Torgeir Fyhri 1998

The Gambia: 
The Complexity of Modernising the Agricultural Sector in Africa

Chapter 5: The Development of the Agriculture Sector 1983-96

Contents

CHAPTER 5: The Development of the Agriculture Sector 1983-96
5.1 Production alterations
5.2 The relevance of the bio-system
   5.2.1 Precipitation
   5.2.2 Soil productivity
   5.2.3 Productive land
5.3 Labour availability
5.4 Technical improvements
5.5 Agricultural policies
   5.5.1 Price policies
   5.5.2 Fertiliser polices
   5.5.3 Credit policies
   5.5.4 Food policies
5.6 The relevance of the farming household to the farming system
   5.6.1 Farming practices
   5.6.2 Social organisation

OTHER CHAPTERS
Contents
CHAPTER 1: Introduction
CHAPTER 2: Methodical Considerations
CHAPTER 3: Theoretical Framework
CHAPTER 4: Introducing The Gambian Agricultural Sector
CHAPTER 5: The Development of the Agriculture Sector 1983-96
CHAPTER 6: Responses to household Constraints and Farm Risk
CHAPTER 7: Testing of hypotheses and Theoretical Discussion
CHAPTER 8: Concluding Remarks
REFERENCES
APPENDIX - Interviews conducted in villages around Farafenni

5 THE DEVELOPMENT OF THE CROP SUB-SECTOR 1983-96.
This chapter sought to point out eventual alterations together with possible causal relations to the national agricultural production of The Gambia 1983-96. As discussed in the former chapter, concrete changes in agricultural production were stated as aims of the modernisation strategies. Therefore, the national agricultural production is a significant measure of the degree of success of the modernisation strategies. Have agricultural authorities reached their aims and what are the main causes to the eventually failure or success?
To look more closely at the whole questions of causes, a causal model discussed in chapter 2, presented below, will be used as a guide to the analysis. The main possible variables influencing agricultural development are integrated in the model.


Figure 5.1: A model for the evolution of the agricultural production in The Gambia

Besides explaining eventually production alterations, the variables will be discussed in relation to each other. It is however; important to point out that not all of the variables of the model will be discussed that thoroughly in the analysis. The relevance of the model is rather to visualise the complexity of developing the agricultural sector.

However, according to the problem of this thesis, the household as an actor in the system is emphasised in the analysis. Further, both the model and the analysis aim at pointing out the spatial dialectics of the agricultural development process.

According to the model above, figure 5.1, the eventually alterations in production are related to changes in production, area, and area productivity of the upland cereal crop early millet, the upland cash crop groundnut and swamp rice. Data on these alterations are presented in sub-chapter 5.1. The rest of the chapter is consecrated analysis of the different variables, and their relevance to the production and the system as a whole.

Firstly, sub-chapter 5.2 discusses the relevance of the bio-system to the farming system. Secondly, sub-chapter 5.3 analyses the importance of labour availability also in relation to productive land. Thirdly, the importance of eventually, technical improvements will be analysed in sub-chapter 5.4. Fourthly, sub-chapter 5.5 discusses the relevance of the agricultural policy system. This sub-chapter discusses implicit the relation to the economic system and the demand system and in turn their gathered relevance to the production alterations. Finally, sub-chapter 5.6 discusses the farming household as an actor in the entire system.
The analyses are mainly based on quantitative data collected by the National Agricultural Sample Surveys (NASS), The government of the Gambia, USAID and FAO.

However, both the data and the discussion concentrate on the development of the crop-sub-sector mainly at the national level. One of the variables of the farming household system, responses, will be analysed in relation to more qualitative data in a discussion at a lower geographical scale in the next chapter.

5.1 Production alterations
During the studied period, early millet, groundnuts and swamp rice are the three main crops cultivated in The Gambia, constituting in all for 83.6 % of the total area cultivated in 1983 and 81.6 % of the total area cultivated 1996 (DoP 1995a, 1996). Below, in the tables 5.1a,b and c, production alterations of these three crops are presented. The figures 5.2a, b, c, 5.3 and 5.4 supplies the tables where data back to 1974/75 are added. The numbers are collected from National Agricultural Sample Surveys (NASS).

Table 5.1a: Early millet production in The Gambia 1983/84-1994/96, Cultivated and Harvested area are measured in `000 Ha. Yield are measured in kg./Ha. Total production are measured in `000 tones. (Source: DoP 1995a, 1995b & 1996)

Table 5.1b: Groundnut production in The Gambia 1983/84-1994/96, Cultivated and Harvested area are measured in `000 Ha. Yield are measured in kg./Ha. Total production are measured in `000 tonnes. (Source: DoP 1995a, 1995b & 1996)

Table 5.1c: Swamp rice production in The Gambia 1983/84-1994/96, Cultivated and Harvested area are measured in `000 Ha. Yield are measured in kg./Ha. Total production are measured in `000 tonnes. (Source: DoP 1995a, 1995b & 1996)

Table 5.2: Production of all crops in The Gambia 1983/84-1994/96, Cultivated and Harvested area are measured in `000 Ha. In % means harvested area as a percentage of cultivated area. Total production are measured in `000 tons. Note: Totals may not add up due to rounding. (Source: DoP 1995a, 1995b & 1996)


As the tables and the figures show, the great annual fluctuations in cultivated area, harvested area, and yield per hectare and total production for each of the three crops tend to be rather conspicuous. In 1989/90 yields of groundnut was 1509 kg per ha, the following year the yields were almost halved to 883 kg per ha. In both years, over 96% of cultivated area were harvested, which is a high degree compared to 83.4% in 83/84 and below 90% in 92/93 and 94/95. For all crops, cultivated area fluctuates dramatically during the study period. In the 1988/89 season 198 740 ha was cultivated for crops while the same number in the 1993/94 season were 157 100. A result, there are unpredictable great annual variations in total production of all crops.

In 1992/93 both low groundnut yields (849 kg/ha) and a small cultivated area (158 570 ha) led to low total production of all crops (136 570 tons). In 1989/90 a rather average cultivated area (180 420 ha) and a positive extreme in yield per ha for groundnuts (1509) made the total production of all crops 230 080 tons. As further discussions of this chapter indicate, these fluctuations are not easily explained. However, the fluctuations might be explanations themselves. Fluctuations in yield per ha might be a source of risk at a farming household level. By this, as discussed in chapter 4, the farming household sought to minimise the risk through their social organisation such as the tenure system or multiple modes, or by adapting their farming practices. Hence, the changes within the social organisation or farming practices do influence either agricultural production directly or through the variables visualised in figure 5.1. In addition, as the director of the Department of Planning (DoP) K. B. Johm (1992) claims, the unpredictable fluctuations of all the measured parameters is problematic when planning agriculture and forming agricultural policies.

The most visible production trend during the period is the changed importance of the different crops within the farming system. Increased groundnut and swamp rice production has been the main aims of Gambian modernisation strategies. However, early millet has become more important while groundnut and swamp rice production has decreased. The Country Position Paper for the World Food Summit (The Gambia 1996a) and Bojang (1996), claim soil degradation through intensive farming and monocropping, farming practices, to be the main causes of these trends. The agricultural policies had led to more intensive farming, increased mechanisation and shorter fallow periods. In the swamp area, salinisation of the soil had limited arable land (The Gambia 1996a). However, the yields of groundnuts and swamp rice have not necessarily decreased. In the case of groundnuts it is cultivated and harvested area, which has decreased and in turn led to declining total production of this crop.

When we know that there has been a rapid population growth during the period, and as a result increased demand for food, other explanations than soil degradation have to be added. Producer prices, and farming input prices have changed during the period due to the liberalisation, resulted by the ERP. Subsidies on fertiliser were removed and output prices on groundnuts have fallen.

Figure 5.3 indicates another interesting trend. The year to year fluctuations of early millet is lesser than that of swamp rice and groundnuts. If we divide the lowest yield per ha with the highest of the period we come out with a coefficient were zeros is lowest and 1 is the highest. For early millet this coefficient is 0.74, for groundnuts 0.56 and for swamp rice 0.57. We must assume that production at a household level is easier to plan when the yields are stabile. Producing early millet tends to be less vulnerable against varying ecological fluctuations such as precipitation. Since early millet is a domestic consumed crop, it is also less vulnerable to price fluctuations. From a farming household perspective, producing the subsistence crop, early millet, therefore, contains less risk than groundnuts.


5.2 The relevance of the bio-system
Both anthropogenic and climatic factors affect the ecological features of farming (SWMU/DAS 1994). Illustrating is declining soil fertility. Given the vulnerable soils of The Gambia, discussed in chapter 4, more intensive cultivation strategies as e.g. monocropping leads to declining soil fertility if chemical fertiliser not does replace the nutrient loss. When fertiliser prices are increasing, farmers might be reluctant to buy it and soil fertility declines. By this, the variables soil fertility and productive land within the bio-system, have to be analysed in relation to other variables such as farming practices and labour availability to state the explanatory value of them to the production alterations.

5.2.1 Precipitation
Precipitation fluctuations are assumingly not human made. However, the significance of rainfall fluctuations and how these fluctuations affect the crop sub-sector is anthropogenic (Mortimore 1989). According to figure 5.5, there is large year to year average rainfall fluctuations in the study period. The average Gambian rainfall was at its lowest in 1983, below 600 mm/annum. The rainfall was highest in 1994 with ca 1000 mm/annum. No direct linkage between yield per hectare and yearly rainfall can, however, be established for any single crop. Groundnut, early millet and swamp rice are not following the same trends in yield per hectare production, which in turn do not follow the yearly rainfall. The lowest yield per hectare of groundnuts was harvested in the 1990/91 and the 1992/93 season, while the 1989/90 was the opposite extreme. According to figure 5.5, the rainfall of the 1989/90 was relative high, but only fairly higher than in the 1992/93 season. Swamp rice production had its very extreme yield per hectare production in 1983/84 when the rainfall was the lowest of the whole period. The greatest part of the water requirement of Swamp Rice is supplied by tidal water, but the nurseries are located in the upland relied upon rains (Alimi 1990). Early millet does not have the extremes in yield per hectare production. It is also more drought resistant than groundnuts (Peters and Schulte 1990).

Figure 4.6 presents seasonal and not yearly average rainfall. Even more important than yearly rainfall are the seasonal and the timing of the rainy season (Alimi 1990). In 1985/86, late rains reduced the groundnut crop. In the 88/89 season the groundnut harvest failed due to uneven distribution of rainfall. Total rainfall was, however, high (Jabara 1990). In the 1996/97 season precipitation failure early in the planting season led to death of groundnut plants in the Farafenni area (Samba Bah: Agricultural Extension Worker, Upper Baddibu district: pers. commun.). Groundnut is relatively drought resistant; water requirement of the groundnut is between 300-600 mm, and dependent on being planted at the right time. Right time means usually first decade of July when average decade rainfall is greater than 50 mm. Delay beyond 15th of July would mean little chance of attaining physical maturity and probably failure of the groundnut harvest (Alimi 1990). However, lack of adequate seasonal rainfall data in required time series make correlation with cultivated area and yield per ha output problematic.

As discussed in chapter 4, and showed in figure 4.6, rainfall in the Gambia follows five years fluctuations. No clearly trend of declining precipitation are visible in figure 5.4 and 4.6. However, figure 4.7 shows a remarkable declining trend the last thirty years:


"Average rainfall levels in The Gambia have followed a long-term declining trend with a high degree of variability over the latter part of this century, with considerably associated implications for field cropping systems." (Willis 1996).

According to Willis (1996), this downward rainfall trend has contributed to the process of salinisation in the mangrove swamps and the banto faros. As discussed in chapter 4, the saline water of the ocean intrudes upstream during the dry season until an equilibrium is reached. In the dry season this interface is 200-km inland while 80 km in wet season (SWMU/DAS 1994). The interface moves inland in drought years and can vary from 120 km in wet season to 250 km in dry season. The movement of the interface affects irrigated rice production. The intrusion of salt water makes the soil of former rice fields in the lowland saline. Intrusion of salt water in early cultivation season kills the rice crop before maturing. In addition, lack of fresh water is an obstacle for effective desalinisation. Approximately 30 000 ha are affected of salinisation (ibid.).

In the upland area, the long-term decrease in rainfall tends to affect the growing season and the choice of crop (USAID 1992). The result is a strengthening of the scarcity of season. Shorter growing season creates a labour bottleneck, which in turn affects the choice of crop (Willis 1996).

Short duration crops as early millet tends to be more preferred now than before. In addition, the shorter growing season has affected the social sustainability due to increased farm risk. The fluctuations of yield per hectare groundnut production has become larger and therefore more risky to produce. However, rainfall fluctuations cannot alone explain the alterations in crop production, but the long-term downward trend might have a devastating impact on the environment (Gambia 1987a). The result is less tolerance for crop production. This decreasing tolerance might explain some the large annual fluctuations presented in figure 5.3 and 5.4. When the trend is decrease in total rainfall in an ecological zone with already little tolerance of rainfall, local day to day rainfall variations not visible in the statistics could have a devastating impact not only on the environment, but also on crop production. However, these suppositions cannot be proved by statistics, especially not at a national level.

5.2.2 Soil productivity
It is an agreement among farmers (field-survey 1996), agricultural authorities (Gambia 1996a) and scientists as Maalenku (1993a), SWMU/DAS (1994), Ridder (1991), Bojang (1996), USAID/OMVG (1985), Willis (1996) and Trolldalen (1994) that land degradation is a major obstacle for both present and future cultivation in The Gambia. There is also an agreement that soil erosion; dismissing soil fertility and salinisation are the main kinds of land degradation affecting farming.

However, it tends to be a disagreement between different scientist when it comes to causes of land degradation. The Government of The Gambia (Gambia 1996a) claims that land degradation is caused by lack on emphasis of conservation of traditional agriculture practises. Contrary, Willis (1996) and SWMU/DAS (1994), claim that the land degradation is a result of agricultural intensification and adaptation of short-term perspective among farmers. We can give either of them neither wrong nor right. Both the bush fallow practise and agricultural intensification causes land degradation.

Increased deforestation has occurred due to agricultural intensification and increased population pressure in The Gambia (ibid.). According to Ridder (1991), 42 000 ha forest were removed between 1980 and 1987. Averagely, 6000 ha were removed yearly. Annual bush fires are estimated to burn 80 % of the forest cover (SWMU/DAS 1994), including both virgin forest and bush fallow areas. Bushfires break out both accidentally or purposely as a part of the fallow practise. Setting fire on the bush in the dry season ease the opportunity to cultivate an area after a fallow period. However, when vegetation is removed through bush fires, nitrogen, sulphur and phosphorus escape into the atmosphere (Maalenku 1993a). In addition, the soil becomes exposed to water and wind, which are transporting particles away from the fields.

Soil erosion by water is claimed by SWMU/DAS (1994) to be the most serious type of land degradation in the country. It has been estimated that continuously cropped fields with slopes of only 2% have soil losses greater than 12,5 tons/ha/year. Soil productivity is declining where fertile soil particles are lost. Siltation where infertile sediments are deposited on exist crop land down slope is another type of land degradation caused by soil erosion.

Soil conservation and fertility maintenance on the uplands of The Gambia has been achieved through the fallow practise and shifting cultivation (see chapter 4) allowed farmlands to recover their facility and protected these against excessive erosion (Bojang 1996). This practise was possible due to low population density and the relative abundance of virgin forestland and effective as long as the majority of the surrounding lands remained under vegetation cover (ibid.).

Rapid population growth and promotion of advanced technologies have also encouraged abandonment of shifting cultivation and the fallow practise. The result has been an expansion of cultivation into marginal land by removing forest and range vegetation. Marginal soils soon lose their fertility because of leaching nutrients, loss of organic matter and soil erosion (SWMU/DAS 1994, Willis 1996).

In the lowlands, salinisation has been the main cause to land degradation. In the former sub-chapter, salinisation caused by a long-term downward rainfall trend was discussed, but salinisation is also man made. In irrigated rice fields water from the river is pumped up in the fields levelled higher than the river, when the water evaporates, salt is left in the soil (The Gambia 1996a).

However, with the basis in agricultural statistics and area data, estimating the impact of land degradation and thereby on crop productions is not a simple task. Using yield per hectare production as an estimate has several limitations. No trend in yield per hectare production (see figure 5.3) can illuminate the impact of soil degradation to crop production in the period. In addition, fertiliser use and different farming strategies affects area productivity. Torrence (1991), The Government of The Gambia (Gambia 1996a), and the farmers in NBD (Interviews, field study 1996), claim the gravity of land degradation as a constraint to crop production in recent years. The long-term downward rainfall trend has decreased the tolerance of the soil and made it more vulnerable to rainfall variations.

5.2.3 Productive land
USAID (1992) claims that pressure on arable land caused by population growth, and decreased soil productivity has resulted in a shortage of land in rural areas. The shortage is operationalised by USAID (ibid.) as decreased use of fallow. According to FAO & MoA, no such shortage exists at a national level (FAO & MoA 1985, FAO 1995). The cultivable land of The Gambia is estimated by FAO to 500 000 Ha. (ibid.). The annual and permanent cropland is measured to 300 000 Ha. Thus, the cultivable land reserve is 200 000 Ha. There has been no alteration from 1985 to 1994 in these conditions. However, the statistics of FAO tend to have limitations. Annual and permanent cropland is not defined. When we know average yearly cultivated area for all crops is below 200 000 Ha. for the period between 1979 and 1996 (see table 5.4), the annual and permanent cropland of FAO tend to be a rather strange extent. The Land Resource Study of 1976 is, according to Johm (1992), the most authoritative source of information of land resources in The Gambia.

According to the Land Resource Study (Dunsmore et. al 1976) and DoP (1996), total arable land in The Gambia is over 550 000 ha. Between 1976 and 1996 the utilisation degree has been between 30 and 35% per year. However, most of the land reserves are placed in the lowlands, while most of the crops are produced in the upland. According to DeCosse (1992), 77% of the upland area is utilised for farming, and there might be some land scarcity in some regions. In the peri-urban agricultural areas of Kanifing land pressure due to rural-urban migration has forced the population to abandon the practise of shifting cultivation and resort to more intensive use of farmland. The result has been, according to NPC (1991) loss of soil fertility and fall in area productivity, which in turn limits total arable land at a lower geographical scale.
We should expect that absolute land scarcity should influence the parameter, cultivated land, in the farming system. By this, according to figure 5.4, harvested area as proportion of cultivated area of all crops tend to have increased during the study period indicating pressure on arable land. However, the total cultivated land of The Gambia tends to be stagnant or stabile during the period studied. In addition, no clear downward trend in yield per Ha production indicate any dismissing soil fertility caused by expansion into marginal land as a result of land scarcity.

5.3 Labour availability
Although the political, economical and physical climate for crop production have changed within the study period, the cultivated area and total production of all crops did tend to be rather stagnant (FAO 1995, DoP 1996). This is a rather interesting point. Causes to this trend are found in the farmers' responses to both ecological variables and agricultural policies, but also in changing land and labour productivity. Table 4.3 indicated an absolute growth in the agricultural workforce. However, due to the increased labour force, why has not totalled production of all crops increased to a larger extent? We should expect that an increased demand and increased agricultural workforce would lead to increased production?

Weil claimed in 1986, that the stock of workers in rural areas had decreased due to rapid urbanisation. Jagne (1996) claims that the displacement of rural population to urban areas will have a general effect on the economy of the Gambia as a whole, due to importance of the agricultural sector. Numbers from the CSD (1994a) shows that while the urban population constituted for 30% of total population in 1983, the number had increased to 37% in 1993 (see table 4.1 and 4.2). However, internal migration has not affected the total agricultural population at this moment. Since 1986, there has been a rapid population growth both in rural and urban areas. In 1985 the total agricultural population was 615 000 with a total agricultural labour force of 290 000. By contrast, in 1994 the total agricultural population was 861 000 with a total agricultural labour force of 397 000 (see table 4.3 and 4.4). Except from some rural regions, there is no decrease in the agricultural population or the agricultural labour force in percent of the total population in The Gambia (FAO 1995). This means that both the total agricultural workforce and population have increased during the study period. A reasonable consequence would be a more labour intensive agricultural production with increased cultivated area and yields. However, the trends in cultivated area as shown in figure 5.4 paints quiet a different picture.

The reason might be the labour bottleneck in the short rainy season. The increased agricultural workforce cannot be fully utilised when the already cultivated area requires more labour than before in the rainy season, due to the long-term downward rainfall trend discussed in sub-chapter 5.2.

It is, however, important to be aware of the changes in the composition of the rural population due the migration process. Mainly men and women in their most productive age are migrating to urban areas (CSD 1994a, Jagne 1996). The result is that: i) The average age of the agricultural population is higher than the urban population. ii) The agricultural workforce contains a greater share of elders than the total population of the Gambia (CSD 1994a).

5.4 Technical improvements
Animal traction has been encouraged to overcome the labour bottleneck created by less rainfall and a shorter production cycle (Willis 1996). In addition, farm mechanisation was seen as the key to sustained rural development, making crop production and diversification more efficient (The Gambia 1987a). For the whole study period, 90 % of the upland area is farmed by using animal traction in The Gambia (Gambia 1996a, DoP 1990, 1996).

Table 5.3: Percent area farmed using animal traction, by type of animal and crop 1989/90 and 1995/96. (Source: DoP 1990, 1996)

More farming activities tend to be managed by animal traction in present days than in the beginning of the period. Table 5.3 illustrates this point (DoP 1996). The degree of utilisation has, however, not been satisfying. The promotion of lighter draft equipment for horses and donkeys have simplified both the cultivation and harvesting process, especially for groundnuts. The super Eco seeder is a seed drill pulled by horses and donkeys. (The seeds are placed in a seed box. In the bottom of the box, a seed plate with holes rotates due to the rotation of the wheels of the seeder and the seeds fall through a tube into the rows. Changeable seed plates with different numbers of holes of different sizes make the seeder useable for all kinds of crops). A sin hoe package outright is a combined plough/harrow, weeder and groundnut lifter with a basic construction and changeable special equipment. Both the sin hoe package and the seeder are used in the upland area and mainly in the groundnut fields (Diallo 1991).

As table 5.3 indicates, mechanisation of the groundnut fields has increased more than coarse grain fields. Farming with this kind of improved technical improvement has, however, certain drawbacks. The use of the seeder requires well-ploughed and harrowed plots. More labour contribution is needed preparing the plot in a busy period, which in turn affects fallowing. As a result the use of fallow becomes even more seldom, due to the extra labour contribution (DeCosse 1992, Bojang 1996).

A greater share of male labour in swamp rice production and male migration to urban areas have forced the women into groundnut production and in turn into the use of draft animals in some regions of The Gambia. However, mainly men invest in this kind of equipment. The equipment is expensive and men have easier access to credit than women do. The traditional household ownership pattern discriminate women who have no right to own mechanised equipment. Mechanisation of women's rice fields has been ignored and since the rice is produced for own consumption, the return of investments in animal traction may be low (Diallo 1991).
Land fragmentation and small size holdings; related to the tenure system is another of the main constraints for utilisation of technical improvements. The agricultural input sector programme (The Gambia 1987a) claims that insufficient cash incomes related to small-size holdings are the main obstacle for investments in machines and animals. Higher incomes could be realised through cultivation of large areas.

Although the uses of animal traction are increasing and farming equipment are improved, the potential of the technical improvements is not yet fully utilised.

5.5 Agricultural policies
Agricultural policies are here defined as efforts aiming at modernising the agricultural sector. As modernising is very close related to the economic system, including export incomes and import cost and the demand system, including demand for food and socio-economic development, these systems will be discussed together with the agricultural policies in the analysis of thesis.

Large annual variations in national agricultural production illustrate the problems of planning agricultural activities in The Gambia (Johm 1992a). Rainfall cannot be planned neither can world market prices on crops. Economic development in The Gambia depends on both variables. Figure 5.6 illustrate this point. According to the figure, the value of exported groundnut products does not correlate with the quantity exported. This finding indicate the relevance of the world market prices on groundnuts to the national economy in The Gambia. The figure also shows the low correlation between total production and exported quantity of groundnuts. Variations in domestic consumption might be the explanation of the latter phenomena.

As we shall see further in this chapter other variables, as the farmer's responses, come into play and make planning even more complicated. Increasing groundnut and swamp rice production through commercialisation and liberalisation of the agricultural sector has been the major aim by Gambian agricultural authorities. The introduction of the ERP was the main effort in reaching this aim aiming at increased groundnut and swamp rice production. However, figure 5.2b and c indicate that both total groundnut and swamp rice production, the crops in which the agricultural policies have put stake on, are declining.

A report presented by the government of The Gambia (Gambia 1987b) indicates that it was much uncertainty of the practical ends of privatisation of the parastatal agricultural institutions. According to DeCosse (1992), the main result of the privatisation policies has been that the development of the crop sub-sector in The Gambia has been disappointing for most of the past decade.

"The impact of macro-economic policy reforms including privatisation, credit and interest rates of liberalisation, and removal of subsidies were found to impact negatively on the agricultural sector resulting in low productivity thereby aggravating food insecurity." (Gambia 1996a pp. viii).

Liberalisation has been the main effort in modernising strategies during the study period. The era of five-year plans, and governmental encroaches upon the market forces were over. However, even if the new strategy was necessary it was not successful. An interesting point is that in a report presented by The Gambia in 1987 (1987a), liberalisation policies during the ERP were seen as the life belt of both food and cash crop production. In a report presented by the same authors in 1996, the liberalisation polices were blamed for the failure of both food and cash crop production (The Gambia 1996a). Below, the efforts of the liberalisation process, the agricultural policies, will be analysed to seek explanations of this phenomenon.

5.5.1 Price policies
As the ecological variables, it tends to be difficult to link crop prices with the concrete alterations in crop production. Coarse grains as early millet is usually domestically consumed, and prices are therefore of less importance. The same tends to concern rice, but the price of imported rice is important. However, in the case of groundnuts, both farm gate price and the border price are of large concern. The International Fund for Agricultural Development (IFAD 1995b) claims that marketing problems resulting from privatisation of the GPMB is as important as reduction in rainfall to the decline in groundnut production. However, way of payment either cash or credit might be of importance as well (interviews, field study 1996). The farmers prefer cash payment for their groundnuts even if prices are higher when paid by credit. It might e.g. be an obstacle for Gambian farmers selling groundnuts to Senegalese buyers when they pay by credit.

Raising through producer price subsidy on groundnuts was the ERP strategy in increasing the production in the Gambia. The price could not deviate much from the Senegalese prices, which were subsidised. If so, Gambian farmers would probably sell groundnuts to Senegalese buyers. Contrary, in according to Lemarchand (1988), more than half of Senegal's peanut harvest in 1985, was smuggled abroad, mainly into The Gambia. Groundnut trade through informal channels is an important constrains to the economy in both countries. However, transport costs and lack of infrastructure limits the importance of price differences between the two countries.

Price subsidies contain additional problems. The GPMB, which had the responsibility of buying the nuts, bring upon large expenses. The pre-ERP price policies had already added a debt burden to the GPMB. Two years after the introduction of the ERP the aim of increasing producer prices on groundnuts was given up. In addition, they did not manage to keep the Gambian prices higher than the Senegalese until France devalued the CFA in 1994 (see figure 5.7) (Carr & Sarr 1994).

Already in the 1985/86 season, when the ERP was introduced, questions were raised whether the economical foundation of the GPMB and the national economy were too weak to accomplish the aim of increasing the production prices on groundnuts (Akinboade 1994, Jabara 1990). World market prices had a sharp decline and the producer prices were subsidised by 46.3% by the GPMB. A flexible exchange rate was implemented the same season by floating the dalasi. As a result the value of the dalasi decreased. The following season, groundnut prices raised 64% due to increased world market prices, but the price subsidy constituted 42.2 %. From the 1987/88 season, Gambian groundnut prices were based on prevail world prices and reduced from D1800 to D1500 per ton. To encourage competition, obstacles to private groundnut traders were removed the same season. The following season, 1988/89, producer prices were reduced to D1100 per ton. The GPMB realised a 13.2 mill. Dalasi loss on groundnut operations (Jabara 1990). Since that season, actual groundnut prices have steadily increased until stabilisation or stagnation at D2000 in 1992/93 to present.

The groundnut price has fluctuated during the whole study period. Even if we should expect the groundnut prices to affect the farmers' will to cultivate groundnuts, there is seemingly no correlation between the price fluctuations and the annual fluctuations in cultivated area shown in figure 5.4. The reason is that the prices both in The Gambia and Senegal are fixed at the end of the groundnut harvest season. The farmers do not know the price when cultivating. However, price fluctuations affect risk. Large price fluctuations increase risk of producing groundnuts.


5.5.2 Fertiliser polices
Fertiliser expansion and increased usage has been a goal in increasing agricultural output in The Gambia (Gambia 1987a). However, the strategies of expanding fertiliser use differ considerably between the pre- and post ERP era. While subsidies on fertiliser ranged from 6 to 80 % in the pre-ERP era, the removal of subsidies has been an outstanding strategy in the post-ERP era (DoP 1995a, Nagarajan et. al. 1993).
The fertiliser distribution was liberalised in January 1986. An outstanding aim was to eliminate the distribution monopoly of the GCU (The Gambia 1987a). In the pre-ERP era, GPMB was exclusively importing and wholesaling fertiliser in The Gambia. Retailing was primarily carried out by GCU (Nagarajan et. al. 1993). 

The reasons for liberalising the fertiliser sub-sector were in the first place to make the distribution more self-sustained. Market forces of supply and demand should reduce costs of fertiliser to farmers through efficiency. Reducing the role of the government was another aim. Liberal credits and heavy subsidies on fertiliser during the monopoly state of the pre-ERP had led to a rapid consumption increase from 2877 tons in 1973 to 12135 tons in 1981 (Gambia 1987a). However, this strategy had not managed to make the fertiliser sub-sector self-sustained. Mismanagement and losses in its groundnut trade added a heavy debt burden to the GPMB. In addition, lack of adequate storing facilities and functioning transport infrastructure of fertiliser gave financial loss. Economic support through subsides from the government were required to sustain the GPMB`s operations (Gambia 1987a, Nagarajan et. al. 1993). Due to economic problems and ERP, GPMB became privatised in 1988 (Woodward 1988).

However, the liberalisation of the fertiliser sub-sector has not met the desirable end. Fertiliser importation and use has declined in the post ERP era. As shown in table 5.4 (DoP 1990, 1996), per cent area cultivated with fertiliser has been dramatically reduced within six years. Figure 5.9 and 5.10 support these findings. Both retail and official prices have increased considerably, because subsidies have been removed. In turn, fertiliser import is reduced due to decreasing demand. However, lack of data over actual use of fertiliser is not excisting. Import data does not account for year to year storage and might have limited value in this analysis.

Table 5.4: Fertiliser use in per cent of total area by crops 1989/90 and 1995/96. (Source: DoP 1990 and 1996) *The numbers are estimates of respectively upland and swamp rice.


If we now compare these statistics with the production trends in figure 5.3 an interesting point becomes visible. Yield per hectare of early millet, groundnuts and swamp rice is not declining significantly, even if fertiliser use is dramatically reduced during the study period. Overuse of fertiliser during the pre-ERP era is claimed by Nagarajan et al. (1993) as a reason for the lacking decline in yields. The effect of fertiliser has not been desirable due to low yield response. A more reasonable explanation might be leakage through re-export of fertiliser through informal channels to Senegal. Prices were far higher in Senegal, due to lack of subsidies (ibid.). According to Akinboade (1994), only between 27 to 36% of the sold fertiliser was used in The Gambia in the years before the subsidies were removed. Following the fluctuations in world market prices added with devaluation in Senegal makes this re-export less profitable (Carr & Sarr 1994). Therefore, fertiliser use has probably not decreased as much as figure 5.10 indicates. However, more expensive fertiliser has affected groundnut production. Output groundnut prices have increased relatively less than input prices of e.g. fertiliser. According to Posner and Jallow (1987), groundnuts require chemical fertiliser for optimal production. A direct linkage between groundnut yields and fertiliser use is not visible in national agricultural statistics. However, e.g. farmers interviewed in the Upper Baddibu claimed that the groundnut yields on their plots have decreased significantly as a result of increasing fertiliser prices (field study 1996). Increased fertiliser prices added with relative slower growth in groundnut prices has made groundnut production less economically significant (Gambia 1996a). Hence as shown in table 5.1a and figure 5.2a, early millet production raised significantly from 1984/85 to 1985/86. This finding indicate that the reliance on early millet increased as the reliance on groundnuts decreased due to increased fertiliser prices.

5.5.3 Credit policies
Both fertiliser and price polices were important efforts in the commercialisation process of the crop sub-sector. These polices cannot, however, be seen independently from the credit policies. The farmers usually buy farm inputs as fertiliser and food on credit as these have to be bought in another season than they get cash income from sales of agriculture products. The credit repayment possibility depends on the groundnut prices. Increasing fertiliser prices lead to more risk in credit operations caused by the increased dependence on the groundnut harvest and producer prices. Harvest failure caused by e.g. late rains added with low producer prices on groundnuts make credit repayments more risky.

Credit repayment deficit caused GCU a heavy debt burden due to outstanding loans. By 1985, the GCU owed D35.2 millions due to poor repayment rates among farmers. ERP credit reforms focused on improving the financial viability of the GCU. In 1987/88 new criteria for earning credit were implemented. The main objective was that private traders should take up the credit business (Johm 1988). Loan defaulters were excluded from new loans. Only co-operative members who get credit in the previous year were eligible for new credit. The result was poorer access to credit for poorer rural household (Jabara 1990). Women tend to be the main losers, producing food for own consumption with lesser opportunity to gain cash income. Already before the reform of the credit system, women received only 20% of the GCU credit funds (Johm 1988). 

Better-off households, that paid the inputs in cash anyway, increased their access to credit (Jabara 1990). The new credit component was not a success, due to less access among farmers. In addition, private traders were less interested than expected in taking up the credit business due to the inherent risk associated with this activity (Johm 1988). According to Akinboade (1994), lesser use of credit led to decrease in agriculture investments among farmers.

Due to failure of both groundnut operations and fertiliser marketing, the GPMB experienced a financial crisis during the eighties. The main objectives of the GPMB, maximisation of producer prices and foreign exchange return to the economy tended to be incompatible (Woodward et. al. 1988). The GPMB became privatised in 1988. However, the economic environment was not favourable for privatisation. Uncertainty among private actors about governmental interventions limited the success of the privatisation both in the fertiliser and groundnut trade (Nagarajan et al. 1993). In the beginning of the 90s foreign investors, mainly Swiss, bought the majority of the GPMB and created The Gambian Groundnut Company (GGC). In the present, groundnut and fertiliser prices are set by the GCU (which owns 10% of GGC) the government, and the GGC.

The parastatal agricultural institutions have become more economically self-sustained as a result of the privatisation process. However, a harder economical climate for the farmers has developed. Government assistance and protection of the farmers are getting weaker with less access to credit, low crop prices and high fertiliser prices. Groundnut production as the main commercial crop depends very much on these three variables. As a result, groundnut production involves more risk than before. Subsistence related early millet production, depends less on these variables. Early millet is seldom sold on the market and early millet cultivation depends less on fertiliser. Remember the decreasing trend of groundnut production and the opposite trend in early millet production. We might say that increased liberalisation led to a downturn in commercial crop production and increased interest in subsistence crop production.


5.5.4 Food policies
As discussed in the former chapter, population growth has led to increasing demand for food. National food production has been fluctuating (see figure 5.2a and c) while consummation has been on continuous increase due to population growth CILLS 1996). However, in a climate of open economies food deficits are theoretical quickly covered by supplies from other economies, so also in The Gambia (Johm 1992). The liberalisation of the agricultural policies has favoured rice imports, as the import restrictions were lifted and retail prices were decontrolled (The Gambia 1996a, Kinteh 1988). Not surprisingly, it is financial problems related to the growth in food imports that led to focus on improving food production in the Gambia. The expenses of rice imports increased sharply (Johm 1996b). In addition, wheat and flour import has increased remarkably. Wheat and flour is used in production of French breads consumed by tourists and urban population. Wheat and flour are not domestic produced and have to be imported.

Table: 5.5: Rice and Wheat & Flour import 1983/84 to 1995 (in `000 tons). (Source: The Gambia 1996a, DoP 1996)

In 1990, the Programme for Sustained Development (PSD) was launched as a successor to the ERP. Failure of the food and credit polices forced the authorities to adapt the agricultural policies implemented during the ERP-era. Survey estimates indicate that women cultivate 92% of rice areas (Johm 1992). Improving women's access to land was (contrary to the ERP) of large interest since both rice and vegetables are grown by women. (Akinboade 1994). The major aims of the PSD in addition to raise food production and reduce food imports are, as in the ERP years, to expand groundnut production, and diversify food and export crop production (The Gambia 1996a, Akinboade 1994).

The aim of increasing food production has as in the whole post-independence era resulted in a variety of large rice projects (see chapter 4). The PSD was no exception. In addition, improved techniques for coarse grains were promoted (Akinboade 1994, The Gambia 1996a). The food security situation at a national level has, however, grown worse. Remember the rapid population growth shown in figure 4.8 and the increase in urban population shown in table 4.2. Swamp rice production as shown in table 5.1c and figure 5.2c has decreased. Early millet has increased, but is seldom sold on the market. The increased early millet production therefore, cannot serve the urban market. Although, the increase in early millet production is far less than the increase in demand for food at a national level. The gap between production and demand is increasing, causing increasing dependence on food imports and thereby increased import costs (Gambia 1996a).

5.6 The relevance of the farming household to the farming system
We have now been discussing the different relevance of different variables to the production alterations discussed in sub-chapter 5.1 and the interplay between the variables. The analyses have shown that no single variable of the entire system can directly explain the most sounding trends within the farming system:

1.There is no downward trend in yield per Ha. production of the three selected crops although;
i) -there has been a long-term downward trend in rainfall
ii)-soil productivity is dismissing due to decreased use of shifting cultivation and intercropping

2.The production of all crops tend to be stagnant although;
i)-the demand for food and socio-economic development has increased due to population growth
ii)-the level of mechanisation has increased
iii)- the labour availability has increased in rural areas
iv)-there is no absolute shortage of productive land at a national level

3. There has been a shift in production both in cultivated area and total production. The subsistence crop early millet has become more important while groundnut and swamp rice has become less important although the agricultural policies are aiming at increase groundnut and swamp rice production through commercialisation of the agricultural sector

These paradoxes might be partly explained by the social system of farming and the farmers responses to changing external influences such as liberalisation of the agricultural polices and the long-term downward trend in rainfall. The risk of farm operations has increased. As discussed in chapter 1, the farmers are active respondents to external influences such as risks. If the farmers were passive respondent we should expect that the all variables under 5.2, 5.3, 5.4 and 5.5 should influence the production alterations discussed in sub-chapter 5.1 directly. According to the theories of multiple modes of livelihood (Mustapha 1992) and risk responses (Corbett 1988, Lipton 1968) discussed in chapter 3, the farmers tend minimise risk through for instance, risk dispersing.

Below, changing farming practices and social organisation of farming will be discussed in the light of the paradoxes presented above. However, the third variable of the farming houshold system, responses, also causing changes in farming practices and social organisation will not be discussed here as whole the next chapter is consecrated the importance of this variable.

5.6.1 Farming practices
According to Willis (1996), monoculture tends to replace shifting cultivation and intercropping in the upland area of The Gambia. As described in chapter 4, the shifting cultivation practise includes in addition to uses of fallow, crop rotation which optimises the proportion between the nutrient content of the soil and nutrient demand of the crop. In the case of crop rotation, lack of this practise is a major drawback to pest management. Crop rotation as a practise is not abandoned, but rotating crop each year could drastically reduce diseases and pests (Canteh & Sanyang 1996). Use of the related intercropping is decreasing, resulting in less organic restoration of the soil and by this dismissing soil fertility (Hyden 1988, DoP 1990, 1996).

Table 5.6: Intercropping in per cent of total area practised by crop 1989/90 and 1995/96.
(Source: DoP 1990, 1996)

There exists, however, no statistical data on fallow use, but as USAID (1992) Bojang (1996), Willis (1996) and (DeCosse 1992) points out, fallow periods are declining dramatically resulting in declining soil fertility. Using decreased fallow periods, as an operationalisation for land shortage at a national level, as USAID (1992) does, might be a foundation of misconceptions. While land resource statistics indicate no absolute land scarcity in The Gambia, land pressure has affected the shifting cultivation practises at a local level (Willis 1996). DeCosse (1992) claims seasonal labour shortage and decreased will to invest labour in farming to be an additional cause of decreased fallow practise. According to Johm (1988a), and Braun & Puetz (1988) field surveys have discovered that the yields of upland crops such as early millet and groundnuts are declining when more labour is put into swamp rice production in the busy season. Import substitution through increased domestic food production might, therefore, lead to decreased cash crop production. Mechanisation of farm operations might be an effort to solve the problem of seasonal labour shortage, but also a cause of changing farm practices (Bojang 1996). The labour cost of preparing a plot for the use of light equipment, such as the seeder, is high compared to cultivating the same plot each year. In addition, the capital cost of buying oxen, which are necessary to cultivate the soils after bush fallow is also high (The Gambia 1987a). The result might be decreased use of fallow and by this dismissing soil productivity

5.6.2 Social organisation
According to Johm (1992), there is a large potential for cultivating more areas, especially in the lowland of The Gambia. However, the tenure systems lack of ability to adapt to demographic processes tends to be a limitation for increased cultivation of the arable land. Even if there is no absolute land scarcity at a national level, land scarcities does exist as a result of the traditional tenure system and the pattern of settlements. The tenure system has evolved for a low population pressure on the land resources and not the population growth of the last decades. According to Willis (1996), the traditionally tenure system has gradually been eroded with increased individual ownership of the traditionally claimed kamanyango land. However, in a large proportion, individual ownership is strictly limited at household level. The maruo fields are still farmed on communal basis (up to 90% of total coarse grain area) and individuals have no ownership rights of the kamanyango land. The result is limited interest in the status of the land they cultivate and limited interest in long term investments in the field. Many young men and women lack the ability to get already cultivated kamanyango land to raise initial capital for clearing and clearing and cultivating new areas. Even for most of the established farmers, the cost of clearing new land is too high (Johm 1992). The limitations of tenure system are, however, a cultural phenomenon rather than an economic question. Freudenberger and Suso (1993) claim the cultural founded awareness of selling unused arable land as a limitation for increasing the cultivated area in The Gambia. The result is, according to Jagne (1996); migration of the landless to urban areas as a risk disperses to the increasing risk of farming.

previous     next

© afrol.com. Texts and graphics may be reproduced freely, under the condition that their origin is clearly referred to, see Conditions. The copyright of this specific text does not belong to afrol.com, but to the author. See special note.    You can contact us at mail@afrol.com