUF exercised a rough kind of control over the rice producing bolilands to the west and south of Makeni, from 1999 to an eventual peace settlement in 2002. 8 We chose to talk to villagers and taxi riders in Small Bo chiefdom (Kenema District) to represent the first phase of the war, when local populations were pressed into service to carry looted items from the Freetown-Kenema highway, regularly ambushed by the RUF, to a key base camp (the Zogoda) in the South Kambui forest reserve.
Today, villages close to the Kenema highway are well served by motorcycle taxis connecting with the towns of Blama and Kenema, but villages deeper in the interior remain accessible only on foot and some are only partially reinhabited, 20 years after fighting ended. Typically, these more isolated rural dwellers will thus have to walk the first stretch of a journey to Blama or Kenema and then take a motorcycle taxi to the main regional road or in some cases all the way to Kenema.
To represent the second phase of the war, we chose to talk to villagers and taxi riders in and around Gondama, a large but isolated village in the foothills of the granite escarpment that divides the country into an upland zone and a coastal plain.
This area was first attacked by the RUF in 1995, seeking food for the large rebel camp in the Kangari Hills. After this camp was dispersed (in 1997), Gondama found itself in a kind of no-man's-land of self-sequestration, fortified by a local hunter militia, during a period in which there was no single acknowledged authority in the countryside.
A further reason for selecting the Gondama case is that we have substantial amounts of pre-war baseline data for this community.
Our third case is chosen from the boliland region of which Makeni is the centre. The RUF used the area for rice supply during the protracted later attempts to find a lasting solution to the war, from 1999 to 2002. Makeni town was one of the places where motorcycle taxi riding was pioneered by ex-combatants after demobilization (Burge, 2011).
Here, our chosen settlement was Woreh Bana, a substantial but inaccessible village sited on a small bluff overlooking major boliland swamps. Motorcycle taxi riders struggle to keep this track passable through these boliland swamps during the worst of the rains, and at times, the village is cut off. 8 For a detailed overview of the various phases of the war, see Peters (2011).
Bike riders in Woreh Bana report that they first established themselves in their trade with encouragement from former members of the rebel movement in Makeni.

| METHODS
We collected data through the medium of group interviews (conducted in April and May 2017) focused on the impact of transport services in general and motorcycle taxi services specifically (given their dominance in rural areas) as tools of agrarian connectivity in three locations in Sierra Leone, with further data gathered through motorcycle taxi operator and passenger surveys. 9 As further supporting information, we ran traffic censuses on two of the tracks in the areas we chose for study-the Woreh Bana and the Fala-Gondama track-on nonmarket and market days. There was no single road in third area (Small Bo chiefdom) where we could run an equivalent census. In no case did we ask participants to identify themselves as former combatants or war victims. The war ended in 2002, and it is locally considered problematic to reopen a 'settled' issue. Where respondents in focus groups spoke about the origins of motorcycle taxi riding and about security risks associated with improved rural connectivity, these topics arose spontaneously in the flow of conversation about the benefits and disadvantages of motorcycle taxis.
The present paper is focused on qualitative evidence provided by the focus group discussions on how villagers perceive these changes. Focus groups were held with women in Woreh Bana and Maboleh (Bombali District/Northern Province), Gondama (Moyamba District/Southern Province) and Bambara (Kenema District/ Eastern Province). Each group involved between 20 and 30 women and included five or six senior women (subchiefs, leaders of women's community organizations and traditional birth attendants). Men's groups were of a similar size, and discussions took place in Woreh Bana, Maboleh and Rolunk (Bombali District) and Gondama (Moyamba District). Each group comprised senior men (chiefs and elders) and youths (including some riders).
Due to an organizational problem, a planned discussion for men in Bambara (Kenema District) had to be cancelled.
According to the generally recommended sizes for focus group work (Morgan, 1997), our groups were large. In our experience, larger groups are preferable in Sierra Leonean village conditions because this allows everybody within the relevant social categories to be invited; social exclusion in villages invites suspicion.
Experienced local research assistants (RAs) worked in pairs, according to a specific protocol for running focus groups in local conditions (Jenkins et al., 2020a). One RA facilitated the discussion, and the other handled the mechanics of participation. The protocol requires operation of a token-based system to allow the speaker, and the order in which points were raised, to be tracked without breaching the anonymity offered as part of the process of informed consent. The token system encourages participation; it is tangible proof that a person has not yet spoken. It also serves as a means to check on the extent to which specific persons or interest groups dominate the discussion.
A prompt list, based on our earlier experiences and a reading of the literature, was used to introduce discussion topics, with some topics provoking much more debate and discussion than others (see Table 2, further down). Prompts included advantages and disadvantages of motorcycle transport, beneficial and negative changes brought about through the motorcycle taxis, attitudes of riders to women passengers, preferences for the gender of the rider and possible advantages and obstacles to women working as riders. The same prompt list was used for female and male groups.
Each session lasted from 60 to 90 min, and the tracking system shows that comments came from nonelite villagers and community leaders. In all, the four women's groups yielded 161 statements. Nineteen distinct topics emerged from the discussions, and statements were coded for these topics. In the four male focus groups, a 9 For details on these survey instruments and the findings, see Jenkins et al. (2020a). total of 254 statements was transcribed (1.5 times the total for women's groups), and coding identified 31 separate topics.

| RESULTS
Our research aim was to build a picture of how men and women saw motorcycle taxis as tools of economic and social change in a post-war period of market integration, with a focus not only on benefits but also on issues of safety. Was there local support for the proposition that better rural-urban connectivity fostered peace? Given that ex-combatants had pioneered the motorcycle taxi (Denov, 2011) was there any long-term legacy of disquiet linked to the gender-based violence of the civil war period? Given the perception that accidents involving motorcycle taxis were especially common, was road safety an issue about which people were deeply concerned, and what did they want to do about it? Were young male riders reckless, and would more female riders improve road safety?
First, we present material covering three main topics-motorcycle taxis and business, motorcycle taxis and medical emergencies and whether or not women should become motorcycle taxi operators. These were topics discussed by all groups. Then, we look at what the participants said about the role of motorcycle taxis in supporting post-war reintegration and rehabilitation processes. Finally, we ensure that the other topics on which female and male focus group members wanted to speak also figure in our account, as we promised.

| Motorcycle taxis and business
A topic brought up by the women's focus groups across all three research sites was the significant contribution that motorcycle taxis have made to women's business activities: 'It has helped to promote our business (charcoal) [by providing] transportation … to sell'. Frequently, women and motorcycle taxi operators (referred to in Sierra Leone as 'bike riders' or 'okada riders' 10 ) form informal cooperative business units in which the rider acts as the courier for the goods a woman wishes to buy or sell: 'The riders have helped us greatly because [we] can just send them to town to buy all the goods we need in town'. The important role of motorcycle taxis in market integration, commented on by the women, recurs frequently in themen's discussions. In Gondama, a man remarked that: 'now there are lot of small businesses in the community', in contrast to the pre-war situation when it would have been hard to buy even a candle or box of matches.
A man (presumably one of the old school of village mercantilists) in Woreh Bana wanted fewer motorcycle taxis to ply the tracks: 'people will not [then] travel to go and get their own goods, so they can buy from me'.
The women also made it clear that motorcycle taxi riders have a rural credit role 11 : 'The riders even loan us if [we] are without money at a particular time'. Motorcycle taxi riders help alleviate the preharvest hungry season: 'During the hungry period in August, we give money to okada men to buy food and bring it for us'. Village money lenders, the pre-war agents of village mercantilism, have seen their activities reduced or gone altogether: 'now we borrow and pay within the shortest period, as okada riders go and get us what we need; [this] minimises our level of borrowing'. 10 Okada was the name of a now-defunct Nigerian airline noted for the convenience and discomfort of its flights; the term probably arrived in Sierra Leone with Nigerian peace-keeping troops during the civil war. 11 Given the short repayment terms (days, rather than weeks or months), typically no interest is charged by the motorcycle taxi operator. The 'interest' or benefit received by the operator is a type of social capital, which can be equally important.

| Emergencies
A major feature of the motorcycle taxi is the speed with which medical emergencies can now be addressed. 12 Even if the rural poor cannot afford transport on a regular basis, many can now access transport in emergency contexts (Porter, 2014). Typically, villagers will wait a day or longer before deciding if a medical case is severe enough to require referral (Richards, 2016: 84).
Families have to decide who will accompany the sick person and provide funds for the cost of the journey, treatment and accommodation. Before the advent of motorcycle taxis, if the case was severe, hammock carriers would have to be hired, at considerable expense. In one Kenema district village, currently without a resident rider, it was reported during the women's discussions that: 'we [still] mostly use hammock to carry the sick to

| Safety issues, speed and speeding
Most participants knew of cases where accidents had occurred, but few had experienced such events themselves.
For example, only four participants in the discussion in Worreh Bana had been directly involved in a motorcycle accident. Nine men had direct experience of an accident in Maboleh, which is on a dirt road graded by the biofuel company, with more opportunity to speed.
Motorcycle taxis often move slowly on rural tracks. A rolling in the mud, with attendant bruises, is more likely than the kind of higher speed spill that would require emergency medical attention. 13 Avoiding a fall depends on the skill and strength of the rider, especially if the motorcycle (as is normally the case) is overloaded. Our traffic censuses taken on market days on farm access roads leading to Woreh Bana (Bombali Sebora chiefdom) and Gondama (Kamajei chiefdom) provide some insight on the extent of (passenger) overloading.
The average loading was 1.53 passengers on the Woreh Bana track (N = 92 motorcycle taxis). Eight motorcycles, however, had three passengers, five motorcycles had four passengers and one motorcycle had five passengers.
The equivalent data for the Gondama track show less overloading, perhaps because the road is longer and more difficult for the rider to navigate (1.43 passengers per motorcycle). Only five motorcycles (out of 124) had three passengers. 12 On the significance of rural transporters in addressing women's access to emergency medical care, see Green et al. (2013) and Sacks et al. (2016). The availability of motorcycle taxi transport greatly improves chances to give birth safely. On-demand service allows quicker transfers to health facilities and addresses the concerns of many women that they might deliver while on route. 13  Skill and strength are particularly needed during the rainy season when road conditions can deteriorate quickly: 'In the rainy season we have falls due to bad road conditions'; 'The tyres mostly are smooth [and] that causes accidents'.
It was generally conceded by both men's and women's groups that four-wheel vehicles are safer, but that motorcycle taxis are quicker. The topic of speeding elicited much discussion. It was agreed that the okada is 'the fastest means of transportation' and that this is helpful for women combining family responsibilities and activities outside the home: 'if [a woman] has to see her husband in another town, the bike is the fastest mean; [motorcycle taxis] don't wait for passengers to fill up before moving like the [four-wheel] taxi does'.
The focus groups were divided on whether speeding is caused by passengers urging the rider on: ' [Passengers] cause speeding because they are rushing to attend to something very important', or by the rider rushing, perhaps for bravado: 'They [show off] to others that they can ride well' and '[they do this] to get admired'. Riders are also thought to speed excessively in order 'to target another passenger'.

| Preferences for mode of transport
A major benefit that motorcycle taxis have over more conventional modes of transport, even where the road condition allows for these modes of transport to operate, is convenience. Four-wheeled taxis and minibuses operate infrequently and in limited numbers, so it is not uncommon for a passenger in a rural location to have to wait for several hours for a four-wheeled taxi or minibus, with no knowledge of whether there would be space for them on the next service (Afukaar et al., 2019;Mustapha et al., 2017c).
The motorcycle taxi does not need to wait for a full complement of passengers before departing, and it delivers the passenger to an exact address. But it exposes the passenger to wind, rain and dust and is perceived to carry a greater risk of accident. In addition, it is about one-and-a-half to twice as expensive per kilometre as four-wheeled forms of transport (Mustapha et al., 2017c).
One woman explained that her choice would depend on the condition of the vehicle: 'my decision is based on the type of transport, whether it is roadworthy'. Others commented that in a four-wheel taxi, one is better protected: '[There is a better] chance of survival in case of accident in a [4-wheel] vehicle than on a [motorcycle], and it is very safe'. Those opting for the motorcycle also considered safety issues. Here, the skill and attitude of the rider matters: 'anyway, all riders coming here are good in riding and in their character towards us'. Lack of professional knowledge was a cause for some concern: 'there is need [to] train the bike riders on how to use the road signs'.

| Women as motorcycle taxi riders
There are a few female motorcycle taxi riders in Sierra Leone, and numbers are likely to grow. The rider's union in Kenema supports a training programme for women riders. No female motorcycle taxi riders were encountered in this study, but facilitators introduced this topic by describing a session they had attended in the training programme for female riders run by the Kenema union (Figure 1). This provoked considerable interest among the focus groups for women. Some participants immediately raised the issue of funding constraints. Business people and relatives will rent a motorcycle on hire-purchase terms to a young man but are sceptical about whether a woman can repay a loan, because there are as yet few successful female taxi riders to provide a precedent. Some participants saw an opportunity for a project to offer loans (to break the purchase bottleneck): Although the idea of women riders was eagerly debated, most women expressed a preference for male riders. Some women also thought that it might not be so easy to 'talk price' with women riders. Feminine persuasiveness was often required, especially (as is often the case) when the load was awkward or excessive. This strategy would be ineffective, it was suggested, with a hard-headed woman rider. Women 'don't joke with their money'.
Opinions were also divided about a choice of male or female rider (if female taxi operators were to be available).
Faced with a thief, or an ambush in the bush, a male rider would be preferred: 'Men can fight any [person who] would want to harm [us] or steal the okada'. On the other hand, a woman rider might be less reckless: '[she] will ride with much [more] care than men, to avoid accident'.
Men's perspectives on the idea women motorcycle taxi riders were, as with the women's focus groups, mixed. A majority of men thought it was not a good idea, but there was a sizeable minority who were in favour.
Those who thought it was not a good idea echoed views expressed in the women's groups about whether women were strong or courageous enough to cope with the challenges of rural riding: '[I say] no [to women riders], they will not be able to fight robbers'; 'No, [I am not in favour, because of their] fear of night riding'; 'Some women will miscarry if they are pregnant'. Socially conservative comments were also made: 'Women are to care for the home', 'there will be no respect for their husband' and 'they will start challenging us [the men] in income earning'.
Other men, however, were positive about the prospect: 'more customers will come their way'; 'Women are careful riders'; '[I would] choose [a] female rider-they don't easily get accidents because [they are] mindful'; 'Yes, women are to be trained, even my wife-I [will] prefer [her to ride], so she can help [support] the home'. One manperhaps reflecting on experience of police harassment-welcomed the idea on the following grounds: 'the government likes women, and if they start [riding], will support riders [more generally]'.

| Bikes and post-war reintegration
The post-war advent of the motorcycle taxi is confirmed in an explicit comment from the men: '[We] started to use okada after [the] war'. Rebel agency in launching riding from Makeni, the RUF 'capital' at the end of the war, is also confirmed: 'Our brothers from Makeni used to come here, and some called us to buy [motorcycles] for us to start employment'. The role of motorcycle taxi riding in drawing young men back into society is also highlighted: '[Motorcycle taxi riding] has made youth to be responsible for themselves and their families'.
Our survey data (Table 1) indicate that a majority of riders own their motorcycles (51%) or are in the process of buying them on hire-purchase terms (18%). Only 20% of machines were 'master bikes' (motorcycles owned by a sponsor and placed with a regular rider).
Sexual harassment (given its prevalence during the war) received little explicit comment, which is perhaps a surprise. Referring to male riders, one woman asserted that the likelihood was uncommon: 'No, we have not heard of such [harassment] in this community and [if there was such a case] it will be treated seriously by the chiefdom authority'. One reason is that rural riders are well-known figures in the communities they serve, and some are resident. Some of the focus group respondents favoured this explanation, commenting that it would not be easy to commit abuses and then hide.
Men's discussions added some important additional information on social networking: '[Bikes have] made us to have more friends, as we move from one place to the next'. The riders themselves are seen by some to have become more community-minded: 'It has changed the riders' mentality, as they even help in carrying people to funerals [for] free'. Riders are also people of substance: '[They] build houses in Makeni'. But this seems to have attracted some jealousy: '[riders] disguise themselves and make themselves superior'.
But the men in the focus groups also spent considerable time discussing a perceived disadvantage of motorcycles-rising crime in rural areas, especially theft of farm animals. '[Motorcycle taxis facilitate] stealing of our animals (goat, sheep, chickens)'. 'Passengers [who steal] instruct riders to run so they can get away from the crime scene'.

| Authority
Men spent time discussing regulation. Commercial motorcycles are taxed (through the road and driver licence fees and insurance requirements) relatively more heavily than four-wheel commercial vehicles. Riders are conscious of this disparity and query it. In particular, they express solidarity with those who evade payment and are punished by jail sentences and confiscation of their bikes. 14 The riders tend to blame the high cost of licences for riding without 14 This is also confirmed by Mustapha et al. (2017b), who found that motorcycle taxi operators generally have to negotiate over the size of the penalty, rather than receiving a fixed penalty, whenever they break the law. Harassment at police checkpoints was also a focus for comment. It was stated to be a reason why some riders, sometimes with less than roadworthy bikes, tended to base themselves in remote locations, where police visits are rare. This is apparent in the substantial number of unlicensed motorcycles logged in the traffic census data. Although perhaps still low compared with some other parts of Africa, they stand in stark contrast with motorcycle taxis operating in urban settings, which are nearly always licensed. One focus group respondent claimed that the police owned many of the commercial motorcycle taxis operating out of Makeni, implying that their riders could break the law and be protected by the owners.
Some participants in the male focus groups questioned whether female riders would cope with police interventions: 'they will not withstand police harassment'. But some riders thought that women riders would experience less difficulty with police. This would provoke the male riders and make any corrupt police officer think twice: '[It will be good to have women riders because they] will have [many male] friends to [help] fight the police and this will make the police afraid to seize [motorcycles]'. Table 2 provides a classification of topics talked about in focus groups and the distribution of those topics by gender and district. Twenty-nine topics are identified in the classification, of which five topics-access, accidents, speeding, gender and emergency response-each accounted for more than 5% of all statements made.

| Distribution of topics
Ten topics were discussed by men but not by women. These included crime, employment, loading and policing.
Three topics were discussed by women but not men-interest in riding, jealousy and money lenders. The male group in the northern sample was particularly concerned with speeding (23 of all 30 responses) for reasons unclear to the researchers. Other topics were more evenly distributed by gender and districts. We cannot claim that the three cases we have chosen are representative of the entire country or even of its more inaccessible corners. However, some well-represented topics are recurrent across both genders and all three locations. These include bad roads, business, medical emergencies, domestic issues and speed (in the sense of convenience, as distinct from speeding). There is some clear evidence of gender-specific and local factors at work in shaping some of the responses. Concern with loading and police was mentioned mainly by men. Rider carelessness was a particular concern of women. An emphasis on the strength needed to control a heavily laden okada was a focus for men but mainly in the northern sample.
As mentioned, the road in question ran through a substantial swamp that was exceptionally muddy in the rains, and indeed exceptional strength and concentration were needed to control the motorcycle. Examples of content by topic classes are included in Table 3.
These provide a summary of the main concerns of the groups. The motorcycle taxi has been transformative of rural life in all parts of rural post-war Sierra Leone, but it is not without problems, and some of these might be addressed through encouraging women to ride.

| Additional supporting evidence
As mentioned, for the Gondama case, we have additional supporting evidence in the form of pre-war baseline data.
These data allow us to claim, with some confidence, that a pre-war regime of mercantilist village grain dealers (Richards, 1990) was replaced by a post-war focus on petty trade, involving an increased number of female petty traders, using motorcycle taxi transport to access rural periodic markets.
In the 1980s, male heads of households mainly made cash income from selling upland rice (typically about 20% of their annual harvest, Richards, 1986). A survey in 2013 showed that rice sales had dropped away, and men now derived a significant part of their cash income from making and selling charcoal.
In the 1980s, most produce sales (mainly of rice and groundnuts) were handled by resident merchants, who loaned seed rice at the beginning of the rains for a return of two or three times the amount of rice at harvest at the end of the rains (when substantial amounts of rice were exported to Bo). Seed rice was still loaned in 2013, but many transactions were now between relatives, rather than between farmer clients and merchant patrons.
The place of the large traders had been taken by female petty traders, selling in Mofombo. Some commercial motorcycle taxi riders also act as small-scale produce buying agents. A bridge built by a relief agency opened the Gondama-Fala track to motorcycle taxi traffic in 2002 and a periodic market opened at Mofombo on the Bo motor road a few years later.
T A B L E 2 Distribution of topics discussed by gender and district Both traffic censuses show more female passengers on market days (Table 4). For Gondama, this is definitely a real change, because (pre-war) there was no periodic market, and to deliver goods to the Bo road required a 40-km round trip on foot. Women were observed hardly to leave their farms during the cultivation cycle but today are regular visitors to Mofombo market on Saturdays. Market integration in this region is now much stronger, than when compared with the pre-war period.

| DISCUSSION
Our focus group discussions were undertaken to provide a local perspective on post-conflict transformation in some of the worst war-affected areas of rural Sierra Leone. The discussions summarized in this paper suggest that better post-war connectivity provided by motorcycle taxis has had a highly positive impact on these vulnerable areas, linking them more firmly to towns and encouraging the spread of market-based commerce. The impact has been especially marked for women.
Women were explicit that motorcycle taxis had made it possible for them to build trading activities outside the home and that the era of exploitation by village-based merchants was at an end. 15 Specifically, motorcycle taxis facilitated participation in rural markets and allowed items to be brought quickly from towns.
Motorcycle taxi riders also increasingly provide micro-credit. They take in cash daily from passengers, but are keen to re-circulate it, due to fear of theft. Some women mentioned that they asked the riders they know well (generally, those resident in the village) to buy items for them on credit in town, which they can then sell from their veranda. This allows a female trader to combine domestic responsibilities such as childcare with petty trade.
The instability of the 1990s was characterized by high levels of gender violence. That many early commercial riders were ex-combatants was confirmed in focus group remarks. We probed this background. Did risk of harassment by riders remain a cause for concern by women?
There seemed to be little experience of this. Even if ex-combatants are still, 15 years after the end of the civil war, more prone to commit gender violence, the overwhelming majority of motorcycle taxi operators these days are too young to have actively participated in the war. Our rider surveys found that on average a motorcycle taxi operator is around 30 years of age. Comments were more concerned with threats of ambush and theft. In fact, some women preferred a male rider capable of fighting to protect the motorcycle, his passenger and her loads. 16 Risk of sexual violence was not seen as a cause for concern. Village-based riders are well known to the local authorities, and women seemed confident that the chiefs would deal effectively with any attempted harassment from a rider. This provides some support for the idea that chiefs are perceived locally as being effective in maintaining public order.
Opinions were more divided on whether women should aspire to be riders. There was a degree of support for the idea and not only from women who saw it as a potential business opportunity. Some were influenced by post-war discourse about women's empowerment, and others saw it as fitting, since it was their right.
Others were more concerned about practical constraints. An overladen bike on a muddy rural track is notably hard to control.
A good number of men, and not a few women, thought that only men had the requisite strength to keep the bike from falling. On the other hand, some focus group participants, both men and women, suggested that women might make better riders; they would be less likely to lose control, because they would be more careful in the first place. 15 Jenkins et al. (2020b) offer evidence that traders still exploit farmers where bikes are few. Middlemen trucking produce to Monrovia offer to pay for labourers to head-load items along local footpaths to the road head at Kpaytuo in Nimba County but extract a premium from farmers to do so. One older farmer stated he could not avoid this payment due to his lack of physical strength. 16 This is consistent with a report by Fithen and Richards (2005) that some women traders in Bo, re-establishing their links with the countryside immediately after the war in 2002, sought out riders they knew to be ex-combatants, because they would be better able to assess the risks of falling into ambushes on rural roads. Much has written about the propensity of motorcycle taxis to be involved in traffic accidents, particularly in urban environments (Bishop et al., 2018). Perhaps surprisingly, accidents were not seen as a major concern by members of our exclusively rural focus groups. This may be because they are less common than might be supposed, on rough tracks where necessarily speeds are low. Mustapha et al. (2017aMustapha et al. ( , 2017bMustapha et al. ( , 2017c highlight these so-called rural transport safety paradoxes: bad roads can cause more accidents, but they force operators to ride more slowly, resulting in less serious accidents; overloading of motorcycle taxis can cause operators more easily to lose balance, but this risk also forces riders to drive more slowly and so on.
Most passengers reporting their views in our focus groups had experienced a spill, but life-threatening incidents were rare. It should be cautioned, however, that extensively sampled accident data on rural roads-and in particular rural tracks-are almost non-existent. Our own provisional view is that serious accidents are indeed rarer than in town and that town-based authorities may tend to overestimate the risks associated with rural riding, given the number and severity of urban accidents involving commercial bike taxis.
It is worth noting the high degree of local admiration expressed by our focus group participants, both male and female, for the skill of the professional rider, without which (focus group members were clear) accidents would be much more common.
Comments on how women would cope with the recurrent adversity of bike riding (punctures, breakdowns and bad weather) clearly reflect a more general local discourse about gender distinctions, shaped by notions of genderbased task complementarity inculcated by the still-influential male and female secret societies found throughout rural Sierra Leone.
Obstacles to women becoming motorcycle taxi riders were a major point of attention and deserve further comment. A major issue is the lack of credit. New motorcycles, our operator surveys showed, are generally purchased by a sponsor, with the operator becoming the owner via some kind of hire-purchase contract. About a third of motorcycle taxi owners stated that they had purchased their motorcycle-often a second-hand one-through the proceeds of farming and charcoal burning.
It is hard to convince an investor-including successful female business persons, we found-that women ride well enough to generate the weekly (or sometimes daily) turnover the sponsor requires to recoup the loan when there are so few, or no, female riders to provide successful precedents.
Information that one of the motorcycle taxi unions (in Kenema) was training a group of women riders generated a lot of interest in the female focus groups. It is especially interesting to note that the 'bottleneck' issue of women's access to bike sponsorship and access to credit was so quickly raised by discussants with no prior knowledge of schemes to promote women's participation. 17 This suggests that access to credit is seen more generally as a major barrier to greater economic participation by women.
Some support for extending ridership to women also came from male riders. This again seemed to be linked to the idea of credit as a barrier to employment. The rider needs a loan to acquire a motorcycle, but the loan is repaid more quickly the more the motorcycle works, so a rider is keen to share it with a responsible partner. 18 Women are seen as reliable partners-both as careful riders and for their perceived greater responsibility in monetary matters.
Some riders would especially welcome it if their partner in the bike-riding business was also their wife.
In general, then, these focus group data support a claim that post-conflict market integration has been marked by a much greater participation by women in petty commodity production and trade and that this transformation is driven by better rural connectivity, resulting in particular from the increased penetration of off-road rural areas by motorcycle taxis. Earlier recognized trouble between villagers and alienated rural youth-once deemed a cause of civil war-is now significantly reduced. 17 For those contemplating a project intervention, note that Porter (2011) found that many women in Ghana, who expressed an interest in riding bicycles, handed over the machine to their husbands once provided with it as an intervention in an action research study. 18 See Table 1 for frequency of ownership/rental arrangements.

| CONCLUSION
It has been argued that greater rural-urban connectivity has been a powerful factor in moving war-affected parts of rural West Africa on from the chaos and violence of the 1990s. We wanted to assess how much that understanding is shared by rural communities representative of those most directly exposed to the earlier violence. Specifically, how did they view the main means of improved connectivity-the rural motorcycle taxi, given that it was a transport innovation pioneered by ex-combatants? Does it deliver the benefits often claimed-of better access to markets and health? Or (on the other hand) do risks of gender-based violence or physical danger of accidents mark out motorcycle taxis as a problematic sphere of service provision? Our method was to encourage informants to talk about these issues in their own terms. The paper has sought to offer as direct as possible a rendering of what was talked about, and the lines the conversations followed.
The insecure rural world of the 1990s has disappeared. New forms of market integration based on cell phones, motorcycle taxis and petty trade in periodic markets are almost without exception considered welcome developments. Women, especially, cannot envisage a return to the drudgery of life tied to the house or the indignity and strain of head-loading their items for hour after hour along dusty or sodden pathways. The motorcycle taxi helps women combine petty trade with household duties, and the availability of motorcycles also assures them of being able to obtain rapid medical help for sick children. Although rural women, including many from the communities we studied, were highly vulnerable to rebel violence, they seem to have surprisingly little concern that such violence might return. They are aware that many of the motorcycle taxi riders of today come from the same class stratum-marginalized rural youth-from which the rebel movement picked and trained its cadres, but a key difference now is that the riders are firmly based in the community. Motorcycle taxi riding (and associated auxiliary jobs, such as selling fuel, repairing bikes or cleaning motorcycles) is offering young people a new livelihood opportunity, where before, living in a village equalled being a farmer. Users know the motorcycle taxi riders' movements and have a sense of where they can get redress if anything untoward should occur. When asked about attitudes to ex-combatants, and specifically those who became motorcycle taxi riders after the war, villagers who suffered under the rebel regime are now more likely to explain that the fighters have 'settled down'.
'They marry, they have children, they build houses, and worry about how to pay school fees, like the rest of us', one villager explained. In Sierra Leone, the international community's preferred prescription for the demobilization and reintegration of ex-combatants was through vocational skills training, and more generally, for the development of the countryside, to stabilize violence-prone young men through the creation of jobs via direct foreign investment in mining and agricultural estates (Peters, 2007b;Peters & Richards, 2011). Evidence presented here suggests that a lower profile solution-agrarian transformation based on intensified short-term rural-urban interaction-is an effective means of promoting peace. As is so often the case with effective development, the most impactful solutions are often local.