Resisting reforms. A Resource-based perspective of collective action in the distribution of agricultural input and primary health services in the Couffo region, Benin - 3. AGRICULTURAL INPUT SERVICES IN THE COUFFO REGION

Resisting reforms. A Resource-based perspective of collective action in the

distribution of agricultural input and primary health services in the Couffo region,

Benin

Dedehouanou, H.

Publication date 2002

Link to publication

Citation for published version (APA):

Dedehouanou, H. (2002). Resisting reforms. A Resource-based perspective of collective action in the distribution of agricultural input and primary health services in the Couffo region, Benin.

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

Thee present chapter deals with background information in order to pave the way for the analyticall presentation in the chapters ahead. Indeed, it is expected that the agricultural input servicess encompass the agricultural sector as a whole. In Benin, however, these services only addresss input demand from the cotton scheme, despite the political rhetoric claiming a full coveragee in agriculture. Consequently, village communities compete to ensure an improved accesss to the distribution of agricultural inputs originally destined to cotton. But, for future developmentt not only cotton production matters, other commodities are also important. The rationalee for the neglect of those commodities and the over-emphasis on cotton will be documentedd in this chapter. A special emphasis will be on the description of agricultural policy sequencingg in Benin.

Thiss chapter will provide sufficient information to set the background for an understandingg of the conditions within which reforms are implemented and performance is achieved.. It is believed that a good understanding of the mechanisms underlying these reforms andd their performance in the cotton sector will enlighten the way in which their objectives may bee expanded.

Recalll that this study is to address the underlying organisational conditions of village communitiess regarding the strategy, structure and core capabilities for the distribution of agriculturall input services. More specifically, the conditions underlying village communities participatingg in collective action are to be assessed. Given such perspectives, this chapter will focuss on the physical, social, and economic as well as on the organisational aspects of the village-levell participation.

Inn order to attain such an objective, this chapter is organised into sixx sections. The first one willl highlight the main features of the people and of the agricultural sector in the Couffo region. Thee second section will concentrate on recent government intervention in agriculture. The third sectionn will briefly present government-initiated farmers' organisations and agricultural input services.. The fourth one will deal with some major issues influencing the distribution of agriculturall input services. The fifth section will present agricultural policy sequencing and the lastt section will draw some concluding comments.

3.11 People and agriculture in the Couffo region

Itt should be stressed from the outset that background information would emphasise the larger entityy that was originally known as the Mono region, of which the Couffo region is an integral part.. Reference will be made alternatively to Mono, the larger entity, and Couffo, its northern part. .

Mapp 3.1 gives a picture of the Couffo region as a densely populated area. The region also hass the highest rate of rural population, 80 per cent against 64 per cent for the national level (INSAEE 1994). The regional population is mainly rural in origin and is engaged in agriculture. Thee Aja were first and foremost peasants, although some changes are observed nowadays. However,, as it appears below, policy discourse on the agricultural development in the Couffo regionn necessarily invokes demographic pressure and land scarcity. In fact, this region is the one withh the smallest per capita cultivable land (.43 ha) compared to the other regions in Benin (Mongboo et al. 1995, CARDER-Mono 1997). These figures suggest both specialisation and high intensificationn processes with regard to agricultural activities and land use, respectively (Boserup

Mapp 3.1: Population Gradients and Agro-Ecological Zones inn the Couffo Region

Less than 200 From 200 to 500 From 500 to 1000 From 1000 to 2000 From 2000 to 5000 More than 5000 BB District capital centre

Boundaryy of district territory

o o o o

thee organisation of the labour force and the land tenure system amongst others, the dearth of a holisticc approach to agriculture intensification from the side of the intervening government organisationss also aggravates land scarcity issues (CARDER-Mono 1997, p. 15; p. 35-36). An overemphasiss on the cotton crop further diverts local as well as government efforts to boost agriculturall development Therefore, agricultural performance hardly keeps pace with demography,, yet the effects are subsequent to population gradients in various agro-ecological zoness (cf. Map 3.1).

Agro-ecologicall zones in the Mono region: Five main agro-ecological zones are delineated followingg soil and climatic characteristics. In addition, they also correspond to different populationn gradients (see Map 3.1). The first zone, the Lonkly Savannah to the north of Aplahouéé (100,400 ha, 24.4 per cent of the regional total), is sparsely populated. This area marks aa transition between two climates, the sub-tropical climate with two rainy seasons and two dry seasonss in the south of the Couffo region, and the Soudano-Guinean climate with one rainy seasonn and one dry season in the north of Benin. Recall that, in the sub-tropical region, each agriculturall campaign consists of two growing periods in which one rainy season is coupled with onee dry season. Recall also that the first growing period usually starts by early March and lasts tilll August, while the second one starts by September and ends with the long dry season from Novemberr to March. The Lonkly Savannah is contained within the research area and is currently subjectt to voluntary resettlement (cf. Map 3.1).

Thee second zone, the Plateaux (179,240 ha, 43.6 per cent of the regional total), of which thee Plateau Aja is an important part, is densely populated. This zone leaks its excess population too the less populated area within and outside the Couffo region.

Thee third zone, the Tchi depression made of heavy clay soils (70,400 ha, 17.1 per cent of thee regional total), is sparsely populated but is still receiving flows of settlers from the Plateaux. Thiss zone is not within the research area, but it receives migrant flows from it.

Thee fourth zone, composed of the alluvial valleys alongside the main rivers (30,020 ha, 9.55 per cent of the regional total), is either densely or sparsely populated, following the Mono andd Couffo rivers, respectively. This zone is partly within the research area.

Thee last zone, the littoral and lagoon area made up of banks of sand (21,960 ha, 5.3 per centt of the regional total), is moderately populated. This zone is not within the research area.

Owingg to the fact that rain-fed agriculture is mainly the basis of the production system, climaticc hazards are very often the cause of failure or success of the cropping seasons in this regionn (RAMR 1987 through 1990). The erratic nature of rainfall causes either drought or floodingg at peak periods of the planting season. Farmers' concerns in this respect are so intense thatt they have devised traditional rites to exorcise the unpleasant effects (Agbo 1995). In the samee vein, on-farm researchers have suggested an environmental index to account for the variabilityy of crop yields caused by soil as well as climatic factors. This index represents the averagee yields of all sample experiments for each variety of crops in one agro-ecological zone. Forr the same planting dates, RAMR (1987 through 1995) found large discrepancies between thee lower values and the higher ones. This is a justification for the central influence of uncontrollablee environmental factors on the agricultural results in the Couffo region.

Withh respect to population density, the regional average has increased from 121 per km2 inn 1982 to 168.7 in 1992, due to natural demographic growth (see Map 3.1). There are, however, significantt differences between sub-prefectures. TovMin has the highest population density (410 inhabitantss per Km2), while Dogbo has the lowest (202 inhabitants per Km2). Whereas some areass are very sparsely populated, such as the Lonkly zone in northern Aplahoué (42 inhabitants perr Km2 in the south and even 20 inhabitants in its northern part), others have population densitiess well above 200 inhabitants per Km2 (Azovè in south-east Aplahoué and other

communess in Djakotomey, Klouékanmè and Dogbo). Obviously, the demographic pressure has bearingss on the access to and the control of land resources, a theme that is very relevant to the organisationall advantage of a village with regard to agriculture in general, and cotton in particular.. In the same vein, the effects of this demographic pressure on the existing cropping systemss will be discussed below.

Landd tenure and the existing cropping systems in the Couffo region: The land tenure system iss very characteristic of the high demographic pressure in the Plateau Aja zone (land market, rentingg for short and long periods, sharecropping, etc.) and the low pressure in the Lonkly

SavannahSavannah zone (inheritance, borrowing, etc.). In the former zone, a large proportion of the land

iss under permanent cultivation (tree and oil palm plantations in Toviklin and Dogbo, for instance).. In addition, sharecropping is much practised in this zone, more than elsewhere in Benin.. Thus, there seem to be increasing gradients of land availability south-north, counter to the existingg pattern of population densities. However, this is not to assert that relatively large land-holdingss do not exist at reasonable distances from most settlements in the south. Nor are there systematicallyy large land-holdings in the north. Although the Aja ethnic group is characterised by thee absence of collective rights on land (Biaou 1996), it should be acknowledged that, over time, theyy have devised strikingly efficient land-use arrangements that are geared towards overcoming landd scarcity issues. In fact, per household within the villages, relatively comparable parcels of landd under cultivation are observed.

Land-resourcess are transferred through two channels, the first involving financial or in-kindd compensation, and the other one involving no compensation (cf. table 3.1). If the former channell could be characterised by its market-orientation, the latter one reflects the existing social ties,ties, lineage, kinship, marriage relationships and the like. Each channel is made up of different modess of ownership transfer or simply use access. Use access derives logically from the resource-basedd perspective that stresses the distinction between resources as stocks, and resourcess as flows, the services that may be obtained (cf. Foss 1997). Trus perspective applies withh a greater significance to land resources because farmers in the research area cultivate relativelyy comparable parcels of land, irrespective of their ownership.

Tablee 3.1: Distribution (%) of cultivated land (area) following different modes of transfer and access in two

agro-ecologicall zones of the Mono region

Agro-ecological l Zone e Lonklyy Savannah 'Villagee Sinlita' Plateauu Adja 'Villagee Touléhoudji'

Channelss for land resource exchange and use access 'No'' compensation INN NPI GI BO 344 2 5 57 111 3 1 7 Totl l 98 8 22 2

'Cashh or in-kind* compensation PUU LR SR SC PC CO 11 1 -144 29 3 25 6 1 Tot2 2 2 2 78 8 TOTAL L 100 0 100 0

Source:: Extracted from Biaou (1996)

Notes-- IN = Inheritage; NPI = Non-partitioned inheritage; GI = Gift; BO = Borrowing; Totl = Total number one;

PUU = Purchase; LR = Long-period renting; SR = Short-period renting; SC = Sharecropping; PC = oü Palm-contract;; CO = Collateral for loans.

Regardingg the non-market transfer, Biaou (1996) found inheritance as one of the most important modess in Sinlita (commune of Godohou in the Lonkly Savannah\ where it represents 34 per centt of the total area underr cultivation. Li another village, Touléhoudji in the Plateau Aja zone,

thee author found 11 per cent for this mode. He further uncovered that this mode of land ownershipp transfer discriminates against women, who only inherit in exceptional circumstances. Inn certain cases, so he found, inherited land-resources are attributed to the heirs in block, with thee purpose of preventing them from selling out There seems, however, not to be a significant differencee between both villages in this respect (2 per cent in Sinlita and 3 per cent in Touléhoudji). .

Giftt is the next common mode of use access (5 per cent of the total area in Sinlita). This occurss within households when the male household-head is still alive. Different variants are observed,, e.g. use access granted by the latter to his children, pure gift when a child can plant treess on the land, and loan without compensation when this right is temporary. A similar mode off land attribution exists between household-heads and their wife or wives. Land scarcity might explainn the low score of gift in the Plateau Aja zone (1 per cent in Touléhoudji).

Denn Ouden (1989,1997) and Biaou (1996,1997) provided a symbolic categorisation of fannerss with respect to agricultural resources, for instance, the accumulation of farming assets suchh as land, labour, livestock and others. According to their nomenclature, beginners are thosee farmers at an early stage of accumulation who are not different from non- accumulators inn their initial resource endowment, but who readily innovate their resource management. The

accumulators,accumulators, on the other hand, are farmers with land, labour and material assets. The

referencee made to resource management is very instructive and calls attention to the concept off competitiveness. For contiguity reasons, a village with innovative accumulators may foster itss position because farmers replicate successful management practices of neighbours.

Inn addition to land ownership, which confers to the beneficiary endless utilisation rights, theree are different modes of use access for a limited period. Borrowing is such a short-term mode,, as opposed to renting, which involves compensation. Biaou (1991,1996,1997) found that borrowingg is the leading mode for women's access to land. Women even borrow land outside theirr household, either from their own family or from their in-laws. In the less populated area of thee LonJdy savannah, borrowing is foremost among the other modes of use access (58 per cent off the total area). While this mode is not the most important one in the Plateau Aja, it is by far thee leading mode of use access without compensation (7 per cent of the total cultivated area).

Thee second channel for ownership transfer involves compensation. The land market is one modee through which ownership is transferred and acquired. This process, as Biaou (1996) argued,, is increasingly getting momentum in the Couffo region, with significant differences amongg sub-prefectures. Mongbo et al. (1995) call our attention to Wartena (1987), who found significantt ratios of land purchased to total land under cultivation (16.4 per cent) and to total landd owned (23 per cent) in Toviklin. In line with Wartena's figures, 14 per cent of purchased landd areas over the total lands under cultivation (cf. table 3.1) underscore the alleged acute land scarcityy in Touléhoudji (sub-prefecture of Toviklin). Den Ouden (1989) also inquired into land tenuree in the same village, and he found the villagers much less captured by the land market, as originallyy expected. It appears that many conflicts are related to purchased lands, although this modee probably represents the most secure way of land acquisition. By contrast with the market rationale,, people resourcefully devised new access modes for not selling out their means of livelihood d

Rentingg is the most important access mode, with several variants. One variant is with

financialfinancial compensation. Although this mode is well developed in the Plateau Aja zone, differentiationn is observed following local circumstances and the knowledge of agricultural

risks.risks. Consequently, the time horizon and discount rates are increasingly incorporated in this mode.. Biaou (1996) recorded a disproportionate share for long-period renting (29 per cent of the totall land under cultivation, rented for twenty and more years) compared to short-period renting (33 per cent of land under cultivation, rented for less than six years). Another variant of renting is

withh compensation in kind, the so-called sharecropping. The author found this mode to represent onee fourth of the area under cultivation in Touléhoudji (Toviklin). In fact, this mode entails three differentt aspects of compensation. One is through sharing of the agricultural output, where the landownerr receives one third or half at harvest. A second aspect covers the share of both annual andd even tree crops. The first two variants are those described under the heading of sharecroppingg in table 3.1, while a third aspect of compensation involves specific contractual schemess between landowners and tenants. For instance, the oil palm contract is a contractual modee between landowners and tenants according to which tenants enjoy the agricultural produce onn a yearly basis, while they plant and maintain the oil palm trees for the benefit of the landowners.. Although all three aspects are found in the Plateau Aja zone, Biaou (1996) claimed thatt the last one is underrepresented, owing to the alleged high demands in practice (6 per cent of totall area under cultivation in Touléhoudji).

Concerningg the specific case of the Plateau Aja zone, differential access modes are in use withinn the villages or some more aggregate territorial units such as the sub-prefectures. Mongbo ett al. (1995) once again call our attention to Wartena (1987) and Ahlonsou (1987), who found thatt land renting for periods from 2 to 6 years is very common in certain villages, while people inn other villages prefer sharecropping. Den Ouden (1989), on the other hand, compared

sharecroppingsharecropping and oil palm contracts while differentiating between the sub-prefectures of

Dogboo and Toviklin, respectively. He found a significant difference between the Aja living in thee sub-prefecture of Dogbo and those living in the other four sub-prefectures of the Couffo region.. According to him, the former are less involved in agricultural work and thus highly value sharecropping,, while the latter are great farmers and prefer oil palm contracts to sharecropping. Thee author also found no proof of any correlation between wealth and land property. This result iss similar to that of Agossa et al. (1997), who found no relationship between poverty and landownership.. Den Ouden (1989) traced the accumulation process through borrowing and rentingg land in the first place, and then purchasing land with, as a primary condition, the control off and access to the agricultural labour force.

AA final variant of interest occurs when land is used as collateral for cash loans. According too this scheme, the landowner recovers his rights on the land only after paying back the loan. However,, only an insignificant proportion of this variant was found in Touléhoudji, as originally expected. .

Tablee 3.1 also depictss a contrasted view of two zones with regard to land tenure, one zone withh almost no market transactions of land (98 per cent of the total land under cultivation) and thee other zone with various modes of land transfer and access under compensation auspices (onlyy 22 per cent of the total land under cultivation). As could be inferred, the differential modes off land resource access and control described earlier may highly influence the existing local farmingg systems. In this respect, differential organisational advantages may be derived accordingly. .

Croppingg systems and agro-ecological features: A dynamic feature of commercial agriculture iss occurring in the Couffo region, due to the different contacts with rural development institutionss in general and with the extension service in particular. The adoption of cash-crops, suchh as groundnuts and cotton, has markedly changed the agricultural practices. Traditional cropss like maize and cowpea, often known as food-crops, are increasingly marketed in order to respondd to subsistence and market demand. Intercropping turns out to be the farmers' response too the increasingly higher demand on land-resources.

Intercroppingg is meant to sum up the agricultural practices consisting of planting one crop first,first, and planting another crop in-between later on. Those practices reflect on the population-landd gradients and on the agro-ecological suitability.1 Table 3.2 illustrates the practices of

intercroppingg for one production season in the Mono-Couffo region. The interest of inquiring intoo such a cropping pattern for this particular season stems from the reliability of the data at handd and the holistic picture it provides for the purpose of analysis. The reliability assumption derivess from the sampling procedure based on aerial photographs of the whole region, with detailedd information on the five agro-ecological zones (see CARDER-Mono 1996, p.8). The secondd assumption, on the other hand, is related to cotton, which is only cropped during the secondd growing period of the agricultural season. As such, cotton is hardly intercropped with others,, justifying the reason of not having cotton in table 3.2. Because of the need to spray insecticides,, intercropping cotton with another crop is forbidden and sanctions are unequivocal.

Intercroppingg ratios of maize crops carry the highest values in all five zones. This may be becausee maize is the major staple food-crop in the Mono region. In this respect, maize is mainly croppedd first at the start of any growing period. The observed intercropping ratios not only illustratee the existing population-land gradients, but also strikingly represent people-cultivable landd ratios. The littoral and lagoon zones, for instance, are not endowed with cultivable lands, leadingg to the highest intercropping ratios for maize crops. The Plateaux are second with their highh population density, hence high ratios of people to cultivable land. The Tchi depression, on thee other hand, carries most of the lowest scores for all crops because of suitability requirements. Itss heavy clay soils are unfit for root crops as well as any ground-fruit crops such as cassava, potatoes,, groundnuts, etc. This justifies why the mono-cropping of maize is widely spread in this zone,, suggesting both subsistence and specialisation orientations.

Tablee 3.2: Ratios of intercropping for five major crops in different agro-ecological zones of the Mono region, 2ndd growing period of the agricultural season 1995-96.

Croppingg systems Maize e Maize/others s Cassava a Cassava/others s Cowpea a Cowpea/others s Groundnuts s Groundnuts/ / others s Tomatoes s Tomatoes/others s Zonee 1 Lonkly y Savannah h Areaa Ratio 297 7 3877 1.30 2495 5 5799 .23 177 7 822 .46 2271 1 4511 .20 47 7 .00 0 Zonee 2 Plateaux x Areaa Ratio 13728 8 208522 1.52 13038 8 45322 .34 5031 1 3555 .07 5425 5 58966 1.09 381 1 .00 0 Zonee 3 Tchi i Depression n Areaa Ratio 8124 4 90022 1.11 1838 8 1411 .08 58 8 .00 0 13 3 .00 0 246 6 .00 0 Zonee 4 Alluviall valley Areaa Ratio 1243 3 18577 1.49 194 4 77 .04 30 0 588 1.93 17 7 .00 0 70 0 133 .19 Zoness 5 Littorall and lagoon n Areaa Ratio 31 1 1211 3.9 477 7 33 .01

Source:: adapted from CARDER-Mono (1996)

Note:: each ratio is obtained by dividing the area intercropped with other crops by the total area mono-cropped

Thee intercropping ratios rank second for cowpea mainly because of zones 1 and 4. The rationale iss that weed pressure is very low in both zones, whereas on the Plateaux cowpea is mono-croppedd and densely planted in order to control weeds. In addition, cowpea is extensively cultivatedd in the latter zone, suggesting a process of specialisation. In contrast to cowpea, groundnutt has a higher ratio in zone 2, where it is even more extensively mono-cropped than cowpea.. Indeed, but for soil quality requirements, its short vegetation period fits in with the existingg cropping season in most agro-ecological zones.

Cassavaa is often cultivated in the end of a cropping cycle, explaining its low intercropping ratioss in most zones. Obviously, its long vegetation period and its rusticity make cassava crop

appropriatee for restoring sou fertility. These reasons certainly explain its extensive cultivation in thee Plateaux zone and elsewhere on suitable lands, in addition to its primary role in the cottage food-industry.. The observed high intercropping ratio for zone 2, although relatively lower than thatt observed for maize and groundnut, is revealing of the acute land scarcity issue. Undoubtedly,, if planting cassava helps to reverse sou fertility decline, the observed ratio suggestss that the expectation of a quick fertility recovery prompts farmers to renew the cropping cyclee with this crop.

Tomatoess are intercropped with other crops during the two growing periods or are solely plantedd as off-season crops. This explains why the intercropping ratios are almost nil in most dry zones,, except for the alluvial zone. In the latter zone, farmers adopt off-season cropping because off the availability of wetlands and as a response to the great market opportunities for off-season agriculturall produce, tomatoes and vegetables, among others. Although farmers of zone 4 enjoy aa great deal of competitive advantage for off-season farming, they face a high competition from thee Plateaux zone, where the most important markets are located. Farmers of the Plateaux zone alsoo adopt the off-season mono-cropping of tomatoes and other vegetables, investing the existingg family labour in intensive watering and care.

Owingg to the intercropping practices discussed earlier, it is unlikely that farmers do pursue aa specific input-output goal that can justify the use of agricultural inputs on a targeted crop. Cottonn crop may be an exception to the rule, although intercropping practices literally invade the wholee production system. Blaikie et al. (1994) called our attention to the goal pursued by farmerss when their livelihood is at risks: they can adapt the pattern of labour and other inputs as

wellwell as outputs to changing situations, particularly those which arise in the aftermath of a hazardhazard (p. 48). Therefore, intercropping itself is one aspect of the farmers' response to risks

impingingg upon the production system. Where agro-ecological conditions are suitable, the intercroppingg practices may seriously influence a village's organisational advantages in agriculture. .

Thee community input system and agricultural practices: The community input system is adaptedd from Cromwell, who defined the community seed system [as] a wide range of informal

groupsgroups who share seeds amongst themselves, often on an irregular basis (1996: p. 4 ; [ ]

emphasiss added). The rationale for this adaptation derives from the generalisation of such a practicee to other inputs, including soil fertility conservation, pest control and the like.

Regardingg soil conservation, it should be stressed here that fanners make attempts to reversee yield decline drawing from their traditional practices. Two major types of farming practicess are relevant to yield improvement, viz. soil fertility enhancement and crop-seed selection.. Brouwers (1993), among others, found that farmers of the Couffo region are very innovativee in their attempts to reverse soil fertility dechne. A full description of those practicess is beyond the scope of this study. Nevertheless, it is instructive to note that green manuree management and tree plantation, involving intensive labour input, are part of the local innovativee initiatives. What also strikes most researchers is the innovative and dynamic land tenuree system that favours rather than constrains soil fertility management. Although the existingg land tenure prohibits users other than the owners from planting trees, oil palm

contractscontracts and long-period renting are such schemes geared at maintaining a certain level of

soill fertility. These two access modes offer countervailing arguments to the chronic land deprivationn that might be expected owing to the existing population-land ratios. In such circumstances,, what really matters is the time consumed, travelling between often distant and scatteredd small individual plots. Consequently, being a land-owner or a user makes little differencess if an appropriate use access is ensured. When farmers endeavour to correct

discrepanciess related to information dissemination through the exchange of innovative practicess and even inputs, sensible organisational conditions are taking shape.

Concerningg crop-seeds, on-farm selection as well as a progressive renewal of local geneticc stocks through the adoption of new varieties is a strategic part of the battery of local integrall responses to yield decline. The so-called local varieties are not local as such. Farmers extensivelyy rely upon breeds and crossbreeds of plant material resulting from cross-pollination betweenn locally used varieties and new ones, imported from other areas or from regional and nationall research stations. Local plant material testing and experiments do pursue quantitative yieldd increases, as do their counterparts commonly found in the agricultural research stations. However,, unlike the modern research experiments, empirical evidence demonstrated that yield iss not the overriding goal in the indigenous plant selection process (Richards 1985; 1993). Attributess such as taste, colour, processing time and technology, ease of conservation and the likee are all very relevant to the adoption and the extensive cultivation of most crops at the communityy level (RAMR 1990). This is not to assert that such attributes are not accounted for onn research stations, but the formal selection process is rather guided by one or two attributes thatt may help achieve improved quantitative yields, given other non-controlled attributes. Therefore,, the complexity of relevant attributes encompassed by the natural selection used by farmerss experimenting in the real environment, challenges the formal selection process. More importantly,, the ease of diffusion of innovative seed-selection results derives fundamentally fromm the existing community seed system. It follows that organisational advantages also derivee from the dynamics of the community input system.

Owingg to the agro-ecological conditions, the population-land ratios, the land tenure system andd the like, farmers from each one of the five agro-ecological zones are more or less specialised inn the production of certain crops. Specialisation is also observed at a lower territorial level such ass the villages. However, local specificity coexists with some more general features at the regionall or even the national level. It is the mix of all relevant features that makes a village uniquee with respect to the organisational dimensions of agriculture.

3.22 Recent government intervention in agriculture

Recalll that cotton is the most important crop concerned with government intervention in the agriculturall sector. Government intervention is also addressed to other crops, but to a lesser extentt compared to cotton. Due to some peculiarities of agricultural production, government effortt hinges on a number of context-specific factors to achieve desired outcomes. Some of those underlyingg factors are described below, but their full rationale will certainly take shape throughoutt this study.

Althoughh this study does not aim at addressing the inter-regional issues related to agriculturall production, they are touched upon below, drawing from the perspective of the distributionn of agricultural input services.

3.2.13.2.1 The importance of the Mono-Couffo region within the cotton sector in Benin

Thee overall performance of cotton production indicates a low profile to the Couffo region comparedd to the northern regions of Benin (cf. table 3.3).

Inn comparison to its counterparts of the southern part of the country, on the other hand, the Couffoo region performs better in absolute terms. Therefore, this region ranks fourth behind the

Borgou,Borgou, Zou and Atacora regions for cotton outputs on a yearly basis. Concerning the growth

performance,, the Couffo region has the lowest growth indicator (only 85 per cent between 1985 andd 1996). By the same indicator, the Ouémé region has the best figure (more than 930 per cent,

followedd by Zou (700 per cent), Borgou (242 per cent) and Atacora (226 per cent). The inter-regionall situation of cotton portrays a relatively steady performance for the Couffo region.

Tablee 3.3: Comparative evolution of cotton production (in tons of cotton) in Benin, 1985-97

Years s 1985-86 6 1986-87 7 1987-88 8 1988-89 9 1989-90 0 1990-91 1 1991-92 2 1992-93 3 1993-94 4 1994-95 5 1995-96 6 1996-97 7 1997-98 8 1998-99 9 Borgou--Pendjari i 60058 8 82243 3 44353 3 70000 0 66607 7 87089 9 113407 7 102993 3 175840 0 154058 8 205339 9 257449 9 209344 4 167389 9 Zou-Collines s 11443 3 2360 0 13921 1 18899 9 19066 6 32025 5 32675 5 37091 1 62629 9 70313 3 93116 6 96138 8 100000 0 106429 9 Atacora--Donga a 7538 8 11918 8 4351 1 10683 3 10143 3 11089 9 15242 2 8729 9 19033 3 18533 3 24569 9 38279 9 46926 6 58103 3 Mono-Couffo o 9350 0 12582 2 7101 1 7923 3 6707 7 10152 2 12760 0 11396 6 16625 5 18898 8 17330 0 7854 4 8302 2 6856 6 Ouémé--Plateau u 926 6 2316 6 471 1 1139 9 2136 6 1750 0 3008 8 1371 1 2898 8 4436 6 9551 1 42 2 35 5 Total11 1 89315 5 132729 9 70197 7 108644 4 104659 9 145105 5 177092 2 162000 0 277000 0 266238 8 349915 5 430398 8 386402 2 359331 1 Source:: S O N A P R A

Note:: * Total figures include the *Atlantique' region with around 10 to 15 tons a year.

33 2.2 The importance of cotton production within the Mono-Couffo region

Tablee 3.4 presents the intra-regional situation of cotton production in the Mono-Couffo region. Alll five sub-prefectures selected in the Couffo region carry the best scores, in both relative and absolutee terms, compared to the rest of the Mono-Couffo region. All five sub-prefectures producedd 99 per cent of the regional total for the agricultural season of 1992-93. Although not so noticeable,, from 1992 onwards, the share of the Couffo region declined over time. The rest of thee Mono-Couffo region, on the contrary, increased its shares of the regional production.

Tablee 3.4: Intra-regional production of cotton between 1992 and 1997

Aplahoué é Djakotomey y Dogbo o Klouékanmè è Toviklin n Research h area a Lokossa a Lalo o Athiémé é Bopa a Houéyogbé é Comè è Restt of Mono o Monoo total 1992-93 3 Outputt %total 7021.5 5 1958.9 9 1456.3 3 502.1 1 348.3 3 61.6 6 17.2 2 12.8 8 4.4 4 3.1 1 1993-94 4 Outputt %total 11287.11 99.0 48.9 9 60.1 1 1099 1.0 11396.11 100 9434.8 8 3277.2 2 2254.4 4 873.8 8 510.1 1 56.7 7 19.7 7 13.6 6 5.3 3 3.1 1 16350.33 98.3 226.8 8 48.2 2 1.4 4 .3 3 2755 1.7 16625.33 100 1994-95 5 Outputt %total 11225.6 6 2882.8 8 2036.1 1 1862.3 3 515.4 4 59.4 4 15.3 3 10.8 8 9.9 9 2.7 7 18522.22 98.0 330.0 0 36.9 9 9.0 0 375.9 9 18898.1 1 1.7 7 .2 2 .0 0 2.0 0 100 0 1995-96 6 Output t 9417.8 8 2822.7 7 2020.0 0 1984.7 7 658.2 2 16903.4 4 300.8 8 118.1 1 7.2 2 %total l 54.3 3 16.3 3 11.7 7 11.5 5 3.8 8 97.5 5 1.7 7 .7 7 .0 0 1996-97 7 Output t 11923.6 6 4166.8 8 2337.5 5 2747.3 3 595.7 7 21770.9 9 354.4 4 125.1 1 34.8 8 A 9 9 %total l 53.4 4 18.7 7 10.5 5 12.3 3 2.7 7 97.6 6 1.6 6 .6 6 .2 2 n n 426.1 1 2.5 5 17329.55 100 22314.99 100

Att a lower territorial level, the sub-prefecture of Aplahoué progressively lost its relative weight forr cotton outputs, although it successfully maintained the regional leadership of the Mono-Couffoo region. Indeed, starting with 62 per cent of the regional total in 1992-93, Aplahoué ended upp with 53 per cent in 1996-97, thus losing 9 per cent of its initial share. Concomitantly, the sub-prefecturess of Klouékanmè and Djakotomey increased their shares, whereas those of Dogbo and

ToviklinToviklin decreased theirs.

3.2.33.2.3 Location-specific agro-ecological conditions f or crop husbandry

Concerningg agricultural production figures, areas cropped and crop yields have followed rise

andand fall patterns for about one decade (see graphs 3.1 and 3.2). According to both graphs,

productionn figures appear erratic, with increasing trends for cowpea, groundnut and improved maizee crops.

Forr cassava, the sudden rise from 1995 onwards is unexpected, given the past shrinking trends.. The unfavourable climatic conditions of the late 1990s, on which most blame of agriculturall failure has been put, have not impeded cassava production. More precisely, it could bee claimed that the biological struggle against the mealy bugs, which cause cassava plants to fadee and die gradually, certainly reached some conclusive results. Recall that this pest stroke in thee Mono-Couffo region during the late 1970s, with production losses amounting to more than 500 per cent in certain areas under severe attacks (CARDER-Mono 1980 through 1985). Thereforee the observed production recovery, if sustained, could be acknowledged as a victory of aa collective action between researchers, agricultural extension officers and farmers.

Thee case of improved maize is one example of government intervention in the field of agriculturee originally suspected to track a sustained production growth. To the contrary, however,, improved maize crop is characterised by insignificant absolute values of production, althoughh these show relative growing trends. Farming system researchers, who have been very activee in the promotion of improved maize crops, claimed that impediments, resulting from eitherr the process of innovation generation or that of intervention, block farmers' access to and theirr adoption of improved maize crops (Koudokpon 1991; Koudokpon et Huijsman 1992). An illustrationn of the tentative dissemination of improved maize crops will be documented later on.

Cowpeaa and groundnut crops both depict unsteady trends with high bearings on agriculturall product markets. Unlike cowpea, which faces important storage losses, groundnut lendss itself to different transformations into by-products with high potentials for storage. Therefore,, production issues are more relevant for the promotion of groundnuts than storage issues.. The production of cowpea requires many skills, from the fields to the storage equipment. Itt is instructive to note that several attempts at intensification of the cowpea production have initiallyy met with enthusiasm, but a wide-spread adoption of new varieties and innovative agriculturall practices is still hampered by deficiencies in the existing storage practices (RAMR

1988,1989,1990;; Kossou et Aho 1993).

Regardingg the promotion of groundnut, at least three innovative practices were inconclusivee (RAMR 1987 through 1995). First, in the Plateau Aja and the Lonldy Savannah zones,, production response to fertiliser application was found unpredictable. This result is commonn for most crops other than cotton and also justifies the government's reluctance to expandd the agricultural input services to those crops. Second, improved varieties of groundnut weree hardly adapted to the rusticity of local conditions, implying losses due to ants and other groundd pests. Last, new varieties were out-competed by local rustic ones in their performance withh respect to agro-processing and consumer quality requirements.

AA more detailed situation for each crop is presented in table 1 (cf. annexe C), based on statisticss of cultivated areas and yields from 1985 through 1997 in the Mono-Couffo region.

w w oo o en n -Maize e -Cassava a 10 0 OS S CD D oo o CD D oo o oo o en n on n oo o en n en n _{en n} CM M en n en n COO * enn en enn en in n en n en n CO O en n en n r--cn n en n Years s

Graphh 3.1: Evolution of production figures for two food-crops, maize and cassava, between 1985 and 1997 inn the Mono-Couffo region

-- Cowpea -- Groundnut Improved d Maize e m m oo o en n oo o en n oo o en n eo o oo o en n en n oo o en n o o en n en n ,_ _ en n en n ears s CM M en n en n i" i" m m en n en n T" " ^--cn n en n *» » in n cn n _{en n} *" " co o cn n cn n h» » cn n en n

Graphh 3.2: Evolution of production figures for three crops, bean, groundnut and improved maize, between 19855 and 1997 in the Mono-Couffo region

Maizee crop: Maize remains the staple food-crop of village communities in the Couffo region, as itt is for the populations of the south of Benin. This is clearly shown through the absolute figures forr the cultivated areas (cf. table 1, annexe C). Given the rising population figures, the cultivated areass under maize crops are expected to escalate significantly. To the contrary, however, farmers hardlyy increased the absolute land area for such a crop from 1985 through 1997. The promotion off improved maize crop only helped to compensate the declining trends observed for the local varieties.. Therefore, the overall cultivated area for maize crop stagnated during the period of concern.. So did yields under traditional practices. The Couffo region is originally categorised as Self-sufficientt in maize production, though serious shortage crises occurred in years when farmerss inadvertently sold out their reserves. The latter situation is more acute in the south of the Couffoo region (Dogbo, Toviklin and Djakotomey) than in the north (Aplahoué and Klouékanmey). .

Differencess between the south and the north of the Couffo region also lie in the characteristicss ascribed to the subsequent agro-ecological zones. The LonMy Savannah in the northh is more fertile than the Plateaux zone in the south, where farm-size is the lowest when comparedd to the former zone. Consequently, yield differentials under traditional practices are high.. In addition, in the north, the relative availability of land resources is favourable for extensivee cultivation. The north of the Couffo region has competitive advantages in the productionn of maize crop compared to the south, because of favourable agro-ecological conditions. .

Regardingg maize husbandry, rustic local varieties are better adapted to local conditions comparedd to their improved counterparts. More importantly, yield differentials between improvedd and local varieties hardly match the costs incurred for both agricultural inputs and the additionall labour (RAMR, 1988 through 1995). This, amongst other conditions, probably explainss the low profile of improved varieties of maize within the local production systems. Cassavaa crop: The situation of cassava, another staple crop, stagnated in the past. However, bothh cultivated areas and yields significantly improved lately (cf. table 1, annexe C). Cassava productionn constitutes some bulky raw materials that enter the local cottage-food industry and helpp sustain people* subsistence and market integration. In certain villages of the Plateau Aja, specialisationn in agricultural processing is a solution to slack labour, while in other villages cassavaa output is transacted as raw materials. This is at thee core of the complementary relations betweenn different agro-ecological zones and more specifically between the north and the south off the Couffo region.

Onn the input side, cassava production requires the least agricultural input service comparedd to other crops. Planting materials are the only input in use and are essentially managed att community level. Exceptionally though, in the case of new promising cassava varieties, plantingg materials are provided by the intervention services.

Cowpeaa crop: Cowpea is a food-crop which lends itself to market transactions, the same as maizee and cassava. Similar to maize crop, cultivated areas under cowpea stagnated between 19855 and 1997 (cf. table 1, annexe C). However, yields have significantly improved lately, althoughh achievements may be considered to be far behind expectations. The introduction of improvedd varieties and the failure of pest control by farmers were blamed for the low yield averagess observed. Indeed, in certain villages with specialisation, yields are selectively higher thann in others with much less enthusiasm for improved practices. Because of intensive labour requirements,, the south of the Couffo region appears more appropriate for cowpea than the

LonJdyLonJdy Savannah zone, given a similar level of input use. However, farmers from the north, who

Unexpectedly,, though, labour intensity takes the lead, and the southern villages are more specialisedd in the production of cowpea than those of the north. This is to assert that the agriculturall input services will address more cowpea-related production issues in the south than inn the north of the Couffo region.

Groundnutt and cotton crops: Groundnut and cotton are two cash crops. In addition, groundnut iss at the heart of a dynamic cottage-food industry, which supplies high-valued locally made vegetablee oils and other products to most urban areas in Benin. Moreover, groundnut by-productss store well, command a ready market for cash, and can be used as an additional inducementt for its extensive cultivation. If groundnut crop is well rooted in the local cottage-foodd industry, cotton crop, as the raw input for textile, is entirely sold out. As noted earlier for mostt food-crops, figures for cultivated areas under groundnut stagnated from 1985 through 1997 (cf.. table 1, annexe C). Likewise, yields stabilised at a level beneath 0.8 ton /Ha. For cotton, on thee contrary, cultivated areas significantly increased with a ratio of one to three during the same period.. This is not, however, the case for cotton yields, which seriously declined during the last fourr years.

Thee expansion of cultivated lands under cotton crop and the implications for the labour demandd are at the core of the stagnation observed for most food-crops and even groundnut (cf. tablee 1, annexe C). As observed earlier, figures for cultivated areas under cotton escalated at the expensee of the other crops. This suggests either a limited labour capacity or the alternative use of slackk labour for goals other than agricultural production. Both aspects of the management of labourr resources have been touched upon earlier. For now, what emerges from the conflicting resultss achieved for cotton and most other crops is that production figures may improve with a betterr management of the physical resources, land and labour, and appropriate knowledge of modernn agricultural practices. Indeed, yields for most crops are still at the lowest level compared too what has been achieved in Asia, suggesting the possibilities of improvement (Dalrymphe and Srivastavaa 1994). This is also thee case for cotton output quality, which drastically declined for thee sake of the expansion of cultivated lands.

Thee figures in table 1 (cf. annexe C) give a general configuration of the Couffo region with respectt to various crops. There is no doubt that what appears characteristic of the study area still enclosess specific variability from one sub-prefecture to the other, one commune to the other, one villagee to the other, and so on. Agro-ecological variability exists between the north and the south,, and this may justify the access to and the use of agricultural inputs. Village communities inn the south of the Couffo region organise themselves to perform labour-intensive agro-processingg activities, whereas farmers in the north emphasise agricultural production-based labourr organisation (cf. table 2, annexe C). This is not to assert that occupation in agro-processingg is almost absent in the north. More importantly, agricultural produce from the north iss not only exported outside the Couffo region, but it is also sold to the south, as input for the locall cottage-food industry. For the purpose of specialisation, however, certain crops (such as tomatoes,, groundnuts, soya bean and the like) are grown in the south. Therefore, self-sufficiency inn crop production hinges on the characteristics of the crops and on those of the agro-ecological zones. .

3.2.43.2.4 Government failure to promote unbiased agricultural input services

Tablee 3 (cf. annexe C) strikingly depicts the failure of government intervention in the promotion off agricultural input services for improved varieties of maize. The evaluation report of the

integratedintegrated rural development project of the Mono-Couffo region (PDRIM) concluded to the

absencee of intensification in the agriculture sector (CARDER-Mono 1997, p. 15). Recall that thiss projectt was the one in charge of the promotion of alleged new promising improved varieties

off maize. After more than five years, however, the consumption of agricultural inputs resulted inn declining trends between 1992 and 1995. The quantities of unproved maize-seed grown by farmerss dropped from 17.747 tons in 1992 to 3.503 tons in 1995. Similarly, figures for improved cowpeaa and groundnut seeds grown dropped from 8.692 tons in 1993 to .0 ton in 1995. Concomitantly,, the quantity of fertilisers purchased in 1993 was 27 tons, whereas in 1995 it was onlyy 1.5 tons.

Thee dynamic features of agriculture in different agro-ecological zones described earlier contrastt to a great extent with the inconclusive intensification attempts readily documented. The questionn is whether people's reluctance to abide by the agriculture intensification rules set by the public-sectorr intervening organisations explains the unsteady production figures for most importantt crops in the Couffo region. As logically derived from the illustrative cases cited earlier,, technical as well as marketing inefficiencies may hamper most intensification attempts. But,, do technical problems necessarily appeal to technical solutions? The scarcity of ready-made solutionss to the issues impeding the agriculture sector, or, put in another way, the lack of adequatee inputs to meet farmers' needs, has been amply documented elsewhere (Daane and Mongboo 1991; Dèdèhouanou 1994; Dèdèhouanou et Zannou 1997; Mongbo 1994; 1995). For instance,, RAMR (1987 through 1995), relying upon the farming system approach, had to contendd with agricultural inputs destined to cotton for its experiments with, and promotion of improvedd varieties of maize, cowpea and groundnuts within the research area. It is instructive to notee that the real issues for crops other than cotton are related to the distribution of crop-specific agriculturall inputs.

3.33 Government-initiated farmers' organisations and agricultural input services

Inn addition to the failure of the government intervention schemes to bring about improvement of thee conditions in which agriculture has performed so far, the proposed organisational reforms seemm to fall short of the existing social, cultural and organisational potentials of rural people. In orderr to mobilise production factors and organise output markets, for instance, farmers' organisationss have been advocated at the village level. Although the process started long ago, beforee the reforms, people's participation in these organisations only increased to not more than 322 per cent by 1996 (CARDER-Mono 1997). To date, the level of people's membership has improvedd to some extent in certain villages while it is losing ground in others. Of 431 farmers* organisationss found in 1996, only 39 percent adopted their group charter and 12 percent formally enrolled.. More importantly, their village-level distribution is very skewed. Furthermore, as has becomee clear from the performance of the intervention schemes reported earlier, farmers' organisationss are more engaged in the marketing of cotton produce than in the distribution of inputss for intensification purposes.

Apartt from the formal organisations in charge of agricultural input services at the village level,, there are local organisations entrusted with the same institutional goals. The latter organisations,, though relevant to the understanding of the institutional contradictions that hamperr the intervention processes, have been extensively documented in chapter 2. For now, a snapshott of the formal organisations will be pursued below.

Farmers'' organisations (GVs) in charge of the distribution of agricultural input services Tablee 3.5 provides a snapshot of the intra-regional situation of farmers' organisations, the GVs. Recalll that GV stands for farmers' organisations actively initiated by the government sector to undertakee activities within the cotton scheme. As for cotton production, the Couffo region is welll endowed in formal organisations compared to the rest of the Mono-Couffo region. This is

veryy characteristic of the cropping pattern that perfectly integrates cotton production; henceforth farmers'' involvement in the government- sponsored distribution of agricultural input services.

Tablee 3.5: Total number of farmers' organisations (GVs) and their status in the Mono-Couffo region, 1995/1996

Sub-prefectures s Aplahoué é Djakotomey y Dogbo o Klouékanmè è Toviklin n Researchh area Lokossa a Lalo o Athiémé é Bopa a Houéyogbé é Grand-Popo o Come e Totall number of Organisations s 83 3 75 5 61 1 60 0 152 2 431 1 Restt of Mono-Couffo Mono-Couffoo total 23 3 47 7 14 4 10 0 51 1 50 0 43 3

Organisationss with group charter Number r 41 1 30 0 38 8 25 5 35 5 169 9 238 8 669 9 14 4 13 3 08 8 04 4 30 0 20 0 15 5 Percentt of total 49.4 4 40 0 62.3 3 41.7 7 23.0 0 39.2 2 104 4 273 3 60.9 9 27.7 7 57.1 1 40 0 58.8 8 40 0 34.9 9

Organisationss already registered Number r 13 3 08 8 19 9 11 1 0 0 51 1 43.7 7 40.8 8 01 1 0 0 0 0 0 0 0 0 0 0 0 0 Percentt of total 15.7 7 10.7 7 34.5 5 18.3 3 0 0 11.8 8 52 2 .1 1 0 0 0 0 0 0 0 0 0 0 0 0 7.8 8 Source:: CARDER-Mono, 1996

Note:: GV stands for 'Groupement villageois' or village-level farmers' organisation

Inn connection to the adoption of the GV-charter, however, the Couffo region has a lower profilee compared to the rest of Mono. This is mainly due to the sub-prefecture of Toviklin, which scoress high for the number of GVs (around 22.7 per cent of the regional total of farmers organisations)) and very low for those with a group charter (23 per cent of the total at the sub-prefecturee level). With respect to the formal registration of the GVs, on the other hand, the overalll profile of the region is strikingly scanty (7.8 per cent of the regional total). However, suchh a low profile is more prominent in the rest of Mono (.4 per cent) compared to the Couffo regionn (11.8). * i_ ** n w

Althoughh table 3.5 depicts a cross-sectional situation of the GVs in the Mono-Courro regionn for only one growing season, it readily portrays the relative importance of each area or zonee according to an evolutionary perspective.2 It was pointed out earlier that an evolutionary processs corresponds to trial-and-error mechanisms.

Apartt from the formal organisations in charge of the agricultural input market, some

parallelparallel channels are very active in the Couffo region. These channels will be extensively

definedd later. Considering the alleged growth of these parallel channels, it seems relevant to questionn the appropriateness of the formal organisations for input distribution.

Regardingg the institutional goals, these may certainly be better achieved through a comparativee approach in which formal organisations will be assessed alongside with the existing

parallelparallel channels. Due to the volatility of the latter channels in the face of restrictive policy

measures,, however, a comparative assessment is unlikely. What remains is an intermediate processs assessment based on people's perception of both informal and the parallel channels.

3.44 Major issues influencing the distribution of agricultural input services

Thiss section will address the conditions under which the access to and the use of agricultural inputss may be hampered or facilitated. It endeavours to inquire into how policy recommendationss may be broadened to ease reforms, or, put in another way, how agricultural policyy may create opportunities for or threaten village-level response to the distribution of agriculturall input services. Accordingly, it is organised in four sub-sections. First, people's purchasingg power will be examined in the light of the existing input-output relationships. Next,, the extent to which knowledge influences the input-output relationships will be documented.. Then, input-output relationships will be discussed from different perspectives. Andd finally, the need for extension services will be substantiated, followed by the concluding comments. .

3.4.13.4.1 Agricultural input services and people's purchasing power

Inn general, the extent to which people make use of agricultural inputs hinges on their purchasingg power. Regarding their access to inputs outside the cotton scheme, people's purchasingg power is certainly a barrier. In Benin, household surveys substantiated that direct investmentss in agriculture are very rare (Czesnik et al. 1992, Agossa et al. 1997). For instance, agriculturall investments do not figure within the numerous expenditure categories that make individuall household portfolios. However, it should be stressed that this result departs from thee growing parallel channel of the input market. Recall that the parallel channel of the input markett concerns non-certified inputs as well as selling points. Owing to the reforms, people's reliancee on these channels rather expands nowadays. This applies with greater significance to thee south when compared to the north of the Couffo region.

Thee rationale for household surveys not accounting for agricultural investments may probablyy lie in the neglect of this investment area during the data-collection process. Agossa ett al. (1997, p. 212) draw our attention to the respondents' own neglect of these expenditures, whichh is in contradiction with the relatively high weight attributed to both credit and wages as constraintss to non-poor rural households. Although quantifiable data are missing, this suggests thatt non-poor rural households substantially invest in agriculture within the limits of their purchasingg power. Another reason for neglect may derive from people's perception of agriculturall input expenditures in relation to household portfolios. Most agricultural investmentss are either in kind or insignificant with respect to other expenditures incurred by rurall households. A further reason lies in that no provision is made for a household's intermediatee consumption of goods or services.

AA significant part of household investments in agriculture consists of family labour. Withoutt an adequate valuation of labour resources, households will seldom put value on agriculturall investments. Family-labour investments in agriculture are higher in the north than inn the south of the Couffo region.

3.4.23.4.2 The relevance of knowledge to the agriculture input services

Inn the production sector in general, there is always a reference to factors or inputs by which performancee goals are achieved. In the agricultural sector, for instance, traditional factors are land,, labour and capital. In economics, however, management is often considered as a fourth input.. Owing to such a broader perspective of production factors, the creation, diffusion and applicationn of knowledge are crucial for furthering agricultural development. In this respect, thee concept of resource, as extensively discussed earlier, applies more than that of input. It shouldd be stressed from the outset that the agricultural policy reforms do not integrate knowledgee in order to boost production. For instance, reforms stress the appropriation of the

distributionn of agricultural input services by farmers, but elude the supportive function of the existingg extension services. The success of the farmers' appropriation of the service may be location-specific.. It is expected that the north of the Couffo region, which experienced more intensivee extension services, performed better than the south.

Inn the absence of a full-fledged extension service and a dynamic agricultural research service,, the agricultural policy reforms may not generate any knowledge. The rationale underlyingg such a claim derives from the cotton bias initially involved with the formal agriculturall input market. It is suspected that discarding the extension service from the cotton sectorr would put an end to the diffusion of agricultural knowledge. Although cotton scores behindd maize with respect to both land allocation and output figures, the extension service dependss on financial resources derived from the cotton sector (cf. table 3.5). Two other reasonss may justify the need for extension service, the increasingly large proportions of inputs deviatedd to uses outside the cotton scheme, and the development function entrusted to it.

343343 Input-output relationships in the agricultural sector

Ann assessment of the distribution of rural economic services cannot dispense of the nature of thee input-output relationships. In fact, efficiency and effectiveness can only be inferred from thee different trade-offs that are needed for the distribution process to be either cost economisingg or goal-effective. In this respect, several issues are at stake.

Input-outputt relationships in the agricultural sector must be envisaged with the perspectivee that a pure technical assessment is not sufficient. One reason derives from the influencee that farmers' characteristics in conjunction with policy measures have on input use, thee so-called Institutional Environment. Recall that 'institutional environment' refers to the ruless of the game that are set by the government as well as by farmers. Owing to such an influence,, the utilisation of agricultural inputs must be considered as a performance variable inn its own right, as people's participation in the agricultural input services. A second reason evolvess from the fact that several attempts to achieve results, obtained on research stations, havee met with disillusion in real conditions. Most on-farm experiments of new inputs demonstratedd large discrepancies between yields in controlled and in real-farm environments (RAMRR 1988 through 1995). Specificity in the physical and natural environment of both researchh stations and small farms is blamed for not being favourable for the replication of excellentt results from controlled conditions. But, other factors still play large roles in the disappointingg farm-level results. For instance, complementary inputs as well as catalyst infrastructuress and services are missing for an objective evaluation of the input-output relationships.. More importantly, the divisible nature of the input package also raises the complexityy of the assessment.3 Regardless of those limitations, numerous attempts to evaluate input-outputt relationships on farmers' fields advocate effective extension services to bridge thee gap between results from research stations and those on real small family farms.

Agronomistss analyse input-output relationships with reference to quantitative figures of production,, given certain levels of inputs. In this respect, referral figures are used.to^both inputt and output. Examples of on-farm experiments carried out from 1987 through 1995 will bee documented below. Farmers, under a close supervision of researchers, have conducted on-farmm experiments. The subsequent results are believed to reflect the farmers' context of production,, including the varying agro-ecological conditions of small family farms.

On-farmm experiments carried out in the Couffo region demonstrated that, under harsh conditions,, local production practices perform far better compared to improved packages (RAMRR 1987 through 1995). However, it must be stressed that certain inputs are more relevantt than others to some zones. Because of soil-fertility decline, a mineral fertiliser is relevantt to agriculture in the LonJdy Savannah zone and the Plateaux zone. Yet, zero response

too mineral fertiliser was found in certain soils of Toviklin in the Plateau zone, owing to pronouncedd soil degradation (RAMR 1988). Likewise, the application of insecticides outside thee cotton scheme is also relevant in certain villages with vegetable crops such as cowpea. Farmerss turn to the use of insecticides in the presence of pest attacks on vegetal materials, the intensityy of this use then being proportionate to the degree of attack. Regarding seeds, on-farm experimentss conclude to the rusticity of local clones compared to improved varieties. For instance,, Kpodji-Guèguè and Botogbohoungbé are locally adapted clones of cowpea crops withh upright stalks. Those crops produce two times more compared to improved varieties underr severe agro-ecological conditions (RAMR 1989, 1990). When conditions are favourable,, the difference between local clones and improved varieties hardly justifies the adoptionn of the latter. More importantly, the fact that local clones produce twice as much vegetall material as rival improved varieties also explains their extensive cultivation. Indeed, leavess from the two local clones are incorporated into the local diet. These also constitute the bulkk of organic manure for community soil-fertility regeneration practices (Brouwers 1993). So,, input-output analysis is still valid although it might involve very complex rather than straight-linee relationships. From a straight-line perspective, both local clones are unfit as soon ass they are sold at half the unit price of improved cowpea varieties. The rationale of their adoptionn afterwards derives from their higher use-values for fanners.

Thee economic analyses of input-output relationships impute costs of production other thann family labour, like farm-level depreciation costs. Put another way, attempts at making a valuationn of certain costs may hardly justify the economic rationale underlying farmers' adherencee to certain inputs. For instance, the economic analysis of the cotton scheme demonstratedd that this crop is not economically attractive to most farmers (Biaou et Hoogeveenn 1994). This debate ensued the CFA currency devaluation after prices for imported agriculturall inputs escalated. Suggestions that the national marketing board (SONAPRA) shouldd raise producer prices of cotton from 165 FCFA to 200 FCFA a Kilogram echoed the debatee (Saka Kina 1995). The implementation of the price increase during the growing period 1996-97,, originally expected to boost production, met with mitigated results. Indeed, quantitativee figures escalated, due to extensive cultivation. However, the quality of the fibre drasticallyy declined, questioning the trustworthiness of Benin cotton on the world market (Le Bourgee 1997). Yet, claims that farmers benefit from the price increase remain shaky for severall reasons. One reason is that the increase in prices was followed by a reduction of the collectivee financial resources that usually accrue to villages.4 Another reason derives from the factt that costs are moving targets, which hardly increase at the pace at which political and economicc negotiations evolve. An additional reason derives from the fluctuation of cotton pricess on the world market. The increase of the announced price was followed by a declining worldd market price. A further reason appeals to the difficulty of setting labour costs and assessingg some depreciation costs in the presence of high unemployment and inflated agriculturall labour force. A final reason relates to the prejudice against admitting the full economicc rationale of many forms of valuable knowledge, which are intrinsic to the use and combinationn of inputs.

Fromm the preceding discussion, it could be derived that straight-line input-output relationshipss would forbid the use of agricultural inputs. The production techniques have high bearingss on the distribution of inputs. However, it was stated earlier that only inputs destined too cotton are available and within the reach of farmers through the formal channel Therefore, veryy complex adjustments of input to output justify people's adherence to certain inputs. In thee cotton sector, for instance, farmers substitute intensive labour resources for most other inputs.. This is the only rationale for the increasingly higher production figures and the concomitantlyy lower fibre quality. In other sectors, especially maize, farmers either use