Torgeir Fyhri 1998
Chapter 6: Responses to Household Constraints and Farm Risk
CHAPTER 6: Responses to Household
Constraints and Farm Risk
The former chapter concluded that the national agricultural production were influenced, through the farmers action, by risk reduction at a farming household level. As discussed in chapter 3, the farmers tend to gear their production towards risk avoidance's and food security rather than to fluctuating prices. Producing the main cash crop, groundnut includes more risk than producing the domestic consumed crop early millet. In addition, subsistence production is less risky than cash crop production due to safer and less cash requiring food situation. At a national level the crop sub-sector is stagnating as a result of this development, due to less national income of cash crops and less self-sufficiency on food (Willis 1996). At a farm level, this is a matter of risk responses (Taal 1989). These responses are rapid and basic features in the Gambian crop sub-sector. The Gambian farmers have to respond rapidly to the variable ecological cultivating conditions of the country. In present days, this variability also includes a capital factor, which means different or additional responses. Income substitution is a new kind of risk response as the income of cash crop production has decreased while the demand for cash income in the rural farm household is increasing. As discussed in chapter 3, the risk responses vary both in time and space, because the risks vary in time and space. Both chapters 4 and 5 demonstrated failure of planned national agricultural policies in the post independence era resulting into lower national income and self-sufficiency degree. Lack of attention to farm risk and decisionmaking at a farm level might be an important shortcoming of these policies.
The questions raised by the production statistics of chapter 5 are not easily explained. Even with access to statistical agricultural data, phenomena as the large yearly fluctuations in yield per hectare of the crops and cultivated area are still questions. To illustrate and go thoroughly into the discussions of chapter 5 and the response issue, studies at a lower geographical level of the Gambian crop-sub sector will be presented in this chapter. The quantitative data used in the former chapter have limitations because little can be said about the causes of why farmers act like they do. There is a need for additional data found at a local level or at a divisional level. When studying responses at a local level, farmer constraints might be recognised.
The starting point is a field study in villages around the town of Farafenni in the Upper Baddibu district, North Bank Division, conducted in November and December 1996. The field study is supplied by a tenure and resource study of the same area, conducted in 1994 by USAID and the government of The Gambia (Bobb et al. 1994). Other examples and local and divisional statistics are also added. This is, however, an illustrative case and not a complete analysis of different geographical conditions and variations of risk responses in The Gambia. Therefore, care must be taken to avoid generalising from some of the findings, as only 6 villages were visited.
6.1: Map of Upper Baddibu (Bobb et al. 1994)
The Gambia is split up in divisions as illustrated in the map of The Gambia, Figure 1.1. The North Bank Division (NBD) is the main agricultural area of The Gambia, with an agricultural population of 133 245 (Bobb et al. 1994). In the 1995 season the NBD had the greatest production of both swamp rice, early millet and groundnuts in the whole of The Gambia (DoP 1996). In a farming constraint analysis presented by Torrence (1991), the NBD was also claimed to be the division with most farming constraints, especially related to land and labour availability and tenure. The field survey and questionnaire which form the main data background of this chapter, was conducted the Upper Baddibu District of the NBD. In the 1993 population census (The Gambia 1993) 55 438 lived in the Upper Baddibu District. 20 956 lived in Farafenni which is the rapid growing town of the district. The rest of the population lived in villages spread around the town, mainly living of agriculture incomes. Land scarcity and tenure constraints, which are related to the expansion of the town Farafenni and the establishment of the Pakala Forest Park have made the future in farming uncertain (Freudenberger 1993a, Bobb et. al. 1994 and Willis 1996). The district includes the Chamen Self Development Centre, which is the main school of agriculture in The Gambia.
The district, and especially Farafenni, almost placed at the border, has been a major area in the re-export trade to Senegal. The main road connecting the northern and the southern part of Senegal goes through the town and makes Farafenni an important trade centre. This is also of importance to the farmers. Groundnuts can easily be sold to Senegal, due to short transport distance. In addition, the border in the district is almost non-existing and uncontrolled, except from the checkpoint of Farafenni. In this district 6 villages were visited and farmers living there were interviewed. The six villages visited, as shown in figure 6.2 were Maka Farafenni, Dutabulu, Tankanto, Kalataba, Kubendar and Balingo. The villages differ from each other in both location and size. 675 people lived in Maka Farafenni in the 1993 Census (The Gambia 1993), while only 36 lived in Tankanto at the same moment. To correlate size and location with different findings of the questionnaire, some of the findings are presented schematic in the tables 6.1, 6.2, 6.3 and 6.4. The questionnaire is rendered as a whole in Appendix.
Table 6.1: Inhabitants in the in the six studied villages of Upper Baddibu 1993. (Source: Gambia 1993).
In the analysis of the questionnaire other problems and questions were raised than in the analysis at a national level in chapter 5. Below, the opinion of the farmers and responses to resource conflicts, farm risk and farming constraints in the district will be analysed and related to the findings of land, labour and tenure in chapter 5. 1
Some of the findings of this case study are rather area specific, but other findings are also found in corresponding studies in The Gambia. However, other farming constraints become visible in this local study than in the national statistics. Therefore, both the questionnaire and the corresponding studies confirm the viewpoint that the agricultural development in The Gambia cannot be fully understood without studying different geographical and decisionmaking levels.
Figure 6.2: Villages around Farafenni
The village of Maka Farafenni is placed in the upland area 4 km west of Farafenni, close to the Senegalese border in north and the Pakala Forrest Park in west (see figure 6.1 and 6.2). Most of the compounds and villagers are Wollof, mainly into agriculture and crop production. Almost all the available land of Maka Farafenni is used for agricultural purposes. These agricultural lands are classified into three distinct soil types: jorr, bahaleh and dak. Jorr is a reddish sandy soil, considered as a very good soil for groundnuts, if fertiliser is available, and good for millet. It is easier to cultivate than the more fertile bahaleh, but only one-quarter of all families, those with founding linkages in the village, have primary rights to the fields on the jorr. However, farmers without primary rights are able to borrow jorr land from the Senegalese neighbouring village Ngueur Anglais. Lower population density due to male emigration from the Senegalese village has left a surplus of land on the jorr to the villagers of Maka Farafenni.
Bahaleh is found in the transitional zone between the jorr and the dak. It is more fertile than the jorr, but more difficult to cultivate due to a high clay content. A mix of early millet and groundnuts are usually planted on the bahaleh. The founding families of the village have user rights to three-quarters of the bahaleh land.
Dak is silty clay, fertile, but difficult to cultivate. Therefore, dak land is the least preferred agricultural land in the village, used mainly for groundnuts and millet, by families without access to land elsewhere (Bobb et al. 1994).
One of the conclusions of the former chapter was that even if there is enough land in absolute terms in The Gambia, relationship between land and labour had led to fixed land scarcity at a local level. The problem is even more complex, as discussed above; land varies in quality. In addition, founding families have more secure user rights to land than late settlers do or slave descendants (Bobb et al. 1994, Willis 1996) do. According to Torrence (1991), NBD is the area of the Gambia with the greatest pressure on land (Torrence 1991). The viewpoint is confirmed by the tenure and resource study of Upper Baddibu (Bobb et. al. 1994), were shortage of land available for agricultural production was claimed as the farmers greatest constraint.
In the field study conducted for this thesis, the farmers of the six villages were asked about the land and labour situation in their villages (table 6.2). However, land scarcity was not a constraint in all villages, as discussed in chapter 5. Tankanto and Balingho had enough land, but scarcity of labour. Contrary, Dutabulu, Kalataba and Maka Farafenni had enough labour, but scarcity of land. In Kubendar neither land nor labour availability were claimed as constraints.
However, in each village the farmers claimed that some households were lacking either labour or land. This point is illustrative. Also within the villages, a misrelationship between land and labour was existing. In Maka Farafenni landholdings are parcelled among the male members. Women cultivate individual fields given them by their husbands or their family. The system is a constraint to individual ownership, because secure user rights to land are fixed, and mainly held by founding families. Land transfer is restricted, and men wishing to cultivate larger areas have to borrow land from neighbouring villages (Bobb et. al 1994).
Table 6.2: Land and Labour availability in the Upper Baddibu
On the question if there had been any change in the cultivated area the last ten years within the village, all the villages except Kubendar and Maka Farafenni claimed that there had been no change due to either lack of labour or land. In Kubendar, former pastures and bush were cultivated at a large scale the last ten years. Contrary, Maka Farafenni has lost 100 ha to the Pakala Forest Park.
The establishment of the Pakala Forest Park has affected the access to land in the surrounding villages. Both Chamen and Maka Farafenni have lost agricultural areas and the ability to expand their agricultural areas. In addition, the expansion of the town of Farafenni has been a problem for the surrounding villages, especially Dutabulu (Bobb et al. 1994). Statistics from the Department of Physical Planning show a large area expansion of the town within few years (DPP 1985, 1988). The residents of Dutabulu have traditionally been in a tenure relation to Farafenni. The present village is only thirty years old and the farmers in Dutabulu have borrowed land from the landowners in Farafenni. However, the expansion of Farafenni has made the land more valuable and the farmers have to pay the landowners by cash to secure a field. Thereby, the land cultivated by Dutabulu farmers are decreasing. Less availability of land has mainly affected the women of Dutabulu leading to less access to individual fields. Besides affecting the women's access to individual land, the accelerating land shortage has affected late settlers in both Dutabulu and Maka Farafenni. They have less secure rights to land than the founder families and borrowing land has become more problematic (Bobb et al. 1994).
Figure 6.3: Photo: Interview situation: Villagers of Balingho
Figure 6.4: Photo: Sorgum drying in Fatoto, Upper River division
The use of fallow tends to be decreasing amongst the six villages. Neither Maka Farafenni, nor Dutabulu, nor Kalataba nor Balingo were using fallow any more. In Kubendar the farmers were practising fallow every second or third year. In Tankanto most fields were left fallow for two to four years and then cultivated for one year. Other fields were not left fallow at all.
Table 6.3: Present fallow practise in the Upper Baddibu
In Maka Farafenni, Dutabulu and Kalataba fallow had not been practised since between 20 and 30 years. In Kubendar and Balingo the use of fallow had decreased during the last decade. In Tankanto, no change had occurred due to abundance of land and shortage of labour. Shifting cultivation, however, was common in all the villages, especially between early millet and groundnuts.
In all the villages, the farmers claimed decreasing yields as a major constraint. Several even claimed a halving of the groundnut yields. Less soil fertility and less rainfall were seen as the main causes of this development. Manure was used in all villages to maintain soil fertility, but groundnuts do not respond to manure (Posner and Jallow 1989). Chemical fertiliser has become too expensive and is only used by few farmers.
The choices of crop were also discussed in the questionnaire. The main trend did correspond well with findings in chapter 5. In all villages early millet has become more important now, than ten years ago. Contrary, the importance of groundnuts and rice is declining. Less rainfall was claimed to be a reason for the change. In addition, limited access to the rice fields makes the rice production too labour intensive. However, groundnuts and rice are still important crops at the moment. The Agricultural Extension Service has promoted gardening, but little access to water has limited the success. New varieties of millet, sesamy and groundnuts have been introduced within the last years. However, the new, more drought resistant variety of groundnuts (73-33) has been less successful in NBD than in the divisions upriver (especially M.I.D north, see fig. 1.1) (DoP 1996). The new variety of early millet is more resistant against pests, and therefore promising.
The promotion of technical improvements by the Agricultural Extension Service has resulted in more widely use of the Sin-Hoe-Package-Outright and timely planting in the villages. In Kubendar a new well was drilled. All the villages claimed more use of draft animals than before. In Maka Farafenni, two-thirds of the families own draft animals. The remaining third borrows animals (Bobb et. al. 1994). Borrowing secures all families access to animal traction and manure, but the borrowing one-thirds access is restricted to times when the owners do not require the equipment. These families cannot practise timely planting, and in such climates planting at the right time is vital to secure the yields.
Some goods produced at the farm are sold. Livestock, chickens, vegetables and sesamy are sold at the weekly market (lumo) in Kerawan or in Farafenni. Rice and early millet are consumed at the farm The groundnuts are mainly sold to the co-operative (GCU), to get access to loans and credit. The farmers were afraid to jeopardise the relationship to the co-operative when selling to other traders. Some nuts were, however, sold to Senegalese traders when the Senegalese traders themselves came to the village to buy. Otherwise, the transport costs were too high when selling in Senegal.
The villagers were asked about their thoughts about the general development in farming the last ten years. There was disagreement among the villages and the villagers on this question:
Table 6.4: Views at the general development of the last 10 years
In Balingho the villagers were content with the situation. Their situation had become easier because of better communication (a new road had been built), better co-operation with neighbouring villages and technical improvements. All in all, there had been a positive development. Also in Kubendar, the development was claimed to be positive. In Maka Farafenni the farmers were generally positive, due to the establishment of a new processing mill for sesamy creating more jobs and in turn, less emigration of young men. However, the foundation of the Pakala Forest Park was rather negative for them due to land scarcity. The farmers in Kalataba stated easier work due to technical improvements as positive, but as the farmers in Tankanto and Dutabulu, they saw the general development trends as negative (see table 6.4).
Men and women were asked separately about the major constrains in agriculture within the last ten years:
All in all, except the farmers in Tankanto and Dutabulu, the villagers were rather optimistic for the future in farming.
As concluded in chapter 5, the farm risk has increased during the last decades of ecological, social and economical-political reasons. In the Upper Baddibu case, all these features were recognised as farming constraints. In this sub-chapter, further analysis of constraints and risks of farming in The Gambia will be presented. However, the following discussion will also contain analysis at a divisional level.
According to Torrence (1991), labour shortage is more a problem on the north bank than on the south bank. The labour shortage might be related to the emigration from the north to the urban coastal areas in the south. The traditional response to labour shortage within the agricultural system is hired labour. This response is however more widely used on the south bank than on the north bank. Mechanisation and labour saving techniques tend to be the response to labour shortage on the north bank (Willis 1996, Torrence 1991, and DoP 1996). The result has been increased labour productivity in divisions as NBD and MID-north and even released labour (DeCosse 1992). However, adoption of animal traction among women is still very low throughout the country. Women have less access to credit than men do, and cultural barriers to female use of animal traction are still existing, and are a constraint for further mechanisation. Women constituted for about half the agricultural labour force in 1996 (DoP 1996).
Even if there is labour shortage in many regions of the Gambia, this shortage is related to the busy wet season (Willis 1996). The demand for labour is low in the dry season. The seasonal demand for labour is both a national and regional constraint, which is related to the scarcity of season as discussed in chapter 4. Land scarcity tends to be a marginal problem in the lowland areas, where the labour demand is high. As the women in the Upper Baddibu District claimed, the roads to the swamp rice fields are of poor conditions and too much time is spent walking to the fields. The time is better utilised in upland area in the busy season. The result has been land scarcity in the upland area and expansion into marginal land (ibid.). The output of swamp rice production is also much lower than for upland crops and the mechanisation level in the lowlands is limited (Kargbo 1983, MANR 1995a). As in the case of Tankanto traditionally cultivating swamp rice, emigration and labour scarcity were basic problems, while in Maka Farafenni, cultivating upland only, land scarcity was the main problem.
Land scarcity was also stated as a basic constraint in the land resource study of Upper Baddibu conducted by the USAID (Bobb et al 1994). However, land scarcity in absolute terms is seemingly not a problem at the north bank as a whole. According to table 6.5, cultivated upland as a percent of upland areas suitable for cultivation are exceeded in the southern McCarthy Island Division, while in the Western Division, North Bank Division and Lower River Division only between 46 and 56% of suitable area is cultivated.
Table 6.5: Land use and availability for upland crops. (Source: DeCosse 1992)
However, the numbers of the table are problematic. Suitable upland is estimated by Dunsmore et al. (1976), while average cultivated area is based on NASS reports from 1986 to 91 (DoP). As discussed in chapter 2 and 5, estimates of both available land and cultivated land differ considerably between different sources. Neither suitable upland nor cultivated land can be estimated exact based on these sources. 110 per cent cultivation of suitable upland in the MID south, as claimed by DeCosse (1992) and presented in table 6.5 is not easily understood. When fallow area is added, estimated by DeCosse (Ibid.) to concern 25% of cultivated area, suitable upland is exceeded also in the Upper River Division. There are two ways to deal with this paradox, either has better equipment and new techniques made more land suitable than estimated by Dunsmore et al. (1976), or, the same estimates are made too low. In both cases, the conclusion is rather obvious: Absolute land scarcity is not a satisfying term. More marginal land can be cultivated with better equipment and inputs such as improved ploughs and more use of fertiliser if land pressure increases. Anyway, as concluded in chapter 5, if we add the lowlands, all estimates of land availability (Dunsmore et al. 1976, FAO 1995, Ridder 1991 and DoP 1996), indicates that there are still large reserves of potential agricultural land in each division and The Gambia as a whole.
However, table 6.2 contains valuable information, even if the reliability of the numbers is not satisfying. The low degree of utilised upland in the NBD is an interesting observation. According to Willis (1996) and CILLS (1996), shorter fallow periods are a result of land scarcity and land pressure, especially in the NBD. However, shorter fallow periods are not a usable operationalisation of land scarcity. The divisional differences in the fallow practise are quiet illustrating. Estimates of DeCosse (1992) show that the proportion of upland area left fallow in 1990 only was 16% in North Bank Division and 17% in the Upper River Division, while 49% were left fallow in the western Division.
A question can be raised: Why do the farmers in NBD practise continuous cultivating when large land reserves can be cultivated? Below, the tenure system is analysed in the light of this paradox.
The tenure system of The Gambia has several major constraints. These constraints have become more visible in present due to population growth, as discussed in chapter 5, and urban growth as shown in the case of Upper Baddibu. Remember, as discussed in chapter 4, that social mechanisms of the tenure system help to ensure that new settlers are able to borrow land of founding families with secure ownership rights. When the population density is low, borrowed land would not be reclaimed, and payment of borrowing would not be required. It was no uncertainty borrowing land. However, land pressure due to population growth and urban growth has weakened these social mechanisms.
Illustrating is the case of Dutabulu, were the landowners of Farafenni have reclaimed farming land from the borrowers of Dutabulu, as the town is growing. The access to land in Dutabulu has become more uncertain. In Tuba Mbake, a village near Dankunku in the McCarthy Island Division, there is now chronic land scarcity due to the late settlement of the village added with population growth. Large landholders were reported to lend land on seasonal rather than long term basis. Borrowed land can now be reclaimed any time (IFAD 1992). Even if the problem is most visible in the upland, were land is more scarce, land borrowing tend to be more common in the lowlands probably leading to further uncertain user rights (MANR 1995a).
Other related problems are resource problems related to the unmanaged village commons as pastures, forest and mangrove swamps. These are commons neither controlled by households or linkages, but "belong" vaguely to one or more villages. The result in both cases has been the adaptation of a short-time perspective. Land borrowers with an uncertain future of access to land tend to be more unwilling to practise soil conservation, as e.g. fallowing, than settlers with secure land use rights. People with primarily rights to land tend to make permanent investments in land as e.g. tree planting, while secondary rights to land (borrowing) limits the incentments to do such investments. The uncertainty of the borrowers increases, when the value of land changes as in cases of urban growth e.g. in Farafenni. In the case of unmanaged commons, users are not interested in doing labour or capital investments with a long-term perspective. Ecological degradation might be a result (Freudenberger 1993a).
Another constraint is, in addition to the will to invest in farming, is the possibility of investment in farming. As stated in chapter 5, farm gate prices on groundnuts has increased less than retail prices on chemical fertiliser. In the Upper Baddibu case, many villagers claimed that groundnuts were a less important crop than before. Early millet has become more important, but is domestically consumed and therefore not a source for cash income. However, cash is required for buying inputs as mechanical equipment as seeder and sin hoe package outright and chemical fertiliser. The farmers in Upper Baddibu and the north bank claim cheaper inputs in general or easier access to credits as solutions to this problem (Bobb et al. 1994, Willis 1996, and Torrence 1991). As stated under 6.2.1, farmers on the north bank invest more in mechanical equipment than the farmers on the south bank do. Women have less access to credit, due to restricted ownership rights (IFAD 1992). However, the high level of investments the last decade in farming equipment on the north bank, and The Gambia as a whole, indicates that access to credit or cash not are critical constraints for many male farmers (DeCosse 1992, Willis 1996).
Investments in farming include labour as well as cash. Due to scarcity of season, there is labour scarcity during the wet season. Cultivation of fallowing land, and swamp rice production requires high levels of labour input. According to the farmers of Upper Baddibu, the result has been decreased swamp rice production (Bobb et al. 1994). Swamp rice production requires more labour than upland farming and labour saving techniques are few (IFAD 1992, 1995b). The labour tends to be more effective utilised in the upland, resulting into more reliance on upland cultivation.
However, investments in farming, both in the case of labour and cash, both in the upland and the lowland, has become more risky than before:
"...Increased reliance on equine draft traction observed over recent years, and decreasing returns to agricultural labour relative to non-farm activities, have made the arduous and time consuming job of clearing land left fallow for a number of years physically and economically unviable." (Willis 1996 pp.12)
The paradox of the limited use of fallow practise in NBD might be explained by this observation. It is no longer economically viable leaving an area fallow in region were the reliance on equine draft traction is as high as in the NBD. Both the labour and financial cost of using fallow is greater than the benefits. The main result of this development has been soil degradation, and in turn decreased yields. Added with the increasing fertiliser prices the farmers have to change their responses to minimise the increasing risk (Taal 1989).
As discussed in chapter 4 and 5, modernising the agricultural sector of The Gambia through commercialisation and liberalisation has been the major aim since the introduction of the ERP. The agricultural authorities expected the farmers to respond by increasing the efficiency of their farming activities through more commercial production (USAID 1992). The farmers have responded in quiet different manners. The choice of crop has changed. The production of the commercial groundnut has decreased, while the subsistence crop, early millet, has increased. According to Corbett (1988) and Taal (1989), as discussed in chapter 3, changes in cropping pattern as the shift from groundnut to early millet are an important response to changed farm risk. However, due to the fundamental changes in the Gambian farming system resulting from modernising policies the change of crop is not enough. Cash requirement of the farmers' household is not decreasing and less income from groundnut production makes incomes elsewhere important. Many households have bought farming implements on credit, which have to be repaid and imported rice has become an important source for food even in the farming household. According to the Central Statistics Department (CSD 1994), the rural household incomes have not been matched by an increase in the production of other crops neither in the case of income nor consumption.
When it comes to the Gambian agricultural sector, modernisation is defined by the national agricultural authorities in economically terms. In other words, making suitable for present days means increasing the GDP. As a result, commercialisation of the agricultural sector has been the main aim of the agricultural policies. Increased groundnut production involves increased export incomes. Increased commercial swamp rice production involves increased self-sufficiency degree, and by this, decreased import expenses.
However, doubt can be raised whether the commercialisation of the agricultural sector of The Gambia is a gain to the farming household. Present days needs seen from a farming household involve social sustainabilty, which in fact is more important to the farmers than increasing the GDP. By this, a question can be raised about the importance of the national economy. Both under the regime of a more planned economy, as in the pre-ERP era, and now under the regime of market economy, the farmers are responding in the context of their own household economy, where social sustainability is the main aim. According to Lemarchand (1988), this phenomenon is seen in many African countries independent of the prebendal state claims a Marxist or capitalist orientation. We might call this economic regime, in which the social sustainability is the main power of motivation, for the "local economy". The basis of the local economy is cultural determined perceptions, such as the risk responses, deeply rooted in the Gambian farming (Taal 1989). Further, a clear dividing line has to be drawn between the economy of the divisions or the district and the local economy. Local economy is an economic pattern, based on culture, thoughts and use of resources, within a limited geographical area. Local economy is subsistence economy added with utility maximising. Utility maximising is not profit maximising, but the largest amount of income gains possible without increasing the risk (Lipton 1968). Another basic feature can be added: More incomes than actually demanded to secure the level of social sustainability are seemingly not required.
As we see, the modernisation is defined differently by the agricultural authorities and the farming household. The farming household secure social sustainabilty through multiple modes of earning a livelihood, including both economical and social factors. Thus, the Gambian farmers are in fact "modernising" their agriculture to adapt to present conditions and needs.
Below, multiple modes such as agricultural diversification, income substitution, off-farm income and migration are discussed as modernisation strategies at a farming household level.
The agricultural authorities have introduced efforts aiming at increased agricultural diversification. Agricultural diversification might spread risk both at a national and a household level. At a national level, less reliance on groundnut export might secure export incomes when world market prices are decreasing. Cotton production, which is the second cash crop of the Gambia, has however been minimal for the most of the past decade (The Gambia 1996). The other national aim of agricultural diversification has been food security and better nutrition. Maize is the crop in which the policy has succeeded. The national maize production has increased, and the crop has become more important than traditional crops as upland rice and sorghum (DoP 1996). Horticulture production has been promoted both to improve the nutrient standard of the farmers' diet and as a potential source for export income (DeCosse & Camara 1990, and Beverly 1991). The export of vegetables has increased considerably from 1988 to 1995 (CILLS 1996). However, farmers interviewed in Upper Baddibu claimed that access to water, as every household would need its own well, was a limitation in horticulture production leading to less interest in horticulture. All in all, the policies aiming at agricultural diversification has not improved the national food situation due to rapid population growth. The growth of early millet, maize and horticulture is too small (Akinboade 1994). Rice imports therefore continue to increase.
On farm level, less groundnut production has become the practical importance of agricultural diversification. The farmers cultivate crops with more stabile yields, which are more secure sources of food (Taal 1989). However, crops as early millet are not sold at the market and therefore not a source for cash income. The farmers have to find additional sources for income.
Sale of livestock is one way to secure cash income in the Gambian household economy (ibid.). Livestock is traditionally kept for the security reasons, but has in the present become more important as a source for income (The Gambia 1991b). Farmers we interviewed in Dankunku in MID, told that livestock was sold to invest in education for their children. Some farmers in Upper Baddibu sold livestock to manage large farming investments. However, keeping livestock for security reasons is still more common than keeping livestock as a source for income in the Upper Baddibu.
The women of Upper Baddibu claimed that access to off-farm incomes was important to minimise risk in the farming household (Bobb et al. 1994). Both men and women are taking part in these activities. Petty trading, mainly practised by women, is a less risky alternative to farming, even if trading not necessarily leads to increase of income (Ahmed et al. 1994, The Gambia/UNSO/DANDIA 1984). According to DeCosse (1992) and Dumbuya (1994), the opportunity to work outside agriculture has helped the farmers reducing farm risk with a greater number of income sources:
"Given the new external conditions, farmers keenest interests has become the production of the same amount of food and the release of the family's labourforce to pursue opportunities elsewhere." (Ibid. pp.1)
As discussed under 6.2.1, the labour productivity in the Gambia as whole has increased to mechanisation. In addition the agricultural population has increased. However, the increased production capacity has not resulted in any increased production as discussed in chapter 5. The increased labour productivity has instead, as DeCosse (1992) claims, resulted in release of labour in the villages. The farming household has an improved opportunity to receive income from off-farm work, either through seasonal migration, weekly commuting or overseas emigration (Dumbuya 1994, Ahmed et al 1992, DeCosse 1992, and MANR 1995a). Migration of family members to urban areas is not necessarily a constraint, but a benefit. We have to remember that the household and the family are more important than the individual in rural economy (Taal 1989). Family members, mainly men, are getting jobs in the service sector of urban areas or enterprises in rural areas as petty trading or crafts as blacksmithing and artesian work (DeCosse 1992, The Gambia/UNSO/DANDIA 1984). Cash remittances from overseas or from the urban areas are spent on farming equipment. In addition, more livestock is sold now than before. This is seemingly the explanation of why credit availability not is a real constraint in the farming household as discussed under 6.2.3. The farmers buy more expensive farm implements than before, even if the access to credit has become less and incomes from agricultural activities are increasing.
The migration goes on fuelled by the stories of money earned in urban areas (DeCosse 1992). Illustrating is the estimates of CSD (1994) in the Household Economic Survey Report: In 1983 77% of economically active persons were agricultural or fishery workers. In 1993 the same number was only 64%. There has, in other words, been a rapid move out of agriculture as an occupation during these ten years.
Studies of Gambian agriculture tend to concentrate around traditional statistics as yields and total production (DeCosse 1992). However, questionnaires as the one conducted in the Upper Baddibu for this thesis are necessary to go beyond the "traditional" data and understand the meaning of them. In addition, statistical data on where, how and why the farmers get their off-farm income, could explain more of the processes visible in table 5.1, but updated data on e.g. the degree of off-farm income in rural areas is almost non-existing. The farmers are seemingly not willing to provide data on their income (CSD 1994). Old estimates shows that farm income were 77%, while 23% were non-farm income of total rural income in 1972. In 1989 farm income was 57% and 43% was non-farm income of total rural income (DeCosse 1992). These statistics are confirmed by Jabara (1990), claiming that almost half the rural household income comes from off-farm enterprises. In a survey conducted by the Ministry of Agriculture and Natural Resources (MANR 1995a) 168 out of 366 households were reported to receive income from non-farm activities, while 119 received remittances from absentee relatives. However, the validity of this survey is limited by lack of information of the importance of these incomes compared to agricultural incomes.
Even with limited or old data a trend becomes visible: Getting out of agriculture has become the new risk response in the Gambia (DeCosse 1992). According to Ahmed et al. (1992), farming will be the main source for income in the rural household economy for years, but off-farm income has become more important and will become more important. The farmers tend to invest in the farm only if the food production level of the household can be kept, and the returns become as high as returns of off-farm income (DeCosse 1992). The household farming strategy has changed:
"To farmers, the relevant issue when considered adoption of new technologies is not whether they will generate more returns than their cost, or whether yields per hectare will increase, but whether they will generate more than the farmer could have received elsewhere" (ibid. pp. 16)
As we see, increased groundnut production and increased self-sufficiency degree of food at a national level in The Gambia do not mean meeting present days needs at a farming household level. Instead, responding "positively" on the efforts following these aims might decrease the degree of social sustainabilty at a farming household level. The actual responses to risk tend to be rather dynamic and rapid, meeting present days conditions and needs. The change in fallow practise is illustrative. Decreased use of fallow at the north-bank has become a risk response, because the labour and capital cost of the practise has become too high. Traditionally, as discussed in chapter 4, the fallow practise was itself a risk response due to the vulnerable ecology. The farmers are modernising their agriculture although the commercialisation strategies of the agricultural authorities do fail. Hence, the agricultural sector isn't stagnant, but transforming.