1Small and medium-sized towns in Africa play a long-established role of relay and redistribution centres in trade relations between town and country. Their location on trade routes, the provision of rural areas with commercial facilities and equipment, and their position as intermediaries in the networks have all allowed the development of many secondary urban centres. In border-located towns, these factors are coupled with transnational relational logics, such as interface areas, offload points and customs posts.
2However, recent socio-economic changes resulting from globalisation and the liberalisation of markets have tended to modify the acquired relational logics. New actors, of both a private and international nature, have positioned themselves in these trade relations and influence the part typically played by actors of these intermediate urban levels. Within the context of multi-scale trading systems, market activities, arising from spatial complementarities and relying on an open-market system, have participated in the redefinition of the role of secondary towns. Although not the driving force of the globalisation dynamics, they participate fully in the process, being more than mere external logics relays. Through their integration into broader trade systems, secondary towns contribute to the structuring of their hinterland in new and novel ways.
3The present paper builds on results of research studies conducted within the CORUS 6165 programme: “Mountains and medium-sized towns in East Africa. Flows of people, resources, and management of environments” (2008-2012), and the “Rurban Africa” project funded by the 7th Framework Programme of the European Commission (2012-2016).The Mount Elgon region on the border between Kenya and Uganda, south-west Tanzania, the Uporoto Mountains and Mount Rungwe region, close to the borders between Malawi and Zambia, are particularly interesting for observing how the mutations of market relations, influenced by local developments and the logics of globalisation, have transformed the classic role of these urban levels and redefined the relational scales (Figure 1). They are mountainous areas, framed by medium-sized towns (Mbale in Uganda with 80,000 inhabitants, Mbeya in Tanzania with 400,000 inhabitants) and small towns (Kapchorwa on the foothills of Mount Elgon, Tukuyu in Mount Rungwe), particularly structured by merchant systems related to recent agricultural developments (rapid growth of the food-producing trade) and border integrations. The article will highlight the role of interface and the integration functions of secondary towns, and then the profusion of new market players before showing the territorial dynamics related to the growth of the markets in both the urban and rural sectors.
Figure 1. Study area
Source: authors
4The importance of small and medium-sized towns in urban-rural relationships has been largely analysed, especially in African settings. A border location strengthens regional integration for towns into a peripheral position within their national space.
5Although the role of secondary towns is not limited to the sole function of market activities (they are also administrative areas, service centres and places of social and economic interactions), this role is particularly important in Africa. Researchers have shown the place of such towns in the outward orientation of rural economies (Rondinelli 1988, Baker 1990, Simon 1992, Baker and Pedersen 1992, Giraut 1994, Owuor 2006). This function has undergone considerable developments as the result of agricultural changes. Colonial cash crops such as coffee, cocoa and tea are largely competing with, or have been replaced by, new productions. Urbanisation has become a driver of agricultural growth, and scales of analysis must incorporate the town and the country as interacting, even interdependent, elements (Coquery-Vidrovitch et al. 1996, Chaléard and Dubresson 1999, Pélissier 2000, Tacoli 2002, Charlery et al. 2009). Market gardening and subsistence farming were developed by farmers and actors within the marketing channels to satisfy the urban consumer market. In the mountains and highlands of Africa, the number of these new farming productions, which enjoy the physical benefits of high altitudes, has soared (potatoes, cabbages, tomatoes, etc.). The western highlands of Cameroon, the region of Moshi at the foot of Mount Kilimanjaro, and the foothills of Mount Kenya act now as supply centres for the large cities (Hatcheu 2003, Brient, 2007).
6The observed territories in East Africa form a wide fertile crescent, the former vector of mobility for populations attracted by exceptional ecological conditions (on Mount Elgon, the Manafwa and Bududa rural districts have between 800 and 1,000 inhabitants per km2). They were sought-after lands for commercial agriculture, itself the mainspring of colonial and national economies. Today, the towns have become part of the mountain territories of East Africa, which are encircled by markets, smaller communities or medium-sized towns with assertive commercial and logistics functions, a network structured a priori by the interplay of geographic scales, with the proximity of the small centres giving access to the cities in the foothills. The relations between highlands and lowlands and between town and country structure the territorialities of these “bastions of rurality” (Charlery et al. 2009) and the secondary towns fully act as intermediaries. In the Mbeya region in Tanzania, the cultivation of bananas has largely replaced the coffee plantations of Mount Rungwe as a source of income, while market gardening has replaced the pyrethrum (merely residual now) in the Uporotos. In Uganda, on the slopes of Mount Elgon, there is a differential and more or less marked neglect of the coffee tree, along with that of the banana tree, in favour of vegetable crops.
7The function of markets is essential, owing to their role of redistribution of food products. In intermediate towns, markets are bustling, with several hundreds of producers selling their products or middlemen sourcing in mountain areas and coming to the market to supply merchants and resellers. The function of redistribution of agricultural products is easily identifiable in these towns, through the omnipresence of shops and market stalls selling vegetables, fruit and dairy products (Figure 2). At Mbale, resellers of vegetables are supplied three times a week from the mountain areas (with potatoes, sweet potatoes, carrots, cabbage, onions, and bananas). Middlemen play an important part in this trade: they travel to rural areas to buy products and bring them to urban retailers. It is thus possible to find goods from the entire area of Mount Elgon, from Kapchorwa in the north to Bududa in the south. A large share of the mountain production is also sold in the other urban centres (Jinja, Kampala) and dispersed across the country via towns that serve as places of redistribution and distribution outside rural areas.
Figure 2. Mbale Main Street on market day
Source: authors
8Medium-sized towns are also gateways to rural areas. They are important in the distribution of products, often manufactured or ‘imported’ from other regions or from abroad. Many products come from outside the mountain system: in Mbale, from Kampala, from nearby Kenya or from abroad via Kampala. Mbale is thus an outpost from where products are redistributed (Figure 3).
Figure 3. Mbale, an example of urban-rural interface
Source: authors
9Owing to their market activities, secondary towns play an important role in the connection of their hinterland with national and international markets. The intermediation role of these secondary towns is both external (gateway to outer markets, logic of networks), and internal (towards proximity areas) with essential utility functions (service centres, influence on the hinterland, a logic of proximity). This long-standing function of intermediate towns as hubs of redistribution contributes to anchor rural (in this case, mountain) economies in national and international economies. Such secondary towns are the best example of intermediate towns that ensure the transition between large cities or the capital and the rest of the towns and rural areas, and vice versa, from the basic level to the top. They thus play an intermediation role in a hierarchical structure, between an urban centre and the zone for which they constitute the hub of economic services. The intermediate town is one that is placed between extremities (between the small and the large; between the close and the distant) and that develops intermediation functions between a wide variety of spaces and scales (local, regional, national, global).
10Both towns studied here illustrate well the intermediary function of these secondary towns (major regional towns) in structured urban networks, in a peripheral location, or even away from their country’s economic capitals (Mbeya is 820 km from Dar es Salaam and Mbale 250 km from Kampala). Their role is reinforced by the polarisation they exert on the regional urban network of small-sized towns, market towns and the surrounding rural areas. Thanks to these towns, the mountain areas are integrated into international markets and through these market activities the towns enjoy a significant demographic and economic dynamism (Figure 4).
11Agricultural inputs, such as seeds or plant protection products, are widely used and increasingly so, in line with changes in farming practices. In Mbale, merchants who sell such food and chemical products for agriculture and livestock source from Kampala or nearby Kenya, where industrial development is on a larger scale. The zone of influence of these specialty shops covers the entire district of Mbale and the surrounding areas (within a 40 to 50 km radius). The border area location of Mbeya, on one of the main East African railways (Tazara) and highways (Tanzam), allows relationships with all of the sub-region. The agricultural production of the Uporotos and of Mount Rungwe are routed preferentially to the national market (Dar-es-Salaam or to other towns in Tanzania), but also exported to neighbouring landlocked Zambia and Malawi (Racaud 2013a). Mbeya is incidentally the dry port of Malawi.
12Further remote destinations such as Botswana or the Democratic Republic of the Congo are also significant while occasionally, particularly in times of drought, agricultural products such as corn can be exported to Kenya or Uganda (as was the case in 2011). Conversely, Mbeya is the point of entry of products and productions from outside the region, mainly via Dar es Salaam (manufactured products imported from China, plant protection products), but also agricultural products from other regions (oranges from Tanga, onions from Iringa), fresh or dried fish from the Great Lakes, and products imported from neighbouring countries (sugar from Zambia or Malawi, loincloths from the Congo).
Figure 4. Mbeya, trade network and regional integration
Source: authors
13The position of Mbale and Mbeya in these commercial networks, including products for/from agriculture, demonstrates their centrality within regional (mountain) systems, but also within cross-border sub-regional areas. Mbale is an urban hub of greater importance than Tororo, which has a population of only 45,000 inhabitants but is better located on the main route between Kenya and Uganda’s main towns, Jinja and Kampala. Malaba, on the border with Kenya, is only a border post. The vitality of Mbale can be explained by its combining of the functions of a town polarising a dynamic rural hinterland and its proximity to the border (which increases its influence, including by the "incoming" flows visible across the range of manufactured goods imported from Kenya or via Kenya). Mbeya displays a similar configuration. Strictly speaking, this is not a border town (Zambia and Malawi are over 100 km away), but the town polarises the trade flows of the entire sub-region, much more so than border posts (Tunduma, on the border with Zambia, is a small-sized town of 35,000 people; Ipinda, on the border with Malawi, has 16,000 inhabitants).
14Supplying the urban consumer market is very competitive (other agricultural areas, sometimes closer or better connected, also supply the large East African cities). The border location allows diversified trade opportunities. Through their commercial activities, these peripheral towns are thus integrated into a wide regional network. This is particularly true in the case of Mbeya, nicknamed the “Dubai of Malawi” by the people of Karonga (Northern Malawi) due to the diversity and quality of the products found or coming from there.
15With intensified market relations and the implementation of liberal policies in countries such as Uganda and Tanzania, an increasing number of actors are involved in urban-rural relations, especially in secondary towns. These actors prove to be very innovative and, as mentioned by Walther (2014b), it is important to analyse social networks by exploring how they have progressively adapted to social and spatial changes in economic activities.
16Small and medium-sized towns play an important role in the supplying of public and private services, and the infrastructure required for the development of very small enterprises related to the economy and non-farm employment. Many rural (agricultural and non-agricultural) enterprises, involved in food processing, construction, transport, retail or wholesale, financial services and services to individuals are located in these towns, serving urban and surrounding rural populations (Rondinelli 1988). Services to farmers and rural populations are also a good indicator of urban influence. All these entrepreneurs, either local or from outside the region, have their strategies, combining economic, social and spatial aspects, and using cross-border and rural-urban complementarities – which is typical of border markets, as analysed in West Africa elsewhere (Walther 2012).
17With the turn to the open market, the old cooperatives, like the Tanzanian Farmers Association (whose regional headquarters are in Mbeya), are being challenged by these new (and often small) structures, which offer better prices to members and better payment opportunities. Associative structures such as Savings and Credit Cooperative Societies (SACCO) have developed under the impetus of non-governmental organisations, but private commercial structures have followed the same trend. Secondary towns benefit from agencies of national structures, which have often been in operation for some years and offer agricultural producers and also non-agricultural entrepreneurs of the region easy access to credit (microcredit). More broadly, the market of products for agriculture (inputs, animal feed, veterinary products) is undergoing growing privatisation, as exemplified by the proliferation of shops in urban centres (Figure 5).
Figure 5. Agro input shop in Mbale, Uganda
Source: authors
18Within the market system itself the number of actors and middlemen has multiplied, along with more complex business practices and strategies at work between the producer and the end consumer. Mountain productions can be collected on-site, at temporary or seasonal rural markets, transported over to rural markets, to urban markets of secondary towns or directly to major urban markets. The actors involved may be rural, urban actors from secondary towns, urban actors from large external cities (wholesalers and traders contracting directly with producers) or foreign actors. A complex chain of actors is thus emerging, involving producers, collectors, transporters, wholesalers and retailers, linking local markets, distant markets, mountains and towns (Figure 6).
Figure 6. Chains of actors of agricultural trade from Mbeya region to the sub-region
Source: authors
19Secondary towns are also nodes of the imported goods network, such as Chinese products that are found in all rural markets as well as in the urban commercial landscape. Cheap items, i.e. low-value and often lower-quality products, are among the main goods. Indeed, their low cost makes them of easy access, whether for consumers with low purchasing power or the diverse range of traders, from street vendors to importers shuttling between the regional metropolises, for example Mbeya and Dubai. The development of this trade is part of a very strong growth of bilateral trade between China and Tanzania, who are traditional partners. This trade serves as a refuge for many town residents, who can start a profitable low-cost activity, even if profits are low. Some traders manage to jump up the traders’ scale and street vendors end up owning a boutique. Shopkeepers generally get their supplies in Kariakoo, the major shopping district of Dar es Salaam. Mobilities between the secondary town and the national capital are developing. Thus, for lack of significant capital and because of the strong renewal of products (especially for head-to-toe clothing), new supplies are frequently available, often several times a month. This trade creates durable links between the Tanzanian periphery and centre.
20However, secondary towns do not only harbour small-scale vendors or shopkeepers. Indeed, in Mbeya, traders directly connect the local and regional markets with big malls overseas. Importers own several stores that they supply from their sources in Dubai or China. Some traders even have shops abroad as well as in the Zambian capital. Dozens of Chinese goods importers operate in the capital of South-West Tanzania. They do not work together because the challenge is to get exclusive access to new products. Other actors take advantage of the development of this sector even if they do not have the capital needed to deal with imports on their own. They are agents who, like their counterparts involved in agricultural trade, do not collect food orders but purchase abroad for several local merchants who do not have the sufficient capital or contacts to import themselves. In secondary towns, merchants of imported goods contribute at their own level to the flow of goods and ideas. Connections are increasingly more numerous between more places of a greater variety, for example between Mbeya and Dar-es-Salaam or Mbeya and Dubai, or even Mbeya and Guangzhou. Mobile actors, with their greater experience of other settings, are vectors of new ideas. In addition, the numerous products they bring back, rapper caps for example, contribute to the evolution of local identities, both in towns and in the countryside.
21The study of such actors (cooperatives, associations, traders, middlemen) shows the importance of the urban network: regional branches are located in medium-sized towns (Mbeya, Mbale), local offices in small-sized towns (Kapchorwa, Tukuyu) and rural agencies exist in the market towns (Bududa, Wanale) as well as representatives in more remote areas. These developments raise the question of the possible cooperation between such actors and of the resulting impact on territorial development. Initially, they strengthen the weight of secondary towns by beefing up their offer of services, and broaden the range of services and prices available to urban and rural consumers. However, the lasting effects are still to be assessed, such as the challenge of the effectiveness of the framework offered by small companies that are sometimes weakly-structured territorially and whose economic viability is not guaranteed in the event of changing economic context, and the risk of actors in secondary towns being ‘short-circuited’ by metropolitan private contractors (carriers for example) directly connecting rural areas with larger national or foreign cities. Direct trade relationships developed between rural areas and main towns, or even global markets, may weaken the intermediary role of secondary towns. Studies conducted in Kenya on the place of secondary town small entrepreneurs in town-country relations (Mainet and Kihonge 2015, Kihonge 2014) tend to show that competition for secondary town actors is most strongly felt in the outbound direction (export from the mountains to distant urban markets) than in the inbound direction (import from the outside to the mountain system).
22Rural areas polarised by dynamic small and medium-sized towns can be integrated into economic markets and extended relational spaces but the process is accompanied by strong socio-spatial contrasts. Market activities are embodied by markets whose geography makes it possible to observe the impacts of spatial differentiation.
23The market is the organised set of tangible and intangible exchanges. It consists of a system of exchange of goods organised as tenders and requests meet, but this function of economic exchange (collection and redistribution centre) is also a social function (exchange of news), a political function (site of power of local traditional authority) and economic openness (distribution to the outside). In Africa, the market occupies a central place in the food trade (Chaléard 1996). The African market for agricultural products is characterised by structural flaws. Indeed, for most producers (who are mostly peasants) access to the market remains problematic and unbalanced as far as infrastructure (transport and exchange), information and rules of exchange are concerned. The weakness of the standardisation of goods (measurement units, product quality control) is often to the detriment of producers, and likewise the lack of access to credit and the informal nature of transactions. Such factors favour the volatility of prices, volumes and quality. Markets are diverse by their nature, their spatial and temporal significance (local/international, daily/periodic), by their modes of regulation, and even by their legality (Minvielle 1999). It is in markets “that the largest amount of trade takes place, that actors and flow of goods can be more easily identified, prices studied, circuits marked, and areas of production localised. Beyond their diversity, markets are privileged places for peasants-traders contact in the countryside, and to supply city-dwellers” (Chaléard 1996: 493).
24A number of parameters for integration in and by the market can be analysed by observing the markets in their spatial anchor. The market (place) can provide insights into economic and territorial integration. Mbeya and the Uporotos are very interesting for observing the changing configurations of places and new trade routes. The trade dynamics of the mountain rely on an exchange network some nodes of which do not always correspond to urban centres. Thus, Kiwira, a village of about 4,000 people, plays a far more important role in agricultural trade than the small town of Tukuyu, which has 50,000 inhabitants and is also a district capital (Racaud 2013b). The collection and distribution function of the markets creates a network of places of exchange whose hierarchy is organised on the basis of commercial criteria such as the availability of products, accessibility, price and quality.
25An intensification of the periodic markets network (gulios) can be witnessed in the Uporotos (it can also be observed on Mt. Elgon, in the Mbale region). A market is created at the initiative of the local community when the volume of transactions and the number of actors are sufficiently great for the district to deem it appropriate to collect taxes. The thus formalised market is then provided with a committee that collects taxes for the district and ensures the basic tasks of management of space, such as infrastructure maintenance and cleaning. The map of market creation shows three subsets (Figure 7). The oldest markets are concentrated in the rural Mbeya district and the Umalila area (slightly landlocked agricultural table-lands) while in the Rungwe most markets date back to the 1980s, at the beginning of the massive conversion to market gardening. Markets along the Dar es Salaam–Zambia route constitute one last set. The recent settlements emphasize the structuring effect of the road (here the main Tanzam highway and Mbeya–Malawi road). The six markets created since 1998 are located within 10 km of the tarred route, which is an indicator of the amplification of interactions between the mountain and the town and of the integration of areas of production into the domestic and international markets.
Figure 7. Periods of creation of markets in the Uporoto Mountains, Tanzania
Source: authors
26There are 38 officially registered markets in the Uporotos area: 12 urban and daily markets, and 26 rural and periodic markets. These markets can be differentiated by typological analysis on the basis of hierarchical logics (size of markets according to the volumes traded), more or less strong specialisation parameters and criteria of differentiated integration into local, regional, national and international markets. This results in a network combining traditional periodic markets, collection markets and urban hubs.
27Periodic rural markets of the Uporotos combine dry goods from productive non-mountain areas (peanuts from Dodoma, rice from Kyela), fresh food items from zones some 10 kilometres away or from the nearby hinterland, and (new or second-hand) imported manufactured products. There is a strong specialisation of some of these markets depending on the productive characteristics of their immediate environment. Thus, the market of Ibililo offers different varieties of bananas (Matoki, Marindi, Kambani, Uganda). The largest of these markets bring together between 200 and 1,000 trade actors (professional traders and peasants selling food), with significant attendance gaps, ranging from several hundred individuals up to 15,000 for Kiwira. These values fluctuate with the seasons. There is a strong complementarity between the markets. Thus, the Ibililo market takes place on the same days as the large Kiwira market: external transporters and collectors are sure to find sufficient quantities of foodstuffs whose collection is facilitated in space and time as they rotate between markets to fill the trucks.
28Beside the official markets (where taxes are collected), there are informal exchange sites in the vicinity of road connections, with some of them being genuine institutions (Ntokela for potatoes, Simambwe for cabbages). Such roadside markets do not have a logistical framework or a local regulatory framework. They are present on the major road that links Mbeya to Malawi.
Figure 8. Market differentiation in the Uporoto Mountains
Source: authors
29Next to the specialised gulios, two market places stand out clearly owing to their collection and redistribution functions and their retail and wholesale practices. These are the Uyole and Mbalizi markets, located at the interface of the rural and urban markets, at major crossroads, giving them obvious comparative advantages (Figure 8). They have a multi-scale production supply, composed of mountain hinterlands and respective production basins which replenish their collection and redistribution centre (the Uporotos and Mount Rungwe for Uyole, the Umalila plateau for Mbalizi). These locations give Uyole the privilege of a more favourable position, its local production supply being wider, its productions more diversified and its volumes larger. In addition, the presence of the road link to Malawi affords easier access to the market and smoother product flow, whereas south of Mbalizi only dirt tracks are available. The location of Uyole, at the intersection of the two international routes, is a crucial asset.
30As noted by Walther (2014b: 10), “the presence of a border market and transport infrastructure is crucial for trade communities and producers because it allows the combination of trading and production activities to conjointly benefit the regional economy”. In turn, the impact of how the markets are structured can also be seen in towns, through the spatial organization and evolution of urban zones. In the Uporotos and Mbeya, market activities structure the flows and differentiate locations. The case of Mbeya is singular (considerable specialisation and differentiation), unlike that of Mbale, where the central market is particularly important and lively (the town is smaller and there are fewer urban markets than in Mbeya). The location (major all-weather roads), the proximity of a border and the accessibility of supply-rich hinterlands and catchment areas are important factors of spatial and functional differentiation.
31The urban markets of Mbeya have distinct characteristics. The dozen registered markets are located in the older parts of the town and on the Tanzam road axis, where population densities are highest. Some markets are specialised, like those of Soko Matola whose reputation is built on the availability of fresh fish and of Uhindini renowned for fresh beef meat (the market burned down in 2011, like many urban markets in Africa where “accidental” fire is a shameful way of urban management), and Soweto market known for its substantial stock of low-priced clothing. Mwanjelwa-Sido is the largest urban market of Mbeya (with close to 1,600 registered shops). Urban markets correspond to their catchment area and prices are adapted to customers, with the lowest prices being charged outside the old town centre. Uyole offers the lowest prices for foodstuffs. Urban markets are organised in a hierarchical operation; some of them supply the others. This makes sense for large peripheral central squares directly supplied by the production basins: merchants in downtown markets refuel in such hubs as Uyole and Mbalizi. Nevertheless, although located in the town, Mwanjelwa supplies goods to other urban markets. This is a particular market as many on-site traders get foodstuffs in periodic rural markets (as in Kiwira, Ibililo). In contrast, a Soko Matola trader does not frequently source in the countryside. Urban market places are secondary markets since, most of the time, products do not come directly from the farming area and such sites are dedicated to retail trade.
Figure 9. Hierarchy of urban markets in Mbeya
Source: authors
32Secondary towns and Mbeya in particular are not only agricultural trade network nodes they are also shopping malls which supply their periphery with imported manufactured goods. In this field, Mbale plays a much smaller part than Mbeya. This can be explained by several factors. Mbale has less than 100,000 inhabitants, while the population of Mbeya is about 400,000. In addition, Mbale is located 225 km from Kampala, which itself is just 215 km from the Kenyan border. Mbale remains in the area of economic polarisation of Kampala and cannot be a shopping centre selling goods imported from Mombasa, the largest port of East Africa. Mbeya is in a totally different situation. More than 800 km from the port of Dar es Salaam, the town, with its logistical facilities (including the dry port), is a sub-regional commercial and logistics centre for its landlocked neighbours. This peripheral relation to Tanzania, but central to the sub-region, is a structuring factor of business development which translates into a specific urbanisation.
33While the agriculture of the mountain hinterland is still the driving force of Mbeya's development, the imported goods trade is rapidly developing and affecting the urbanisation of the Tanzanian regional metropolis. Commercial activities are concentrated in the Mwanjelwa district along the highway that runs through Mbeya, where the commercial function takes precedence over the residential function. This dynamic is the result of the initiative of private operators, although local authorities have tried to take part in the trade. In 2009, a project was initiated by the town (public-private partnership) to build a shopping mall of over 20,000 square metres. However, work was put on hold in 2013 and in 2015 the building was still not completed. Hence, market activities, the main function of the Mwanjelwa district, are being developed by private operators (Figure 10).
Figure 10. Mwanjelwa Market Street and the building site of the shopping mall in the background (Mbeya)
Source: authors
34Opposite the shopping mall, a souk-like interleaving of shops is steadily replacing an area of old houses. In the last 10 years, the shops have penetrated more than 100 m into the living area, and have spread out along the street over a distance of 200 m. This expansion includes several blocks separated by narrow dark alleys (Figure 11).
Figure 11. Alley in the new Mwanjelwa market
Source: authors
35This market sector supplies all the goods consumed by urban and rural populations such as clothing, shoes, blankets, perfumes, baby wear, fashion accessories, loincloths, suitcases, and household accessories, mostly from China. The place is truly central to the merchant network of imported goods, radiating into the Mbeya region but also beyond the national borders. It performs retail functions for the local population, but also wholesale functions for traders from the hinterland and regional towns and from Malawi, Zambia and the East of the DRC. This market district has become the centre of the town at the expense of the old town centre with its administrative functions. As Mbeya’s commercial functions have strengthened, the centre has shifted to the Dar es Salaam–Zambia road. In addition, other commercial establishments including banks and hotels have settled along this thoroughfare in the Mwanjelwa district acting as a commercial showcase created by private actors. As a result of the urban restructuring in this major sector of the town, Malawian traders have aptly named Mbeya the “Dubai of Malawi”.
36Local changes (a shift from commercial agriculture to the food-producing trade) in combination with global changes (a turn to the open market and globalisation) have led to an amplification of market activities and to more complex market and network systems. Urban-rural relations are more important than ever. Depending on the level of analysis, rural and urban territories are differentiated by complementarities, competing logics and differential benefits, some being characterised by comparative advantages, others remaining at the margin of the dynamics of external openness. The road and all it induces, i.e. access to the market, has emerged as a structuring factor of the “terroirs” and markets.
37Secondary towns, each at their own level, perform a structuring role in their locality whose effects extend as far as their production supply areas. Indeed, “border markets are more than places of flows; they also play a crucial regional role for the organization of agricultural activities” (Walther 2014b: 9). Periodic markets are showcases of the hinterland and act as interfaces between the nearby and the distant. The densification of the market network is an indicator of the changed mountain agricultural model and of the integration of the mountain ranges into broader trading systems. The commercial functions performed by private actors are the driving force of the urbanisation of these towns. The case studies of Mbeya and Mbale show the importance of developing an analysis of the role of secondary towns as intermediaries between rural and urban networks, in line with Bell and Jayne’s recommendations on their role in urban studies (Bell and Jayne 2009).
38The dependence of these mountain systems on external economic logics bears witness to the frailty of the integration model based on the development of commercial food crops. Integration into the regional market is a line of development that applies particularly to the Uporotos and Mount Elgon. This is particularly true for the Mbeya region by virtue of its strategic location at the gateway to landlocked countries (the differential is reversed in the case of Uganda, which depends largely on supplies from Kenya). Regional structuring of the potato and banana sectors, crops much in demand by Malawi, Zambia, the Democratic Republic of the Congo and even Botswana, could be an instrument of regional integration and a means of strengthening the centrality of a town such as Mbeya at the regional level and of reducing its remoteness at the national level. The international border acts as an amplifier of the growth of the commercial functions of these secondary border towns, which are nodes of a wider merchant trading system. However, this perspective, if made their own by private actors, including, those from outside, may merely reinforce the dependence of these border regions on external centres of impulse and decision.