Torgeir Fyhri 1998

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

Chapter 7: Testing of Hypotheses and Theoretical Discussion

Contents

CHAPTER 7: Testing of Hypotheses and Theoretical Discussion
7.1 Testing of hypotheses
   7.1.1 Examination of hypotheses IA & IB
   7.1.2 Examination of hypotheses IIA & IIB
   7.1.3 Examination of hypotheses IIIA & IIIB
7.2 Theoretical discussion
   7.2.1 Modernisation, stagnation or transformation?
   7.2.2 The relevance of macrotheories and the present findings

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

This chapter sought to summarise and discuss the findings of chapter 5 and 6 in relation to the problem presented in chapter 1 and the theories presented in chapter 3.

7.1 Testing of hypotheses
The hypotheses presented in chapter 1 are discussed in this chapter with the basis in the empirical findings of this study. The findings of chapter 5, related to the causal model, figure 5.1, are compared to the findings related to the case study of chapter 6. Hence, the quantitative data becomes adjusted by qualitative data and the, national agronomic data becomes adjusted by lower scale data.

7.1.1 Examination of Hypotheses IA & IB

IA Due to the decreasing prices on groundnut, removing of subsidises on fertiliser and increasing prices on food grains, the farmers have turned back to more subsistence agriculture to secure the food availability.

IB Increased demand for food caused by population and urban growth added with less interest in commercial agricultural production causing a stagnation in domestic swamp rice production, has resulted in increased food imports.

As discussed in chapter 5, the total crop production has been rather stagnant during the period studied in this thesis, except from the non-marketed, domestic consumed crop early millet. Groundnut production in particular has been declining. The findings related to the model, figure 5.1, shown that no single phenomena can explain this development. However, short duration crops as early millet give more uniform annual yields than groundnuts, and might be preferred, as a long-term rainfall decline has occurred. In addition, prices on fertiliser, on which groundnut production relays on, have been increasing, due to removedness of subsides in 1986. Privatisation of the credit institution led to squeezes of the access to credit among farmers, especially women.

Hence, it might be concluded that farmers have turn back to subsistence production to secure food availability, because the production of subsistence crops as early millet increase while cash crop production decreases. However, as discussed in chapter 6, the farmers do invest money in farming, in technical improvements such as seeder and sin-hoe-package-outright. The investments are managed by access to off-farm income by one or more household members. Off-farm incomes tend to be higher than incomes from farming, and in addition, more stabile. Farming becomes; therefore, a subsistence related activity in a household economy influenced by commercial thinking.

Even if early millet production is increasing, swamp rice production tends to be stagnant or even decreasing. Rapid population growth added with an increasing urban population involves increased demand for food at a national level in The Gambia. Food imports bring upon large national expenses. The agricultural authorities direct their efforts in increase the national food production to the lowlands since upland food crops compete with groundnuts. Large national investments therefore, are done to improve swamp rice production and increase swamp rice production to increase the national self-sufficiency degree. However, swamp rice production includes large labour investments at a farm level since the opportunity to invest in labour saving equipment in the lowlands is less than for upland crops. This labour that could be better obtained elsewhere, either in early millet production or in non-agricultural sector.
Conclusively: Hypotheses IA, must be partly rejected. The farmers do produce more subsistence crops, but due to increased off-farm income invest more money in agricultural production to release labour. Hypothesis IIA is, however, confirmed by the findings of chapter 5 and 6.

7.1.2 Examination of Hypotheses IIA & IIB

IIA The disability of the traditional land tenure system to be adjusted to agricultural commercialisation and population growth leads to fixed land scarcity.

IIB Long-term downward rainfall trends the last decades have created a labour bottleneck in the short rainy season.

Land availability is seen as a barrier to modernise the crop sub-sector. The declining practise of fallow is stated as an operationalisation of land scarcity. The argumentation goes further, land scarcity leads to less use of fallow, the soil fertility dismissing and yields are declining. However, even if the estimates of land reserves in the Gambia differ considerably between different sources, all the sources claim that there is no scarcity of land in absolute terms. In addition, there are no signs that the changing fallow practise is a result of land scarcity. In the first place, the yields per ha are declining for neither swamp rice nor groundnuts nor early millet. The method of estimating soil fertility through by yield per Ha. production is seemingly invalid. As discussed in chapter 6, divisional data shows that fallow practise is less in divisions with largest remaining land reserves.

Labour availability in the short rainy season has traditionally been a constraint in the Gambian cropping system. Rapid population growth the last decades should theoretically impair the importance of this constraint. However, as for the case of land scarcity there is still scarcity of labour in many villages especially in villages mainly living of swamp rice production. Even if there are enough labour in absolute terms at a national level, low mechanisation level in swamp rice production, slow income growth in crop production compared to other sectors and scarcity of season have made the farmers less interested in invest labour in farming. Also this might explain the declining fallow. The labour input of clearing field left fallow is high compared to cultivating an already cleared field.

Long-term rainfall trend has resulted in increased scarcity of season. Increased scarcity of season might explain the paradox of increasing agricultural workforce and labour shortage. In addition, the inability of the tenure system to adapt to the rapid population growth has been an obstacle for utilisation of the increasing agricultural workforce and technical improvements and investments. Even if there exist reserves of arable land at a national level in The Gambia (FAO 1995, DoP 1996), is the tenure system an obstacle for utilising it. Due to the constraints of the tenure system, there are fixed land scarcity at a village level. Population growth and urban growth have made the access to land for households with less secure ownership rights more uncertain. (Land borrowed by these households, might anytime be reclaimed by landowners for other land use such as urban growth.) Ambiguous young farmers are lacking the ability to clear and cultivate new land because land cannot be sold.

In 1990 the State Land Act was designed with the intention of restructuring the land ownership system of The Gambia (Freudenberger 1993a). The act has however, not yet been fully implemented (MANR 1995a, Freudenberger 1993b). In a report presented by the government of The Gambia, Vision 2020 (The Gambia 1996b), reforming the tenure system is at focus, but no concrete efforts were presented. The question can be raised whether such reforms would help removing the fixed land scarcity. Lawry (1994), claims that there is a need for tenure reforms in general, but such reforms cannot alone result in a sustainable resource management. The capital and labour investments of clearing new land is high. Most of the farmers are not economically viable to make such investments. In addition, increased scarcity of season requires more labour input in the already cleared and cultivated areas in the short rainy season to overcome the labour bottleneck.

Conclusively; both hypothesis IIA and IIB is confirmed by the findings of chapter 5 and 6.

7.1.3 Examination of Hypotheses IIIA & IIIB

IIIA Commercialisation of the crop sub-sector leads to increasing farm risk due to higher vulnerability of prices and more vulnerable crops.

IIIB The farmers tend to responding to increasing farm risk through efforts minimising or spreading the risks.

The development of increased subsistence production discussed in 7.1.2 is seemingly caused by the increasing risk of cash crop production. USAID (1992) sees this development as a paradox because liberalisation of the sub-sector was meant to favour cash crop production. The liberalisation efforts, including reforms of the credit system, removing of subsidies on fertiliser and privatisation of parastatals, has, however, to a high degree removed the governmental protection of the sub-sector. It has become more risky investing in crop production and especially groundnuts. In addition, long-term decline of rainfall has increased the risk of groundnut production, which is a less drought resistant crop than e.g. early millet. Fluctuations in yield per ha is also higher for groundnuts than for early millet and as claimed earlier in this chapter, large yield per Ha. fluctuations contain a higher risk than small fluctuations.

According to Boserup (1965, 81), population growth should force the farmers to intensify their production, even with limited land resources. This growth has not occurred. The agricultural production system has showed inability to keep up with the rapid population growth. There is increasing use of improved technical equipment and a growth in the agricultural labour force. This indicates better production capacity and should result in increased production. In addition, shifting cultivation tends to become abandoned. Improvements of farming equipment and greater access to labour have not yet resulted in increased crop production so far. This is a paradox. Farm practices have become more modern while the choice of crop has become more traditional and more subsistence oriented. Adapted risk responses, as discussed in chapter 3, might be an explanation. Risks vary timely and spatial, so also risk responses (Taal 1989). While shifting cultivation and fallowing were the responses to the risk of the vulnerable climate, risk spreading, as earning off-farm income, might be the a response to the increasing economic farm risks of liberalisation efforts (DeCosse 1992). It tends to be safer getting income elsewhere and consume imported rice than e.g. invest in rice implements.

Off-farm incomes are secured either by remittances from emigrated family members, or through crafts in rural areas. Increased off-farm income might explain the paradox of increased investments in labour saving farming equipment when credits are less available.

Conclusively; hypothesis III A, must be confirmed. However, hypothesis IIIB can be partly rejected. The farmers do minimise risk through risk dispersing. However, the farmers are not strictly risk minimisers, the farmers are instead as discussed in chapter 6 more utility maximisers, that seek to gain the maximum profit possible without increasing the risk.

7.2 Theoretical discussion
Some general theories related to the topic of this thesis, were presented and discussed in chapter 3. Something might be said about the relevance of these theories in the light of the findings of this study. We must, however, be aware of not drawing the conclusions too far. Even if some of the macrotheories are of limited value in studies at a lower geographical level such as this, they might be of relevance in studies of the whole sub-Saharan Africa, or regions on the sub-continent such as West Africa. In addition, we must be aware of generalising findings of this study. To adjust macrotheories in this field, or generalise the findings discussed below, additional studies in other African countries and updating of data, have to be done.

7.2.1 Modernisation, stagnation or transformation?

As claimed by Hyden (1988) and found in this study, the agricultural authorities are seemingly not paying enough attention to the social sustainability of the farming household. The agricultural policies are seemingly performed in a Western tradition of modernising and economical development. Strategies has been performed and implemented to force the farmers to become commercial farmers. This policy is based on the assumptions that the subsistence production is economically unviable, and that the farmers will change their perceptions to commercial thinking if forced to. The unexpected responses, such as decreased groundnut production, are according to Maalenku (1993b), a result of lacking knowledge among farmers. USAID (1992) and MoA (1993), claim that the farmers probably do not know their own best as they do not follow the modernising strategy of the agricultural authorities. However, the risk minimising strategy is not a result of ignorance of the benefits of profit maximisation, but a well-developed strategy against uncertain conditions (Lipton 1986). The basic feature of this strategy is rapid adoption of risk cope responses. Adopting the profit maximising strategy the farmer might well benefit economically in the long run. However, such a strategy, including monoculture cultivation of cash crops (Maalenku 1993b), increase the possibility of greater output fluctuations as discussed in chapter 5. The farmers are not able to take such a risk; they are dependent on stabile yields and stabile output (Taal 1989). Benefit in the long run does not matter for the farming household if it leads to failure in the short run, as this might turn out desasterous (Lipton 1968). Selecting a year to year strategy, the farmers tend to be more dynamic in their responses than both agricultural authorities (MoA 1993) and scientifical institutions as USAID (1992) expect. However, the national economy does not necessarily benefit from these responses, but the farming household does. According to CSD (1994), large scale groundnut farmers are by far the poorest group in The Gambia. This is a rather unconventional finding. The large scale farmers would be expected to be rich while small scale farmers are expected to be poor. This would be true if cash crop production was economically viable. However, at the present, groundnut prices are low, while incomes from non-farm activities are increasing. The farmers need cash to buy food, farming implements and consumer goods. Small scale farmers have been forced to obtain off-farm income for years, and are, therefore getting more of their incomes from off-farm work than large scale farmers (Jabara 1990). Job opportunities outside agriculture are increasing and the incomes are far higher than in crop production. According to DeCosse (1992) the result is producing food for subsistence on the farm and obtain cash elsewhere.

The agricultural policies in the Gambia are mostly performed with the basis in the view that agricultural modernisation can be planned at a national level and that farmers are passive respondents forced to follow the aims of the agricultural authorities. However, the analyses of chapter 5 and 6 shown that the actual development within the national agriculture is occurring as a result of these responses rather than as planned for. This is a matter of the dialectics of space where action at all geographical scales influence action at other geographical scales (see figure 1.1). The farmers are actors, or active respondents, that can influence the development or even develop by their own. By this, in the case of The Gambia, we might say that the real decisionmaking is made at a household level, and not at a national level. A logical succeed of real decisionmaking based on local economy (see chapter 6) and not on the national economy is, as DeCosse (1992) and CSD (1994) claims, that farmers are obtaining off-farm income instead of investing in agriculture. Such investments become insecure as a result of weaker governmental protection of the crop sub-sector. However, scientists as Hinderink & Sterkenburg (1987) are assuming that modernising and development are the same as increased output and commercialisation of farming. Therefore, statistical data such as yields per Ha; total production and price level are usually analysed to conclude whether the crop sub-sector is stagnating or developing. With the basic in such statistics, the crop sub-sector of The Gambia is not modernising, but stagnating. However, as claimed by DeCosse (1992) and concluded in chapter 5, such data have limited value in analysing what actually happens and why this actually happens at lower decisionmaking level. He further claims that lack of attention to this issue has resulted in lack of additional data related to responses. Assuming that the local economy is the driving force beyond decisionmaking, and modernising is defined as making suitable for present day's needs, the agricultural sector of The Gambia is modernising. The rapid responses on external changes are in fact real modernising. The agricultural policies are external change leading to increased farm risk. To cope with this risk, the farmers are making their decision suitable for present day's needs, which means risk minimising and utility optimising.

7.2.2 The relevance of macrotheories and the present findings

The premise that a local cultural and resource based economy is of greater importance than the national economy is of vital importance when discussing the relevance of macrotheories in the light of the findings of this study. In sub-chapter 3.2.1, four main theoretical schools discussing agricultural development were presented: The Malthusian, the Boserupian, the neo-liberal economists and the neo-Marxist. These four different schools, should, since they are all related to agricultural development be a the theoretical starting point of this thesis. However, discussing agricultural development at a regional or national level, the theories have limitations in this study of The Gambia. Below, the relevance of the macrotheories for this study are evaluated.

The neo-liberal and the neo-Marxist school are seemingly theoretical extremes. The neo-Marxist views strong governmental interference in the economical activities as necessary for developing the agricultural sector. The colonial heritage and dependence relations to developed countries are barriers for modernising the agricultural sector in countries as The Gambia. Contrary the neo-liberal school is blaming governmental interference of African countries as an obstacle for development of the agricultural sector. Privatisation of agricultural institutions and market value on all inputs and crops will result in agricultural growth. However, both the neo-liberal and the neo-Marxist school are sharing the view that national authorities can plan agricultural development, and that decisions of importance are taken by agricultural authorities. The basic view is, of course, different when it comes to how this policy should be planned.

These two theoretical schools have been dominant in the post-colonial era of The Gambia. For both, agricultural development is the same as increased total output of both food crops and cash crops. Their conceptual link to euro centric modernisation theory thus becomes quiet visible. Commercialisation of crop production is claimed by both as a necessity for reaching the aim of agricultural development. Subsistence production is seen as an obstacle by both. Forcing the farmers to change from subsistence to commercial production was an important effort of the agricultural policies inspired by those theories. Intensive production of both cash and food crops should replace traditional farming techniques, and both should be produced for the market. The belief was that the farmers themselves did not know their own and the country's best. If the farmers where given responsibility for their own production, it was believed that they would turn back to subsistence production. An illustrative example is the failed tractor project in the village of Geniri in MID, discussed in chapter 4. The authorities were reluctant to give the land back to the farmers because it was believed that commercial production would be replaced by subsistence production (Haswell 1991).

Less incorporated in the policies but of importance in foreign aid and policies and in the scientific works on agricultural development in The Gambia and other African countries, are the neo-Malthusian and the Boserupian schools of thought. As discussed in chapter 3, according to a neo-Malthusian outcome, population growth will lead to increased land pressure more intensive cultivation and in turn dismissed soil fertility and declining output. The stagnation of the total crop production, as shown in chapter 5, added with increased demand for both food and cash due to population growth confirms in many ways the neo-Malthusian point of view. Food production has not balanced the population growth. However, when studying a lower geographical level and using more qualitative empiric findings, as in the Upper Baddibu case, a more complex picture becomes visible. It is neither population growth nor land scarcity that affects land degradation, but the constraints of the tenure system that leads to land scarcity and land degradation. In addition, the farmers' responses to changing external conditions limit and even exploit the effects of population growth. Increased population has resulted, as claimed by DeCosse (1992) and CSD (1994) and discussed in chapter 6, to the release of rural population mainly into the service sector as a multiple mode. Even if the development of the crop sub-sector tend to be rather negative when discussing the national level, decisionmaking and responses on household level leading to search for off-farm income have led to risk spreading and less dependence on an uncertain farm income.

Boserupian theories, especially when adjusted to African conditions as by Lele & Stone (1989), tend to be more relevant for the development of The Gambian crop sub-sector than the neo-Malthusian. In a Boserupian outcome, farmers will respond to population growth by improving their farming techniques, turn towards more labour intensive farming with increased output as a result (Mortimore 1993b, Boserup 1965 & 1981, Lele & Stone 1989). However, this development pre-supposes a market value of all crops leading to price increase when running scarce, and that food aid is disturbing the market. The farmers then will invest in farming to increase production and profit. The given presuppositions are, however, rather theoretical. As Kargbo (1983) claims, Gambian rice production cannot compete with overseas rice producers. Further, spontaneous intensives of crop production requires that the farmers are interested in maximising profit. According to Lele & Stone (1989) and Taal (1989), the uncertain ecological, economical and political conditions of African countries have led to risk minimising and risk spreading rather than commercialisation of farming. The farmers, as claimed by Boserup (1965) respond rapidly to external conditions such as population growth and changes in the market, but not necessarily by efforts maximising farming output. Investments in farming, such as mechanical equipment, added with greater access to labour, have not resulted in increased production as expected by Boserupians, but into a stagnant or stabile production and a release of labour.

The discussion above, concerning the relevance of macrotheories discussing population growth and agricultural developments, does not intend to reject these theories. The geographical scale at which the study has been conducted and too little empirical findings is not sufficient for such a statement.
The actual development of the Gambian crop sub-sector has only few similarities with the expected development. Using the "modernisation" of The Gambia as a case, this thesis indicates that theories in the field cannot be suitable when the agriculture sector is isolated from others. The presupposition that it is the governmental policies and population growth that affects the real agricultural production tends to be infertile. The barriers for "modernising" the crop sub-sector are to be found elsewhere, in the system of the local utility maximisation economy. The empirical basis of such a small country has of course limited value when it comes to verifying or falsificating macrotheories. However, the majority of studies in this field are conducted at a national or a local level and here, the macrotheories are of limited value and can hardly be applicated. Instead, as Hyden (1988) claims, scientific works within this topic have to focus on indigenous natural and human resources within each country. Future research on this topic should according to DeCosse (1992) build upon the rural households' interest in risk and income diversification.