The proponents of small town development see an important role for small towns in improving the conditions of the rural poor in developing countries. Firstly, an argument for promoting small towns is that they provide markets for urban consumer goods from higher-level towns and act as trading centres for agricultural goods from rural areas (Rondinelli, 1984; Gaile, 1992). It is believed that urban population growth and agglomeration in small towns creates increased demand for agricultural products from nearby rural areas. Secondly, small towns are seen as possible locations for providing non-farm employment for the growing rural populace in hinterland villages (Gaile, 1992; Leinbach, 1992). With expansion of employment in small towns, they can act as alternate destinations for potential migrants to large cities (Mathur, 1982). The rapid growth of large cities often results in greater regional inequalities, problems of governance and environmental sustainability, all of which call for greater attention to smaller towns from equity considerations as well (ESCAP, 1991; Jones, 1991). Thirdly, proponents of small towns consider them appropriate locations for concentrating public services, such as agricultural development services, health services and educational facilities, for reaching a larger rural populace (Taylor, 1981). It is assumed that the provision of agricultural credit and inputs at the small town level can introduce farming innovations at the lowest level, rather than allowing them to filter inefficiently down the hierarchy of urban centres and transportation routes causing time- and distance-delay effects.

_____________

* The author, who is a Post-doctoral Fellow with the Division of Demography and Sociology, Research School of Social Sciences, Australian National University, would like to acknowledge with gratitude the comments and support of Gavin Jones, Mike Douglass and Brian Murton. Field work for the study was conducted with financial assistance from the East-West Center Population Program, Hawaii, and the International Center for Integrated Mountain Development (ICIMOD).

These propositions, nevertheless, have been tested in areas with particular geographic and socio-economic characteristics. Most past studies on small towns from other developing countries have focused on areas where agriculture has moved beyond the subsistence level (Hardoy and Satterthwaite, 1986; Titus, 1991). In contrast, the people in the hills of Nepal are engaged predominantly in subsistence agriculture, although there is increasing integration of the hills with the larger national economy. In addition, while previous studies have dealt with relatively plain areas, the terrain in Nepal is largely undulating and there exists a diversity of ethnic/caste groups.

Thus, this article attempts to understand the role of small towns in rural development within the context of an area with hilly terrain, severe resource constraints and ethnic diversity. It focuses on the three primary functions of such towns: (a) as market centres, (b) as employment centres and (c) as service centres.

Small towns have been defined as nucleated settlements with between 5,000 and 20,000 inhabitants (Hardoy and Satterthwaite, 1986). In the case of the hills of Nepal, where the population is scattered, even a central place with a population of fewer than 5,000 people can be considered a small town. For the purposes of this article, small towns in the hills are all assumed to be central places providing vital urban functions such as retail services, administrative services, agricultural inputs services and employment opportunities to the rural hinterland population. This definition would generally include towns of between 2,000 and 20,000 population.

The article first discusses the processes pertaining to the growth of small towns in the hill areas of Nepal. Past studies on small towns are then examined as a background for the study. The next section discusses the findings of a case study of a Nepalese town and its surrounding villages, with respect to the three aforementioned functions and linkages of small towns. Constraints of small towns in promoting rural development are identified. An outline for a strategy of small town development in Nepal is then proposed based on the current theoretical debate on the topic and the findings of the study.

Nepal is one of the least urbanized countries in the world, with only about 10 per cent of its population living in officially defined urban areas. Published data on urban areas are available only for those towns with municipal status (with a population of 9,000 or more). Since the census estimate of the urban population does not consider smaller towns that do not have municipal status, many small towns in the hills which are playing an important role as market/service centres in the area are altogether excluded from the urban definition. Hill towns are both few in number and small in size compared with towns in the Kathmandu Valley and the Terai. There are only nine hill towns with municipal status and their populations vary between 9,000 and 21,000 (except Pokhara). The hill towns with municipal status are Ilam, Dhankuta, Bidur, Banepa, Dhulikhel, Pokhara, Byas (Damauli), Tansen and Dipayal. Many other smaller hill towns which do not have municipal status have populations of fewer than 9,000 people. A list of market towns in the Nepal District Profile gives a tentative estimate of the number of small towns in the 32 hill districts as 166, an average of five market towns per district (Nepal District Profiles, 1982; cf. CEDA, 1989).

Many hill towns trace their origin to acting as administrative centres or military outposts. Another impetus for hill town growth was from the movement of Newar traders from the Kathmandu Valley to exploit the limited demand for trading activities in the hills. Most of the small towns grew along major trail routes leading to the Kathmandu Valley.

In the present context, expansion of the central bureaucracy, development of motorized transport and increase in trading activities in the hills have contributed to the development of many hill towns. Some hill towns, which are administrative centres, grew with cash flows from the national government for the maintenance of local bureaucracy and expansion of education, health and banking facilities. The increase in administrative activities at district centres has contributed to the increased opportunities for commerce in these towns to serve the civil servants and people using the administrative services.

With the development of highways, some towns along the highways have gained in importance, whereas others by-passed by such roads have declined. With the removal of protectionist trade policies, there has been increased trading of imported goods in these hill towns. Commerce is the major economic activity in towns along the highways, with few agro-processing and transport-related industries. These towns also have become the new gateways to serve the hinterland, roads from where food-grains, kerosene and salt are distributed to rural hinterland areas and locations where villagers come in search of work (Blaikie and others, 1980). Some towns have become points where travellers stop for meals and overnight accommodation and where tourists disembark for trekking. Most of the hill towns are thus engaged in tertiary activities such as commerce or administrative work.

Historically, subsistence-level agriculture, difficult terrain, limited development of hill trails and exploitative state policies, all have acted as constraints to the development of hill towns. The hill areas had fragmented "pocket economies" with subsistence agriculture and limited economic exchange between different areas in undulating terrain. The hill trails were the major link between different pockets of settlements, which were primarily suitable for local movement of people and goods in small geographic areas, thus limiting greater economic interaction (Shrestha, 1990).

As part of an integrated rural development strategy for the hills of Nepal, the Nepalese Government, in its successive development plans, has suggested the need to promote small towns in the hills. The Government stressed the development of small towns in appropriate locations in the hill and mountain regions so that they could "act as intervening points for migration, development and diffusion of socio-economic activities in the lagging region" (NPC, 1975). Recently, the Government proposed that "rural-urban linkages will be promoted through the development of small town/market centres" (NPC, 1992). However, many of the urban development projects implemented by the Government have focused primarily on the improvement of town infrastructure, preparation of land use plans for the towns and financial assessment for all municipalities, without much emphasis being given to issues of rural-urban linkages.

After the 1991 census, the Government introduced a new municipal act according to which a town can gain municipal status only if it has a population of 20,000 and minimum yearly income of 100,000 rupees (US$1.00 = Rs. 37.2 in 1991). Although this new classification will not apply to existing towns, this type of policy measure will make it unlikely that any new towns from the hills will attain municipal status in the near future, unless political pressure is applied to make exceptions to the rules. Population thresholds for municipal status should be lower for hill districts considering the small size of towns there.

A review of past studies on small towns in Nepal reveals the diverse nature of work done on the topic, as well as the research gap in the area. There are four types of study on the subject. The first type refers to research conducted primarily by anthropologists (Mikesell, 1988; Messerschmidt, 1980; Caplan, 1975), usually a detailed study of a particular town with focus on its economy and polity. Mikesell's study (1988) uses the dependency perspective to examine the historic growth of a market town in the Western Region of Nepal. Messerschmidt's study (1980) shows how a new town along a road developed while an old town declined after being by-passed by the same road. Another anthropologist, Caplan (1975) examined an administrative centre in the Far Western Region of Nepal and showed the dependence of its economy and politics on the infusion of cash from the central Government.

The second type of research on small towns includes studies carried out by geographers which are oriented towards functional analysis of a group of towns in a particular district or region (New Era, 1986; GIC, 1984). The New Era study (1986) found that hill towns acted merely as distribution points for agricultural and industrial products from the Terai and India. The other study (GIC, 1984) contends that, while small towns in the resource-poor hills are not growing, the Terai's more productive agricultural sector has contributed significantly towards the growth of urban centres in the plains.

The third group of studies examines all major towns in Nepal, providing a comprehensive picture of different levels of towns and their hierarchy (Sharma, 1989; CEDA, 1989; K.C. and others, 1991). CEDA's report indicates that hill towns are in the lower tier of the urban hierarchy as compared with Terai towns and that the former are more important as administrative centres. Both Sharma (1989) and K.C. and his colleagues (1991) comprehensively review the migration and urbanization pattern in Nepal.

Besides these three types of study, there are other miscellaneous types of papers dealing with small towns and urban-rural relations (Erthur, 1994; INFRAS, 1991; ICIMOD, 1986; RTI, 1989). Most of these studies advocate the promotion of small towns. ICIMOD's report suggests the need for developing market towns in order to generate off-farm employment in the hills in the context of the increasing inability of farm-based activities to absorb the growing labour force in rural areas. Likewise, Erthur (1994) recently suggested the need to develop a national urbanization policy with a framework of management of small towns and middle-sized cities to maximize the country's rural development potential.

Past studies on small towns in Nepal have all highlighted the importance of small towns and contributed towards understanding the internal structure of small towns and the differences between small towns in the hills and Terai. However, most studies have not made an in-depth assessment of the towns' relationships with surrounding villages in terms of the use of urban services, marketing and creation of off-farm employment. No attempt has been made to study the hinterland villages of small towns, which could better identify the constraints of small towns in promoting rural development. Many past studies are concerned with the small town's role in rural development, but they ignore the importance of hinterland villages in the growth of small towns as well as linkages between small and large towns. These are some of the issues this article attempts to address while examining a small town in the study area.

For the purpose of this case study, fieldwork was conducted in the small town of Banepa and two of its hinterland villages (Opi and Devitar) to attain an in-depth understanding of the nature of urban-rural linkages in the hill area. The linkages between the town and villages were considered from both the town-level perspective and the viewpoint of the villagers. Banepa was chosen because it had many rural development functions suggested in the small town literature. The case study approach has often been criticized on the ground that it is difficult to make generalizations based on a single case. But as Yin (1984) argues, a case study can make generalizations based on a comprehensive understanding of the phenomenon rather than merely focusing on some notion of representativeness. Likewise, Chapman and Prothero (1983:607) believe that "findings from the case studies may be generalized to other cases for which the same contextual conditions hold".

Banepa is an historical market town in Kavre District with a population of 12,537 in 1991. Situated along the Arniko Highway linking Kathmandu with the Chinese border area, this town is only about 26 kilometres (km) east of Kathmandu City (a one-hour bus ride from Kathmandu bus park). The opening of the Lamosangu-Jiri Road in 1984 has enhanced Banepa's access to the eastern hills. The service area of Banepa includes four districts in the area, namely: Kavre, Sindhupalchowk, Dolkha and Ramechhap.

Two major activities dominate Banepa's economy. Firstly, Banepa continues its historic importance as the major wholesale centre for daily essential goods such as rice, salt, kerosene and clothes, catering to people from the eastern hills. Secondly, the various cottage industries, such as handloom, sweater-weaving, carpet-making, rice-milling and dairy industries, generate non-farm employment for the townsfolk and surrounding villages. Recently, an activity which has increasingly gained in importance is transport-related businesses, taking advantage of Banepa's important access to the eastern hills. Many of the well-off businessmen own buses and minibuses running along the local regional routes including the Arniko Highway and the Lamosangu-Jiri Road.

While the town is inhabited largely by the Newar ethnic group, the surrounding villages have two major ethnic groups, namely: (a) Brahmins and Chhetris and (b) Tamangs. Between the two villages chosen for examining the rural perspective on the town, Opi (250 households) is primarily a Brahmin and Chhetri village, whereas Devitar (323 households) is a Tamang-dominated village. Both villages are within walking distance (less than 2.5 hours) of Banepa and are located in undulating terrain. People from both villages are living on subsistence agriculture, growing mainly grains. Some villagers have diversified their agriculture by growing potatoes, yams and fruit for marketing purposes. The proximity of the study area to Kathmandu has facilitated the development of a dairy industry in Banepa to fulfil the increasing demand for milk in urban Kathmandu. People from both villages walk along the hill trails to come to Banepa for marketing as well as accessing other services there.

Among the Brahmin and Chhetri villagers from Opi, there is an increasing shift towards service occupations (guards, gardeners and general labour) and administrative jobs. The range of occupations and distances travelled vary, but clearly the geographic mobility of these people is greater than that of Tamang households. Another important characteristic of Brahmin and Chhetri households is their higher literacy levels as compared with Tamangs. Some of the poorer Tamang households are also increasingly involved in wage labour such as the carpet industry and in brick kilns at Kathmandu.

Role of a small town as market centre

Banepa is the principal distribution centre for food-grains and cloth in Kavre District. Owing to the general lack of purchasing power of the rural households, the range of urban consumer goods sold in small towns primarily cater to the residents of the towns themselves. Based on a study of items sold by the wholesale and retail shops, it has been found that most of the items sold in Banepa's market are not produced or processed locally but are brought from outside the area. The majority of wholesalers in Banepa deal with food-grains and food products for the hinterland villages. The wholesalers get most of the grains directly from the Terai owing to the demand for large quantities of food-grains in remote hill and mountain regions along the Arniko Highway and Lamosangu-Jiri Road. This finding is different from the findings of studies in other developing countries with more dynamic and productive agricultural regions where the rural landlords and newly emerging urban merchants in small towns make huge profits by buying the agricultural produce from the hinterland villages and transferring such produce to large cities (Douglass, 1981; Harris, 1984).

In Banepa, the market is characterized by the predominance of retail establishments whose volume of transactions is low. With a heavy dependence on goods produced outside the area, the market is oriented towards the consumption of goods from outside rather than linked to local production. Thus, Banepa's merchants make their profits by trading in imported goods rather than selling locally produced goods.

Although the literature on small towns suggests that their development is linked to agricultural growth in the rural regions, in the context of Nepal, a small town such as Banepa grows not because of the agricultural potential of the rural hinterland, but because of the town's strategic location as a gateway to the rural hinterland with good transportation access, making it an attractive location as a break-of-bulk point and distribution centre for daily essential goods and services such as credit and agricultural inputs. Other contributing factors for the small towns' growth are the presence of local entrepreneurs and linkages with a large urban area (Kathmandu).

Banepa does not act as a collection centre for local produce as significantly as it does as a distribution centre for urban consumer goods. Only small rice mills collect rice and maize from nearby villages (larger rice mills obtain their inputs directly from the Terai). The retail fruit and vegetable market in Banepa often gets its supply from the Kathmandu wholesale market owing to the irregularity in the supply of vegetables from surrounding villages. Large surpluses of rice, potatoes and fruit belonging to rich farmers from surrounding villages often go directly to Kathmandu, by-passing the small town. In contrast, the small farmers often sell their limited produce directly to the retail establishments in Banepa to pay for items such as salt, kerosene and other daily necessities.

Although the villages of Opi and Devitar are in close proximity to Banepa and Kathmandu, as yet there has been no significant commercialization of agriculture. Most villages grow food-grain crops such as maize and rice. There is little attempt to diversify into vegetables and fruit crops. Most villagers plant small quantities of vegetables such as string-beans, cucumbers and radishes, essentially for home consumption only. A large number of households even have to purchase food-grains from Banepa owing to the small size of landholdings. The average size of cultivated landholdings is about 0.75 hectare (ha) and about 60 per cent of the households in both villages have less than 0.75 ha of land. According to one estimate, the minimum land required for the subsistence of a Nepali hill family with four members is 0.75 ha (Blaikie and others, 1980).

Maize and potatoes are two important items sold from the village. But in Opi village, only 20 per cent of the sample households sell maize, which is the most important crop in the hills. The value of the amounts sold by households ranges from Rs. 300 to Rs. 4,500 per year (average Rs. 1,548 among those who sell), indicating a relatively small volume of these crops for sale. Among those who sell potatoes, the average amount sold ranges from Rs. 150 to Rs. 10,000 per year. The situation in Devitar is not much better, with similar low levels of agricultural produce being marketed. Owing to limited agricultural surpluses, villagers sell their produce directly to retail shops in Banepa. Often the farmers sell the produce on an individual basis, and thus are not in a position to bargain for better prices. Although chemical fertilizer is widely used in the area, the use of pesticides and improved varieties of seed for higher production is limited.

The constraints on Banepa acting as a collection centre for agricultural produce have roots in the small size of landholdings and limited diversification of agriculture in the hinterland villages. The problems are further compounded by the difficult terrain, which limits the amount of cultivable land, and the lack of rural infrastructure (such as irrigation facilities, feeder roads to the village). Also, a marketing culture and entrepreneurial skills among the villagers are not well developed, which further constrains their ability to diversify agricultural production for sale.

Role of a small town in providing employment

In order to examine whether small towns such as Banepa provide an important source of employment for the hinterland population (as suggested by the proponents of small town development), the employment structure of Banepa has been analyzed both in terms of types of job created and the beneficiaries of the jobs. Banepa provides employment for the people in the area through about 55 cottage/household industries, 438 commercial establishments and 13 government/semi-government institutions.

Industries in Banepa are dominated by craft/household industries (e.g. cotton fabric, carpet- and sweater-making) and primary agro-processing (e.g. rice milling and vegetable oil production). Aggregating the employment data of industries in Banepa, it was found that 49 per cent of the employees are from Banepa itself, whereas people from surrounding villages and distant villages constitute only 38 per cent and 13 per cent, respectively, of the work force. In terms of ethnic status, it was found that Newars constitute 80 per cent of employees in the industries, followed by Brahmins and Chhetris (10 per cent), Tamangs (8 per cent) and other groups (2 per cent). While enterprises such as rice mills and furniture industries employ mainly family labour, other industries such as handloom, carpet- and sweater-weaving industries have a much higher proportion of hired labour. The largest handloom industry employs as many as 80 workers from surrounding villages who often weave fabrics at home and are lowly paid on a piece-rate basis.

A review of the employment records of offices in Banepa indicates that, of about 495 people employed in service-sector jobs, 49 per cent are from Banepa town itself, followed by 31 per cent from surrounding villages and 20 per cent from other areas. Often, higher-level jobs are handled by the town dwellers, while people from surrounding villages have lower-level jobs. Newars account for 60 per cent of employees, followed by Brahmins and Chhetris (34 per cent). Tamangs have less than 2 per cent representation in employment in the small town's offices, although they comprise 32 per cent of the population of the district.

When the above employment data are compared with town-level and district-level population data, the under-representation of people from surrounding villages in jobs available in the small town becomes evident. Banepa's population (12,537) constitutes only 3.9 per cent of the Kavre District population (324,329), but about half of the employees in Banepa's industries and institutions are residents of Banepa. Similarly, the local Newars dominate virtually all the jobs in the small town. Thus, while proponents of small towns argue that these towns generate jobs for people in the surrounding villages, there are clear ethnic and spatial biases in the employment pattern in Banepa, favouring the town residents.

The dairy industry is probably the only exception as it does not generate much employment in Banepa itself, but it provides income-earning opportunities in surrounding villages. Banepa's milk industry has created the backward linkage of buffalo raising which provides opportunities to sell milk to about 600 villagers. This dairy industry in Banepa benefits the surrounding villages more than Banepa itself and this has major implications in terms of the small town's ability to generate income opportunities in surrounding villages.

Decline of old and emergence of new industries in Banepa

Two new industries which have grown recently in Banepa are the carpet- and sweater-weaving industries, both of which operate on a subcontracting system from Kathmandu-based business establishments. Part of the reason for the growing carpet industry in small towns like Banepa is the increasing government restrictions on carpet industries in Kathmandu and search for alternate locations with less scrutiny from the Government.

In contrast with the growth of industries operating under the subcontracting system (carpet- and sweater-weaving industries), other industries such as handloom industries, which produce goods that have to compete with mass-manufactured products from larger industries in Kathmandu and foreign countries, have generally declined. These findings about declining old industries and emerging new ones imply the need for selective promotion of certain types of industries in small towns such as Banepa where industries have some chance of success.

Certain industries which have linkages with villages in terms of procurement of raw materials and linkage with larger towns for final output/processing also have potential for growth in small towns. The contribution of the dairy industry in generating non-farm cash income for villagers selling milk is a case in point. In fact, the dairy industry is one of the few industries in Banepa which has largely benefitted the villages rather than the town.

Off-farm employment in villages

There are limited opportunities for off-farm employment in the villages. Most people with small land holdings are involved in wage labour during agricultural seasons for about three months. A few families from the villages go to work in brick kilns and carpet industries in Kathmandu; a greater number of Brahmins and Chhetris from the villages are engaged in service employment in Banepa and Kathmandu. In comparison, the number of Tamangs from Devitar village holding administrative/service employment is almost insignificant. About 30-40 people from the villages have skills in carpentry and masonry and they are engaged in building construction in the villages. But very few of them work in construction work in Banepa and Kathmandu owing to their low levels of those skills.

Surveys in the villages showed that many villagers want to own a buffalo so that they will be able to sell its milk. There are milk collection centres both in Opi and Devitar and they are providing important sources of income for the villagers. Some villagers even rent cows and buffaloes from others and sell milk as a source of cash income. However, owning buffaloes is an expensive proposition for the majority of villagers.

Role of a small town in providing urban services

Urban services in small towns can be divided into two categories: (a) town-serving functions used primarily by town residents and (b) region-serving functions catering to the benefit of surrounding villages as well. The schools in Banepa primarily cater to the residents of the town, in spite of the fact that educational facilities in the villages are quite poor. A survey of two schools in Banepa reveals that 88 per cent of all the students studying in these schools are from Banepa alone while the remaining 12 per cent are from hinterland villages (Bajracharya, 1994). About 71 per cent of the students are Newars and another 24 per cent are Brahmins and Chhetris. Tamangs and other ethnic groups account for only 5 per cent of the students. Literacy data from village household surveys have shown that the illiteracy rate in the Tamang village (78 per cent) was much higher than in the Chhetri village (53 per cent).

Agricultural inputs and credit services in the small town are found to be largely region-serving functions benefitting hinterland villages. Interviews with owners of two chemical fertilizer shops in Banepa indicated that about 65 per cent of their customers are from the villages, and 35 per cent of them are from Banepa town, which demonstrates the rural bias in the use of this service. Villagers agree that there has been increased use of fertilizers in the villages since these products became available in Banepa. The use of fertilizers has increased agricultural production compared with the time when such fertilizers were not used. However, much of the gain in production has been more than offset by the consumption of the growing village population.

Regarding agricultural credit facilities, a study of loan patterns (for 1989-1990) from the Agricultural Credit Bank located in Banepa indicates that it primarily caters to residents of the villages rather than the small town. About 73 per cent of loan takers were from the surrounding villages, while another 24 per cent were from distant villages (Bajracharya, 1994). Only 3 per cent of loan takers were from Banepa. Longitudinal data on loans taken from the Bank in different time periods indicate some gradual increase in loan investments in marketable products and livestock development. Thus, Banepa's agricultural credit facilities have to some extent contributed to the diversification of agriculture, horticulture and livestock development in the surrounding villages.

In the past, many villagers used to grow millet and wheat in the winter. In the last decade, there has been a rise in the number of families growing potatoes instead of wheat, the former having a higher market value. Likewise, there has been a greater use of improved varieties of potatoes for increasing production. According to the villagers interviewed, currently more than 80 per cent of the potatoes grown are produced for the market and only 20 per cent for self-consumption.

However, even among region-serving functions, there are some spatial, ethnic and economic class biases. Villages close to the town have availed themselves of the services more than the distant villages. With regard to credit facilities, there is a linkage between types of loan taken and distance from the town. People from surrounding villages borrow money to grow cash crops (such as potatoes and fruits), whereas those from distant villages took loans primarily to grow subsistence crops such as maize. The major beneficiaries of agriculture-related services are large-farmers with financial resources to present collateral, pay for fertilizer and improved seeds, and obtain access to knowledge concerning the use of services. In this respect, ethnically, Brahmins and Chhetris are the largest beneficiaries of the Bank. People from occupational castes such as Sarkis and Damais, and ethnic groups such as Tamangs have not been able to take advantage of the credit services in Banepa. While the truly needy small-scale farmers cannot obtain loans owing to their lack of collateral, some influential farmers get larger loans than necessary. Many villagers do not have the perseverance and time to go constantly to Banepa to follow up the loan process, which could take months. Also, many villagers often need credit for social purposes such as a marriage or festival, for which the Bank does not provide loans; thus, they have to depend on the local money lenders in the village for such purposes.

With respect to hospital services, Banepa's hospital has been found to have a region-serving function in terms of its use by patients. Hospital records show that, while the staff of the hospital are predominantly from Banepa itself, the patients are largely from surrounding villages and nearby smaller towns, which are within walking distance of Banepa. Household interviews in the villages indicate that the people use the hospital as a last resort. Most villagers have a sequential way of getting treatment, progressing from a local faith healer to a local health worker in the village to private pharmacies in Banepa, finally to the hospital when all other avenues have been exhausted. Owing to the lack of transport, relatives of the ill often find it convenient to go to pharmacies in Banepa, describe the patient's problem or symptoms to the store owner, and purchase medicine without bringing in the patient. The hospital, on the other hand, dispenses medicine only after examining the patient. Therefore, many villagers depend on alternative treatment methods including traditional faith healers.

In other words, while both the Agricultural Credit Bank and hospital have region-serving functions, the schools are primarily fulfilling a town-serving function. If investments are made in region-serving functions, there is a greater potential for these services to benefit the people from hinterland areas rather than those from the town only.

Constraints of small towns in promoting rural development

It is clear from the above empirical findings that Banepa's present status does not match exactly with the rosy picture of small towns as assumed by their advocates. Several interconnected factors account for this divergence of actual and expected roles of small towns. Firstly, at the town level, the local government in Banepa lacks decision-making power, technical expertise and financial resources to play a greater role in encouraging a larger and more diversified economic base of the town and its hinterland. With the centralization of power and resources in the Kathmandu Valley, much of the development budget of line agencies is dictated primarily by the head offices in the capital, without much input from the local area. There is little articulation of the needs and priorities of local people in such centralized decision-making processes where most of the resources also go to larger cities.

The industries in the small town have limited backward and forward linkages with the villages. Currently there is very little processing of local products in Banepa. Except for small rice mills, most larger rice mills in the small town obtain their grain directly from the Terai, thus eliminating backward linkages with the surrounding villages. The dairy industry, which is still in its nascent stage in the small town, is probably one of the few industries which have backward linkages with the surrounding villages. Most industrial and commercial firms have few downward linkages, but they have more linkages with the larger towns in Kathmandu and the Terai for the supply of raw materials and destination of their finished products. Because the productive base of small towns is limited, there is penetration of the local market by outside industries with greater economies of scale, which contributes to a further decline in the cottage/household industries in the region. The existing administrative machinery provides little scope for planning for urban-rural linkages between the small town and surrounding hinterland villages. The local municipality government often does not consult with villagers to identify projects with complementary benefits. The town often plans its development activities without taking the surrounding villages into consideration.

Turning to the village level, the low purchasing power of the rural people prevents the villages from acting as markets for industries in small towns. Small landholdings coupled with limited resources, organization and information constrain the possibility of agricultural marketing through the wider introduction of vegetable and fruit production. The majority of farmers in the villages near Banepa have small landholdings (less than 0.75 ha in size) which are not large enough to support their families using traditional subsistence agricultural practices. With increasing population pressure being put on the limited resources of the hill economy, villagers have to depend increasingly on outside non-farm employment (such as casual labour) to make ends meet. But because of the lack of adequate skills in urban household industries (e.g. weaving, knitting), the lack of education to qualify for administrative jobs, and the limited number of jobs available in small towns, many villagers find it difficult to obtain work even in a small town such as Banepa. There is also a lack of financial capital for starting up rural industries. The poor spatial linkage between the town and surrounding rural areas inhibits rural residents from availing themselves of agricultural inputs such as fertilizers as well as from transporting their products to the town.

The Government's policy to concentrate infrastructural investments in a few larger towns of the Kathmandu Valley and the Terai is having an adverse impact on small hill towns. Likewise, there is an urban bias in many sectoral policies of the Government and these favour the capital city. The low prices for agricultural produce and lack of investment in improving village agriculture, or agricultural credit and marketing, all contribute to the depression of rural areas. Lack of adequate services such as educational facilities in the villages make the villagers less able to compete for urban jobs. The levels of investment in infrastructure and services in the hill area's small towns and villages are significantly less than in the capital city and the Terai.

The general findings which have emerged from the case study of Banepa and two villages are as follows:

1. In the context of subsistence agriculture, a small town plays an important role as a service centre for the distribution of basic essential goods (such as food-grains, salt, kerosene and fabrics) for the hill and mountain areas, but it plays a limited role as market centre for rural produce from local areas and as an employment centre for rural people. A small town in a resource-poor area is less diversified than a town in regions with agricultural surpluses. The hill town has more consumptive functions than productive functions.

2. Limited arable land and lower levels of production in the hill areas of Nepal have an adverse impact in the growth of both the rural areas as well as the towns located in the rural areas. Difficult terrain and lack of access to roads in many villages constrain the movement of people, goods and services in the hills.

3. Nepal's rural society is culturally heterogenous with many different ethnic/caste groups. This study found that ethnicity and place of residence of people are important variables in the access to off-farm jobs and services available in the small towns.

Two major approaches to small-town development have emerged in the literature on this topic and the article discusses each of them briefly to determine their applicability in the context of Nepal (Dixon, 1987; Hinderick and Titus, 1988; Unwin, 1989).

The first approach, called "market-oriented regional development", argues for a greater role of the private sector in the development of small towns and considers the need for local authorities to be more entrepreneurial rather than mere service providers (Gaile, 1992). The focus of the strategy is government investment in infrastructure (power, roads etc.) to enhance the small town's role in agricultural marketing from the hinterland villages as well as to improve the marketing system by skill development, and access to credit for farmers to produce crops for sale on the market. This strategy parallels the earlier approach of "urban functions in rural development" (UFRD) which argues for integrating small towns with the interlinked hierarchy of urban centres for facilitating the efficient exchange of goods and services between rural areas and urban centres (Rondinelli, 1984). The primary focus of both the "market-based regional development" and "urban functions in rural development" approaches is to bring about improvements in the physical access to services and markets for the rural people, without suggesting any changes in the political economy of the rural areas for the benefit of the rural poor.

The other approach of territorial development assumes that small towns within a self-reliant territorial unit can promote rural development through greater local control over resources, complementary urban-rural linkages and decision-making at the local level. The approach suggests measures such as rural industrialization, diversification of agriculture, retention of surpluses at the local level and land reform (Friedman and Douglass, 1978). Linked to the idea of territorial development is the concept of "selective spatial closure" of rural areas to check the leakages from rural to urban areas and encourage greater investment of rural resources in the local economy. Recently, Douglass proposed a "regional network model" in which he argues that a territorial unit with a network of towns and their hinterland rural areas provides a larger diversified regional economy which could be a strong alternative to a conventional, single growth centre founded on urban-based manufacturing (Douglass, 1992). He also highlighted the importance of the region-serving functions of small towns.

The basic difference between the two approaches is that, while the former approach emphasizes the need for market orientation/integration, the latter approach highlights the need for decentralization of power at the local level. These two approaches need not be mutually exclusive; rather they can complement each other. Integration of small towns and rural areas with the larger economy without political and institutional support to the former can work to the disadvantage of small towns and rural areas, with leakages of local resources to the outside. That is why the territorial development approach's suggestion to retain surpluses in the local area is relevant in developing countries. What is advocated in this study is market-oriented territorial development as part of a strategy for hill area develop-ment, drawing from both approaches.

The key elements of the strategy are: (a) market integration of hill economies, (b) decentralized planning in small territorial units which include both small towns and groups of villages, (c) institutional support for the rural poor, (d) expansion of off-farm employment and (e) investing in region-serving functions of small towns.

Market integration of hill economies

Since agriculture is the mainstay of the villages, any programme concerning small town development has to take into consideration agricultural development in the rural areas. In the context of Nepal's hill areas, concepts such as "selective spatial closure" cannot be pursued since the local economy cannot sustain itself by being isolated. As Banskota (1989) argues, hill agriculture needs to be integrated into the larger economy for increasing production, selling the produce in the market and increasing opportunities for off-farm employment. Attempts have to be made to diversify subsistence agriculture to include (high-value) cash crops based on local comparative advantage suitable for hill climate and terrain rather than limiting production to subsistence grain crops. Small farmers need to be encouraged to set aside a certain percentage of land for cash crops (such as potatoes, vegetables or fruits) for selling in the market and generating cash income. To facilitate direct marketing by the villagers, the Government could promote the development of periodic and permanent farmers' markets through the provision of market space and paved areas, weighing/grading facilities and overnight storage facilities.

Instead of merely marketing agricultural produce, the processing of these at the village level offers greater potential for adding value from processing which can be retained at the village-level. Potential industries in the small town could enter into direct contracts with groups of farmers to ensure a regular supply of raw material for the industries. These measures would contribute not only towards improving the local economy by providing marketing avenues for rural products, but also towards providing additional off-farm employment opportunities for the local people.

Decentralized planning in small territorial units

The degree of decentralization of power and resources by the central Government largely determines the role of local people in decision-making. This situation also affects the development of small towns since they are often the local nodes in the rural areas where government investments are decentralized at the local level. In a situation where the different line agencies in Kathmandu largely determine the nature of government investments at the local level, small towns have not been particularly instrumental in encouraging greater involvement of local people nor in coordinating the work of line agencies at the local level. Local government needs more power and resources with further delegation of financial and administrative authority at the local level to stimulate production in small towns and hinterland areas. Such a development would have better chances of success in mobilizing resources and involving people for locally prioritized development.

Under Nepal's "District Development Plan", the basic territorial unit for plan formulation and implementation is the district. But with the present districts spread over a large territories and in mountainous areas, there are difficulties in coordinating district-level activities. In the context of undulating terrain and scattered settlements, a more appropriate strategy would be to divide the district into smaller subdistrict territorial units with at least a small town and a small group of villages having economic, spatial or social linkages to it. Although the service centre concept (with division of the district into nine areas) was proposed during the early 1980s, such service centres were conceived primarily to distribute government-sponsored services to the villages. The division of the district was often done arbitrarily without consideration of the social and spatial linkages between the villages and the service centre. They were not designed particularly to help commercialize agriculture, create non-farm jobs at the service centre level, or to harness the potential of the service centre with private sector efforts.

The current practice of small-town planning, without giving adequate attention to hinterland villages, needs to be changed, too. The services in small towns are not only for the towns alone, but for the hinterland villages as well. Hence, while demarcating territorial units for planning, small towns need to be seen as a part of their hinterland service area. Each of these towns and their hinterland villages, instead of being seen in isolation, need to be viewed together while developing district-level development plans and policies.

Institutional support for the rural poor

In view of the small size of surpluses, support needs to be given to encourage group marketing of hill produce at larger urban centres. Group marketing could be encouraged through a "small farmer development project" which could provide group loans to people for growing a particular crop and marketing it together (for example, for renting farm implements, providing marketing information, helping to negotiate with outside parties, teaching modern methods of grading, weighing, cleaning, storing, packaging and transporting). Group transport would also bring down the cost of transport. Farmers who have a disincentive to produce marketable items owing to inadequate returns and the necessity for individual transport may be more inclined to produce for more distant markets that provide better prices if group marketing and transport are made available. In this way, they would also be in a better bargaining position with regard to the pricing of their products.

The people from surrounding villages often lack adequate skills and education to get jobs in urban areas. Job training and skill development of villagers should be an integral part of the strategy for small town development. Vocational training and adult literacy programmes in the village could facilitate those disadvantaged groups to enter the job market outside the villages. Thus, local governments would need to organize necessary training in production and marketing.

Expansion of off-farm employment

Off-farm employment opportunities in small towns and hinterland villages need to be expanded. For this, the comparative advantages of individual small towns need to be identified. For example, Banepa's comparative advantages - excellent location with easy transportation access to Kathmandu and the eastern hills, the availability of local entrepreneurs and cheap labour - make the town a potential area for expanding export-oriented carpet and garment industries. Such a development could considerably increase employment opportunities in such a small town. Also, owing to the small town's proximity to surrounding villages, the people in the area, instead of relocating to the town, could commute to work from their villages.

The small cities may be encouraged to develop goods and personal services for local consumption as well as to produce intermediate goods for larger industries in the bigger cities. Some of the potential industries for promotion at small urban centres may be the production of household items, handicrafts for export and similar businesses. As the Banepa case study shows, those industries in small towns which compete directly with industries in large towns are in a general state of decline, but other industries such as sweater weaving and carpet manufacture are doing relatively well. Therefore, it would be more prudent to encourage the formation of industries in small towns which would have higher chances of succeeding by working in cooperation with larger industries in Kathmandu or other large towns than by competing with them.

Priority also has to be given to industries which can generate employment for villagers as well as for urban residents. The dairy industry is a good example. Since the dairy industry benefits many villagers, additional efforts need to be made in livestock development by encouraging the raising of animals that yield higher quanties of milk, for example, as well as by providing small farmers with better animal health services, fodder and credit.

Investing in region-serving functions of small towns

The Government needs to invest in region-serving functions rather than town-serving functions while investing in small towns so that the surrounding villages can also benefit from investments in the town. Since there are ethnic and spatial biases in the use of services, greater attention must be paid to reaching disadvantaged groups in the use of services through small farmer development programmes, and minority group economic development programmes.

This study found that, although fertilizers are widely used, the use of pesticides and improved varieties of seed is quite low, partly owing to a lack of information and knowledge about their effective use. In order to diffuse agricultural innovation, there is a great need for agricultural extension services. The extension service could include advice on marketable products and crops, information on market prices, advice on how, when and where to sell agricultural products, the promotion of group marketing activities and facilitating production and marketing arrangements between farmers and mill owners and traders etc.

In order to promote small towns for rural development, a small number of towns in each district which have a potential for developing region-serving functions to the hinterland villages could first be identified. The selection of small towns for further investment needs to be based on such criteria as their capability for agricultural diversification, centrality in serving a larger population, availability of rural infrastructure such as spatial links to villages, potential for processing industries, and scope for creating off-farm employment.

Local area development plans could be formulated, based on the understanding of urban-rural linkages between the small town and its hinterland villages, which later could be part of overall district development plans. These development plans could identify the constraints to marketing, service provision and non-farm employment, for which an appropriate package of investments could be suggested. These could take many forms, such as building town-to-village trails, developing a regular or weekly producers' market in the town, provision of agricultural credit, input and extension, skill training for certain jobs and the development of processing industries based on local comparative advantage. Such a multifaceted approach could provide a useful basis for integrated urban-rural area development for the benefit of rural areas in Nepal.

Bajracharya, B. N. (1994). Small towns and rural development: a study of urban-rural relations in the hill region of Nepal. Ph.D. dissertation, University of Hawaii, Honolulu.

Banskota, M. (1989). Hill Agriculture and the Wider Market Economy: Transformation Processes and Experiences of the Bagmati Zone in Nepal. Occasional Paper No. 10 (Kathmandu: International Center for Integrated Mountain Development).

Blaikie, P., J. Cameron and D. Seddon (1980). Nepal in Crisis: Growth and Stagnation at the Periphery (New Delhi: Oxford University Press).

Caplan, L. (1975). Administration and Politics in a Nepalese Town (London: Oxford University Press).

Centre for Economic Development and Administration (CEDA) (1989). Inter and Intra Regional Flows in Nepal (Kathmandu: CEDA).

Dixon, D. (ed.) (1987). Rural-Urban Interaction in the Third World (London: Developing Area Research Group, Institute of British Geographers).

Douglass, M. (1981). "Thailand: territorial dissolution and alternative regional development for the central plains", In: W. Stohr and D. R. F. Taylor (eds.) Development from Above or Below: The Dialectics of Regional Planning in Developing Countries (New York: John Wiley and Sons).

__________(1992). Global interdependence and urbanization: planning for the Bangkok mega-urban region. Unpublished manuscript, Discussion Paper No. 26. Department of Urban and Regional Planning, University of Hawaii, Honolulu.

Economic and Social Commission for Asia and the Pacific (ESCAP) (1991). Small Town and Rural Human Resources Development to Reduce Migration to Large Cities (New York: United Nations).

Ertur, O. (1994). "The need for a national urbanization policy in Nepal", Asia-Pacific Population Journal, 9(3):19-36.

Friedmann, J. and M. Douglass (1978). "Agropolitan development - towards a new strategy for regional planning in Asia", In: Growth Pole Strategy and Regional Development Policy (Nagoya, Japan: United Nations Centre for Regional Development).

Gaile, G. L. (1992). "Improving rural-urban linkages through small town market-based development", Third World Planning Review, 1(2):131-148.

Geography Instruction Committee (GIC) (1984). Role of Small Urban Centres in National Development in Nepal (Kathmandu: GIC, Tribhuvan University).

Hardoy, J. E. and D. Satterthwaite (eds.) (1986). Small and Intermediate Urban Centres, Their Role in National Development in the Third World (Toronto: Hodder and Stoughton).

Harris, B. and J. Harris (1984). "`Generative' or `parasitic' urbanism? Some observations from the recent history of a South Indian market town", The Journal of Development Studies, 21(3):82-101.

Hinderink, J. and M. J. Titus (1988). "Paradigms of regional development and the role of small centres", Development and Change, 19(3):401-423.

Interdisciplinary Consulting Group for Natural Resources and Infrastructure Management (INFRAS) (1991). Rural-Urban Linkages: A Challenge for Swiss Development Corporation. Interim Report, July 1991 (Zuerich: INFRAS).

International Center for Integrated Mountain Development (ICIMOD) (1986a). People and Jobs in the Mountains. Report of the International Workshop on Off-Farm Employment Generation in the Hindu Kush-Himalaya Region (Kathmandu: ICIMOD).

__________ (1986b). Towns in the Mountains. Report of the International Workshop on Planned Urbanisation and Rural-Urban Linkages in the Hindu Kush-Himalaya Region (Kathmandu: ICIMOD).

Jones, G. W. (1991). "Urbanization issues in the Asian-Pacific region", Asian Pacific Economic Literature, 5(2):5-33.

K.C., Bal Kumar, P.P. Timilsina and P. L. Joshi (1991). "Nepal", In: Small Town and Rural Human Resources Development to Reduce Migration to Large Cities (New York: United Nations).

Leinbach, T. R. (1992). "Small towns, rural linkages, and employment", International Regional Science Review, 14(3):317-323.

Manandhar, M. S. (1985). Off-farm employment: the role of small towns. ICIMOD internal document, Kathmandu.

Mathur, O. P. (1982). "The role of small cities in national development re-examined", In: O. P. Mathur (ed.) Small Cities and National Development (Nagoya, Japan: UNCRD).

Messerschmidt, D. A. (1980). "Gateway-hinter relations in changing Nepal", Contributions to Nepalese Studies, 8(1):21-39.

Mikesell, S. (1988). Cotton on the Silk Road: subjection of labor to the global economy in the shadow of empire (or the dialectic of a merchant community in Nepal). Ph.D. dissertation, University of Wisconsin, Madison.

National Planning Commission (NPC) (1992). The Eighth Plan 1992-97 (Kathmandu: NPC).

__________ (1975). The Fifth Plan (Kathmandu: NPC).

New Era (1986). Development of small towns in Gandaki growth axis, central Nepal. Report prepared for International Development Research Center (IDRC), Kathmandu, Nepal.

Planning and Development Collaborative International (PADCO) (1990). Nepal: urban development policy study. Management Support for Urban Development Project. HMG/UNDP/World Bank/USAID, January 1990, Kathmandu.

Research Triangle Institute (RTI) (1989). Workshop on Market Towns, Kathmandu, Nepal, 11-14 September 1989. MHPP, USAID.

Rondinelli, D. (1984). "Cities and agricultural development: the urban-rural connexion", In: Regional Development Dialogue (Nagoya, Japan: UNCRD).

Sharma, P. (1989). Urbanization in Nepal (Hawaii: East-West Population Institute).

Shrestha, C. B. (1973-74). "The system of central places in the Arniko Rajmarga", Himalayan Review, 5(6):19-39.

Shrestha, N. R. (1990). Landlessness and Migration in Nepal (Oxford: Westview Press).

Taylor, D.R.F. (1981). "Role and functions of lower order centres in rural development", In: F.C. Lo (ed.) Rural-Urban Relations and Regional Development (Nagoya, Japan: UNCRD).

Titus, M. J. (1991). "A structural model for analysing production relations and regional functions of small towns in Central Java", Tijdschrift voor Econ. en Soc. Geografie, 82(4):266-281.

Unwin, T. (1989). "Urban-rural interaction in developing countries: a theoretical perspective", In: Robert B. Potter and T. Unwin (eds.) The Geography of Urban-Rural Interaction in Developing Countries (New York: Rutledge Books, Inc.).

Yin, R.K. (1984). Case Study Research (Beverly Hills: Sage Publications).

For further information on this material please contact: loftus.unescap@un.org