~ANSPORT aid ROAD R.ESEARCH LABORATORY -.—-—.—-——.y. -— LIBRARY Ibmsa@ Accessibility and agricultural development in the Ashanti region of Ghana by J. L. Hine (Transpoh and Road Research Laboratory) J. D. N. Riverson and E. A. Kwakye (Building and Road Research Institute, Ghana) TWSPORT and ROAD RESEARCH LABORATORY Department of the Environment Department of Transport ACCESSIBILITY AND AGWCULTURAL DEWLOP~NT W THE ASHANTI REGION OF GHANA by J L Hine (Transport and Road Research Laboratory) J D N Riverson and E A Kwakye (Buflding and Road Research hstitute, Ghana) The Report is based on a cooperative research study undertaken by the Butiding and Road Research hstitute, Kumasi, Ghana, for the Ghana Highway Authority The work described in this Report forms part of the programme carried out for the OverseasDevelopment Administration, but any views expressed are not necessady those of the Administration OverseasUnit Transport and Road Research Laboratory Crowthorne, Berkshire 1983 ISSN0305 -1315 CONTENTS Abstract 1. htroduction 2. Agriculture in Ashanti Region: the context 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 htroduction An increase in production and sales with no change in farming technology Technical change, labour input and population density Mxed cropping Cocoa 2.5.1 The development of the cocoa industry 2.5.2 Cocoa marketing 2.5.3 Pests and diseases 2.5.4 New varieties 2.5.5 Cocoa production and prices Relative prices and choice of crops The role of extension services 2.7.1 Cocoa Production Division 2.7.2 Ashanti Cocoa Project 2.7.3 Department of Agricultural Crops Extension Division 2.7.4 Other extension organisations The availability of finance and modem inputs Marketing 3. The structure and organisation of the study 3.1 The anrdysisframework 3.2 Accessibtity in Ashanti Region 3.2.1 The dominant position of Kumasi 3.2.2 The measures of accessibility used in the analysis 3.3 Development parameters 3.4 Control factors , 3.4.1 Terrain 3.4.2 Population density 3.4.3 Sofl characteristics 3.4.4 RainfW 3.4.5 Crop diseases 3.5 The survey Wages 3.6 Data co~ection 3.6.1 The main questionnaires 3.6.2 The tilage survey 3.6.3 Sofl samples Page 1 1 2 2 3 3 4 5 5 5 5 6 6 6 7 7 7 7 7 8 8 9 9 9 9 9 9 10 10 10 10 10 10 10 11 11 11 11 Page 4. 5. 6. 7. 8. 9. 3.6.4 me suwey of extension organisations 3.6.5 Other data 3.7 Data analysis Suwey results 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 Population and SONSin the suwey area &nerd holder characteristics hbour input Modern inputs Holder finance Cocoa production Animal husbandry Food production yields Food crop sales Rotten produce and accessibtity Factors affecting the expansion of production Accessibfity transport costs and prices Discussion Achowledgements References Appendix: VWagesuwey data @CROWN COPYRIGHT 1983 11 11 12 12 12 13 13 15 16 17 19 19 20 20 21 22 23 23 23 27 Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged ACCESSIBILITY AND AGWCULTML DE~LOP~NT ~ T~ ASHANTI ~GION OF GHANA ABSTWCT The report examines the relationship between agricultural development and accessibility in the Ashanti Region of Ghana. A wide variety of factors are identified that can influence agricultural development in the Region and some of the problems of its measurement are hi~ghted. Using a cross sectional framework of analysis, data was co~ected from 33 Wages (W but two with vehicle access) in the Ashanti Region of Ghana located between 8 and 102 km from the Regional Capital, Kumasi. By comparing a number of development parameters and the transport costs of moving farm produce between each Wage and Kumasi (and dso between each tiage and its respective district centre) the fink between accessibility and agricultural development was investigated. Within the range of accessibdity considered little evidence was found to indicate that market agriculture was promoted directly by accessibility. However, loan finance was easier to obtain the nearer the farmer lived to Kumasi. OverW there is evidence to suggestthat the most accessibleWages tended to concentrate more on non agricultural activities (such as rural industry and the provision of services,including marketing) wtie the less accessible tilages concentrated rather more on agriculture. The study supports the view that where road investment can induce only a sma~ change in transport costs then little impact on agricultural development may be expected. 1. ~TRODUCTION The planning of rural road investment in developing countries can be improved by an understanding of how that investment may influence agricultural development, and subsequently rural development in general. The Buflding and Road Research hstitute in Kumasi (Ghana) and the OverseasUnit of the TRRL (UK) have co~aborated in a study of the impact of feeder road investment in the Ashanti Region of Ghana; the work was undertaken for the Ghana Highway Authority and partia~y funded by the World Bankl. Road projects are most usudy justified on the basis of the forecast savingsin transport costs gained by road users. ~st ttis is widely accepted as adequate for road investment which caters for inter-urban traffic it is believed that transport cost savingscan only partia~y reflect the development benefits which may arise from improved communications to rural areas. Mthough a number of case studies have been carried out in different developing countries on the relationship between development and road investment it has not been possible to genertise satisfactorily from their results2. In consequence the ability to predict the effect of road investment on rural development is limited. It is against this background that the study was conceived. It was recognised that the best chance of obtaining usable relationships between accessibility and development in the time scale avadable would be to collect field data from viUageslocated at varying distances from a major urban centre. In fact sample data was collected from 33 vmagesin &hanti Region located between 8 and 102 km from the Regional capital, Kumasi. 1 This report mdyses the influence of accessibility on rural development by comparing a number of parameters of rural development with the transport costs of moving farm produce between each Wage and Kumasi, or between a Wage and its respective district centre. The main emphasis of this report is on the relationship between accessibtity and agricultural production. The relationships between accessibfity, transport costs and marketing are considered elsewhere3. 2. AGMCUT~ ~ AS~I REGION: T~ CONTEXT 2.1 Introduction This section discussesthe principal factors that determine and influence agricultural change in the survey area. Its purpose is to provide a context for evduattig the influence of accessibfity on development in relation to the particular conditions in Azhanti Region at the present time. The survey itself is described in Section 3. Many of the major institutions and most of the basic communications infrastructure of the Region have been in place since the 1950s. bring the past twenty years the pattern of rural economic activity has tended to be rather static with some dectine in the important cocoa sector. A description of the data collection and on analysis of results are included in later chapters. Nthough there is some measure of agreement as to what constitutes agricultural development the issue h not unambiguous. Most would accept that high yields, the use of new inputs like fertilizer and improved seeds, and the growth of market agriculture are reasonable indicators of agricultural development. Nevertheless,there is much less agreement between agriculturalists as to what practices should be advocated by the extension service, whetier land should be switched between growing a crop for export and growing a different crop for the local market or what farming practices are best for the farmer and the country. These disagreements stem largely from four problem areas. These are domestic labour input, farmer risks, the long term availability of modern inputs and relative domestic and international prices. Taking each in turn. i) ii) iii) iv) hbour input Farmers are interested in getting the best return from their own labour. They W be natur~y reluctant to undertake new agrictiturd practices which, even though yields may be increased, W nevertheless demand a disproportionate increase in their own labour effort. Farmer risks Farmers W tend to be reluctant to expose themselves to greater risks of substantial crop losses even though ‘on average’they may be better off by adopting a given change in farming practice. The avaflabtiity of modem inputs Many new farming practices are based on modem inputs. If the farmer is to adopt a new practice then he must have confidence that the modem inputs@ be avtiable when he wants them in the longer term. Relative domestic and international prices. It is possible for relative international and domestic-prices to move so far out of he that a farmer would be financidy better off by growing a food crop on his land rather than an export crop like cocoa even though it wodd be better from the point of view of the national economy if the reverse was the case. 2 ~ese issues are considered in greater detati below. 2.2 An increase in production and sales with no change in faming technolo~ Perhaps the most straightforward way for farmers to respond to better accessibfity is by simply employing more inputs to increase production. me relative use of inputs @ remain basicdy the same and the method of cultivation WMbe unchanged. If the better accessibfity lowers transport tariffs to move produce to market then the farmer W gain an effective rise in his farm gate prices. me rise in farm gate prices W then make it profitable to employ more labour to increase the land under cultivation and so a rise in production W result. Bateman4 has calculated a short term price elasticity of 0.22 for cocoa production in Ashanti and Brong Ahafo Regions of Ghana, which means that a one per cent rise in cocoa prices (in red terms) would induce a 0.22 per cent rise in production. However, using Nigerian data Sterns has crdcdated a supply price elasticity of 1.29 for cocoa acreage which does suggest a much larger response. A study by 0ury6 of the supply elasticities for a number of crops in different developing countries gives a range of between Oand 1.5. me range of supply elasticities is to be expected in view of the differences in resource endowment, population density and cultural background found in areas covered by the studies. If the costs of agricultural labour and other inputs dso dectine as a result of the lower transport costs then the farmer ~ find it profitable to employ even more inputs to expand production. Better accessibility W not necessady have these effects for an uncompetitive transport industry may prevent tariff reductions from taking place. Ukewise labour costs may actutiy rise as a result of improved accessibility by enabling agricultural labour to seek better paid employment elsewhere. Camemark, Biderman and Bovet7 have developed a model to cover a range of situations (eg. increased domestic consumption, increase in cropped area, crop substitution and regional deficits) that can be used to predict possible changes which might occur with reduced transport costs. Using their framework of analysis the most critical variables that are needed in order to predict a rise ~ production foflowing a road investment can be identified as foflows:- i) the absolute change in the farmers’ farm gate price ii) the farmers’ price elasticity of supply iii) the change in labour prices iv) the change in other (non labour) input prices. Obviously tie greater the increase in farm gate prices, the greater the reduction in input prices and the more elastic the farming supply curve the bigger the increase in production that @ occur fo~owing road investment. 2.3 Technical change, Iabour input and population densi~ me current pattern of food farming that is practised in the forest zone of Ashanti Region is based on shifting cultivation. ~s is a pattern of farming whereby a piece of land is cropped for up to three years and then left to bush ftiow for up to ten years to regenerate the fertflity of the sofl. men the area is to be cleared again the land is cleared by fire, the large trees and tree stumps are left in the ground and cultivation is carried out with a hand hoe. 3 Food farming in the forest zone of Ahanti Region is, therefore, characterised by virtudy no modern inputs and low total yields. Boserup8 has suggested that the reason shifting cultivation persists, rdong with other agricultural practises givinglow totrd yields is because, with existing population density these practises meet farmers’ needs at least effort. She puts forward the view that technical change in agriculture has taken place principMy in response to population pressure, usudy over a time span of centuries. h tradition societies, for a given level of population pressure, a farming technology is adopted which@ meet basic needs with the minimum of labour input. It is argued that with each change in tradition farming techolo~, moving from forest fdow through the stages of bush fallow, short fdow, annual cropping to multicropping, there are diminishing returns from extra labour input. To obtain increased yields from a given land area it is necessary not only to change farming technology but dso to increase disproportionately the labour input. However, with high population density the scarcity of land is such that intensive methods of cultivation are forced upon the local population even if this does mean a much increased My input of labour per person. From this point of view it is easy to understand some of the reluctance of farmers to respond to advice to adopt new farming practices which may increase farming yields per acre. From the farmers’ point of view it may we~ be an inefficient use of their labour to attempt to adopt intensive methods of farming to increase yields from a sma~ farmed area when less effort applied over a larger farmed area may produce more output. Boserup recognises that with the use of modern labour-saving technical inputs (such as tractors, irrigation pumps, insecticides etc.) the marginal productivity of Iabour may weflincrease with more intensive methods of cultivation, rather than decrease as with the more tradition changes in agricultural technology. Thus, farmers may be wfing to adopt more intensive methods of cultivation provided they can obtain the new inputs cheaply enough. However, the adoption of many aspects of new technology til not necessartiy represent the most efficient use of resources. It is now widely recognised that farmers are aware of a range of agricultural practises but they @ not adopt more intensive methods of cultivation udess population pressure or widely avaflable cheap modern inputs encourages them to do so. Nthough Boserup’s thesis maybe useful in making broad comparisons between different farming areas and different countries it may nevertheless be criticised for paying insufficient attention to differences within farming areas relating to factors such asland tenure, chate, terrain, sofl fertflity, available water or the presence of urban areas. M of these factors are likely to lead to local relative specialisation, trade and the adoption of different farming methods. h practice, of course, a variety of farming technology wdl be adopted within any farming area. 2.4 Mtied cropping Throughout the forest area of khanti Region food is usu~y grown in crop mixtures. Most often three or four dominant crops are grown together with a number of other crops thinly scattered throughout the plot. TypicWy, maize, cassavaand cocoyam are grown as dominant crops interspersed with beans, tomatoes, plantains and yams, but many combinations of these plants are found. h its early stages, cocoa is often grown mixed with food crops, particularly plantain and cocoyam. There are different opinions on the wisdom of mixed cropping 9‘1‘. The evidence on achievable yields per unit area and on returns to Iabour input is conflicting. On balance it seems that mixed cropping has advantages for smd scale farming where no modern inputs are used but “pure stands” are more suitable when hired labour, mechanicrd cultivation and chemical inputs are employed. The advantages claimed for mixed cropping (compared with growing crops in singe stands) are that it lessens the chance of complete crop faflure, it lessens dmage from pests and diseases and it is more effective in conserving sofl fertflity because different crops have different sofl nutrient demands. Mixed cropping can dso be labour saving to the smti sde fam’er because it is possible to undertake several operations such as weeding, harvesting and planting during the same visit to the farm. The disadvantages of mixed cropping become apparent when modern inputs are introduced and a more commercial approach to farming is adopted, as mixed cropping does not lend itself to mechanical cultivation. The use of selective chemical inputs (such as ferttier, weedicides and pesticides) to meet the requirements of particular crops is impractical with mixed crops. To make the best use of chernicd inputs the plant densities of the target crop have to be usu~y at such a high level that interplanting with other crops is not practicrd. With more commercial agriculture the economics of using tired Iabour for harvesting at one time are such that again pure cropping is demanded. To summarise it would seem that shifting cultivation and mixed cropping are an efficient and rational method of producing food by farntiy labour on smti farms. The deficiencies of farming by this method only re~y become apparent when modem inputs and hired labour are introduced on a large scale. , 2.5 Cocoa 2.5.1 me development of the cocoa indus~. The most important development in Ghanaian agriculture was the establishment and growth of the cocoa industry. Cocoa plays a key role in the economy accounting for 60 per cent of Ghana’s export earnings and a third of the Government’s revenues11. Cocoa growing is concentrated in the South of the country and is particularly important in Ahanti Region. Forty per cent of Ghana’s cocoa acreage is planted here and approximately three quarters of the cultivated land in the Region is under cocoal 2. Cocoa growing was developed in Ghana during the 1880s – 1890s by indigenous entrepreneurs who responded to the high prices offered by international traders at the coast. Cocoa was first established in the Nwapim area before the widespread use of motor vehicles. Cocoa growing rapidy spread west and northwest as farmers bought up lands with profits made from existing cocoa farms. Before the end of the 19th century cocoa was established in Eastern khanti. Its spread through the Region was helped by the expansion of the road and rd network in the 1920’s and 1930’s. In the light of the modern approach to agricultural development which cds for extensive extra-industry support it is interesting to note that cocoa growing developed without government supphes of inputs, capital, extension advice, seeds, or insecticides. The main ingredients for success were a profitable market, suitable land, local entrepreneurship, capital, labour and a source of cocoa seedingsl 3. 2.5.2 Cocoa Marketing. The Ghana Cocoa Marketing Board has become responsible for the local purchasing of cocoa from the farmer through buying posts operating from W but the very smaflesttilages in the cocoa growing areas of the country. h contrast to other crops the farmer is paid a f~ed price at the buying post for his produce. The onward movement of the cocoa is then the responsibility of the Board and the adverse effects of inaccessibtity on the smalfiolder cocoa farmer are minimised. 2.5.3 Pests and diseases. Cocoa production in Ghana has suffered considerably from two main causes, swollen shoot disease and capsids. Swollen shoot disease is a virus infection that is spread from diseased to healthy trees by crawling and wind blown mealybug insects. The disease has decimated cocoa production in large parts of Eastern Region in the past, its only cure is to cut out the diseased plant and completely replant. 5 me farnfly of insects known as capsids are a major pest on cocoa trees throughout the whole of West Africa. When feeding the capsid injects its poisonous safivainto the plant which eventutiy causes the tree to weaken and die out. Control is successfu~y achieved by regular spraying with insecticide. 2.5.4 New varieties. High yielding hybrid varieties of cocoa have been introduced which W yield much earlier in the fife of tree than the original strain of cocoa such as the Arnelonado variety. me hybrid varieties have greater resistance to pests and diseasesand have somewhat higher yields. No fertfliser is recommended with any variety of cocoa. me widespread adoption of new cocoa varieties would obviously require an increase in transport capacity however the net effect on average traffic flows would be ne~igible. 2.5.5 Cocoa production and prices. Cocoa production in Ghana reached an d time high in 1964/65 when 557,000 tons were produced. Since then production has gradudy decfined as old and dying trees have not been replanted on a sufficient scale. ~s is thought to be related to relatively poor producer prices and the untingness of young people to take up cocoa farming. Bateman4 has calculated that the price offered to producers dec~ed in red terms from the early 1950’s onwards to reach a levelin 1968/69 of ody 35 per cent of the 1954 price. ~ring the 1970’s the red price dechne continued but at a slower rate to reach a level of 33 per cent of the 1954 price in 1975/76. Between 1976 and 1980 the timing and rise of cocoa prices offered to farmers has tended to lag behind the price increases of other food crops. 2.6 Relative prices and the choice of crops Expected price is a key factor in the farmer’s decision to grow a certain crop. Given production costs the absolute price is needed to determine whether a crop is worth growing at afl. me relative prices of other crops (and their relative costs of production) are needed to fmd which is the best crop to grow. A problem emerges at the national level when relative domestic prices do not adequately reflect relative international prices. In this case a farmer may be encouraged through price incentives to grow one crop which has less value to the country than another crop. Such a situation appears to have arisen in Ghana in recent years in terms of the relative domestic prices offered for food and cocoa. h much of Ashanti Region (as elsewhere in Southern Ghana) land can be used to grow food or cocoa. h many places farmers have dug up their cocoa farms (or fafled to replace old and poor yielding trees) to grow food in response to the relatively higher food prices, w~st from the country’s point of view it can be argued that it would have been better to concentrate on growing more cocoa and less food. If necessary extra food could be imported with the extra export earnings from the cocoa. h August 1978 the cocoa price paid to farmers was approximately 1.5 times the price they would get for se~ing the same weight of maize at the Kumasi Central Market. At the same time on the international markets in Europe, cocoa was worth 18.5 times more than maize14. ~though the ratios are not strictly comparable because of the relative valuation of production costs*, the costs of transport and the fact that cocoa prices can fluctuate enormously (in January 1982 the international * h order to define precisely the optimum it is necessary to &o consider the domkic and international valuation of the production costs. However, in the forest areas of Ashanti labour is the major criticrd resource to grow both crops and so one may expect that the ratio of the domestic valuation of resources to grow maize relative to cocoa WUbe litde different to the ratio of an international valuation of the same resources. market price of cocoa was 9 times the price of maize) 15, the figures are sufficiently far apart to suggestthat cocoa was relatively so under-vdued during the study period that farming decisions on cocoa were Wely to be far from optimal from Ghana’s national economic point of view. 2.7 me role of extension semices Ashanti Region is suppfied with a wide diversity of extension services and purchasing organisations. Operating in different areas these institutions WMsupply advice, credit and a variety of modem inputs to the farmeF. 2.7.1 Cocoa Boduction Division. me prime role of the Cocoa Production Division is to check the spread of pest and diseasesthat affect cocoa. Each year a proportion of cocoa in the Region is sprayed with insecticide by the Division to prevent damage by capsid attack. me whole of the cocoa growing area of Ghana is covered by the Division. Officers of Cocoa Production Division keep detafled records of any outbreaks of swoflen shoot disease. Once the disease is identified the cocoa plant is cut out and replanted. me Cocoa Production Division keeps cocoa nurseries and WMassist farmers with the planting of new stock. 2.7.2 Ashanti Cocoa hoject. me Ashanti Cocoa Project is an independent externa~y financed organisation, operating principdy in the south of the Region. It has 12 district offices, nearly 30 senior officers and over 400 field staff. Officers of the Project have the task of setting up new cocoa farms in their area of operation. ~ey first identify farmers interested in growing cocoa; they wti then measure the farmer’s land and organise a ‘loan application’ for the farmer. Once this has been agreed the organisation wifi clear the land and plant cocoa seedings and for the initial ‘non bearing’ years they WWdso plant food crops in between the cocoa for the farmer. h many ways the Ashanti Cocoa Project has almost taken over the tradition entrepreneurial role from the farmer for establishing the cocoa farm. A complaint often made by field staff of the Project is that farmers take little interest themselves in the Cocoa farm and they appear wing to let the Project staff do everything to establish the farm. Perhaps this is not surprising. 2.7.3 Deparment of Aw.culture, Oops Extension Division. ~s Division covers the whole of Ashanti Region, operating from 6 district centres with 16 supervising staff and nearly 100 field staff. me Crops Extension Officer provides general agricultural expertise to the farmer; he W supply new seeds and fertdiser and dso help in obtaining offlcid loans. h Ashanti Region it appears that particular attention has been given by Extension Officers to maize growing. ~is is probably because of the development of new high yielding seeds which are responsive to the application of ferttiser. Maize thus provides a good opportunity for extension work. Maize storage chemicrds are dso distributed by the Extension Officer. me Extension Officer helps farmers apply for loans by helping them to organise into a loan co-operative. me loan is granted by one of the commercial Banks to the co-operative in the f~st instance and it is then distributed amongst the co-operatives’ members. 2.7.4 Other ~tension o~anisations. mere area number of other extension organisations operating throughout the Region. me Animal Husbandry Department distributes chicken feed and other animal foodstuffs. mere is a sm~ Veterinary Department that pays particular attention to monitoring the health of sheep and goats in the Region, these being the ody large animals kept in the Region in substantial numbers. 7 me Grains and hgumes Board encourages farmers to grow improved seeds. h Ashanti Region its efforts are concentrated in the Offmso area. It lays out demonstration plots, arranges loans and suppties seeds and fertfliser to farmers. Uke the Crops Extension Division the Board givesparticular attention to maize growing. It also has the role of principrd suppher of improved seeds in Ghana. fie Cotton Development Board and the Bast Fibre Development Board promote and purchase cotton and bast fibre in the Region, both concentrating their activities in the North of the Region. 2.8 me availabili~ of finance and modem inputs Finance and modern inputs are critical to agricdtural development in Ashanti Region. As a result of economic difficulties there has been a shortage of many of the key inputs to agriculture. Both fertfiiser and poultry feed have been particularly scarce. hsecticide for spraying cocoa has been more widely avaflable rdthough locrd shortages coupled with some organisational problems in the Cocoa Production Division have caused some farmers to report difficulties in getting their fields sprayed. One problem with the adoption of new farming technology is that it is often necessary to apply different inputs (eg. new seeds, fertfisers, top dressing) and carry out a number of different procedures in sequence if it is to prove worthwtie. Atsu 16 has shown that if only hdf the recommended practice for growing new maize is carried out then the new measures taken wi~ prove to be an expensive faflure. It is for this reason that farmers do need long term confidence that supphers of modern inputs wfll be avaflable before they wifltake to adopting many of the new recommended practices. Finance is commody claimed by farmers to be the critical factor preventing them from expanding their farms. If the farmer is to undertake large scale changes in his farming then a loan W almost inevitably be required. Small scrde farmers find it difficult to gain the confidence of the official lending agencies and they are often forced to go to unofficial sources for loans at very high rates of interest. 2.9 Marketing me relationships between transport, accessibility and marketing are discussed in another report3. me development role of marketing is outlined herein order to complete the description of the principal factors that influence agricultural development in Ashanti Region. Marketing provides a stimulus to grow more than is required for domestic use, it encourages specialisation in food crop production and it provides the farmer with the cash resources to purchase extra inputs which will in turn help to increase production. mere are risks in specialisation and if the marketing system is costly and inefficient then farmers wfll be reluctant to specialise and produce for the market. An inefficient marketing system can be caused by a poor spread of price information, co~usion between market operators, smafl volumes of produce for sale and a relatively expensive, monopolistic and uncertain transport system. Virtually all small scale farmers in Ashanti Region grow food for domestic consumption and a large majority WWdso sell some of their food although cash is also obtained from sefling cocoa, personal remittances and paid employment. ~flst cocoa is sold at the Cocoa Marketing Board buying posts the majority of food is first purchased by trave~ing wholesalers at the farmer’s house and at local markets. A smaller proportion is sold on the farm or taken’by the farmer direct to the larger central markets in the Region. 8 3. T~ STRUCTURE W ORGANISATION OF T~ ST~Y 3.1 ~e analysis framework There are two distinct ways of identifying the impact of road investment on rural development. One method is to survey an area before and after a road investment is undertaken so that an historical comparison may be made. A second method (the cross sectional approach) is to survey a range of areas with varying degrees of accessibility at the same time. A careful interpretation of the”observed differences and of ‘control’ data is required with both approaches. The cross sectional approach was adopted in this study. 3.2 Accessibility in Ashanti Re~’on 3.2.1 me dominant position of Kumasi. Kumasi is the Regionrd capital of Ashanti Region. It is the.second largest town in Ghana with a population of 400,000 which is over ten times larger than any other town in the Region. The road network of the whole Region and most of the central southern part of Ghana radiates from Kumasi. It is an important rafl terminus and has an airport. Besides being a major market town Kumasi dso has some nationrd headquarters and dl of the regional headquarters of the extension services operating throughout the Region. Even at the fringes of the Region the pu~ of other major towns from outside the Region is comparatively sm~ due to their size and distance from the border of the Region. In view of its central importance the travel costs from anywhere in the Region to Kumasi can be used as a convenient measure of accessibility. 3.2.2 me measures of accessibility used in the analysis. Two principal measures of accessibility were used in the analysis and the relationships quoted in this report relate largely to these measures. These are:- (i) The cost to the farmer of moving one standard headload of produce from Wage to Kumasi. (ii) The cost of moving one standard headoad of produce from tiage to the nearest ~strict Centre. Both ‘costs’were based on charges farmers would have to pay to take their individud loads; wholesale transport charges were not readtiy avafiable for the survey Wages. The costs of moving a headoad of produce from farm (rather than from vtiage) to Kumasi and Dstrict Centre were also used as alternative measures of accessibility to check the viabdity of the conclusions. h practice the four measures of accessibfity were found to correspond fairly closely with each other. 3,3 Development parameters In order to assessthe impact of roads it is necessary to measure rural development, however, no sin~e unambiguous indicator was found suitable for the study. The combination of mixed cropping, subsistence farming, a lack of farming records, high rates of inflation and a wide variation in reported district centre commodity prices au contributed to making it impractical to value total farm output. The view was taken that a whole range of social and economic parameters should be surveyed so that a comprehensive view of the effects of better accessibdity may be assessed.To this end it was recognised that data should be co~ected on farm inputs, outputs and on farming technology as well as data on social characteristics and available social facilities. 9 3.4 Control factors At the outset it was recognised that rural development would dso be influenced by a range of factors which were largely unconnected with road access. It was felt that terrain, population density, sofl characteristics, rainfa~ and crop diseaseswere key factors to be taken into account in the analysis. 3.4.1 Terrain. It was felt extremes of terrain should be avoided when locating the survey Wages so mountainous areas were deliberately excluded from the su~ey. Moderately ro~ing countryside is common throughout most of Aahanti Region. 3.4.2 Population density. Boserup has suggested that population density may have a major influence on farming technology (see Section 2.3). h order to accommodate the anrdysisto this factor, district and wage populations were coflected from official sources and household population was coflected in the survey. District population density, tiage population and household numbers per farm area were used as alternative measures of population density, 3.4.3 Soil characten.sties. me SONSin Ahanti Region can be grouped into two broad types, forest SOUSand savannaSOUS.me forest sods are less suitable for mechanical cultivation than the savanna sofls of the north, however, they are suited to cocoa growing. Sod fertflity WWvary even with an area of.uniform SONtype. It was for this reason that soflsamples were coflected for analysis from each survey Wage. 3.4.4 Rainfall. hnud average rainfall varies across the Region from a high of 1800 mm near Bekwaiin the south of the Region to 1400 mm in the north east of the Region. Nthough rainfall in the north tends to be more season~y concentrated than in the south, there is a great ded of local variability in the rainfall patterns month by month caused by the passage of isolated thunderstorms. me influence of rainf~ and other weather characteristics on cropping patterns and yields can be simplified for the analysis. Cocoa cannot be successfu~y grown in the north because of the longer dry spefl. For food crops fike maize variations in rainf~ above a minimum level til have little effect on yields. It is, however, important to establish that this minimum level of rainfa~ occurred. ~s was confirmed by the Ghana Meteorological ServicesDepartment, Ugon, for the 1979 main crop season. 3.4.5 ~op diseases. Data on crop diseaseswas coflected from the farm surveys. A critical factor which has had an important impact on cocoa growing in Ghana is swollen shoot disease. me impact of the disease was particularly fierce in the Eastern Region. Mthough some past data of crop disease was avaflable it was not known what impact this disease or other diseaseshad on current farming decisions. 3.5 me suwey villages Because the main data co~ection exercise was to be carried out by Ministry of Agriculture statistics enumerators, the choice of survey Wages waslimited to those dages currently part of the Ministry Survey. (me Ministry of Agriculture surveys a random sample of 15 per cent of the smti scale farmers located in a random sample of Wages in the Region.) From the Ministry Survey a sub-sample of 33 dages was chosen which were widely located throughout the main inhabited parts of the Region (see Fig. 1) except that the more mountainous northern and eastern areas of the Region were deliberately excluded as were the remote and uninhabited parts of the Afram Mainsin the far north east of the Region. @eraH the chosen survey Wages were broady representative of the agricultural Wages of the Region. ~irty one of the survey Wages lay in the cocoa growing forest zone and two Wages lay in the savanna zone of the Region. me survey tiages lay between 8 and 102 km by road from Kumasi. 10 3.6 Data collection 3.6.1 me main questionnaires. Ministry enumerators administered two extra questionnaires to their normal sample of smtiolders. This additiond data was co~ected from 491 holders in the 33 selected survey Wages. The first questionnaire covered:- farrn size household composition holder education labour inputs finance crop diseases crop production and sales tivestock use of inputs extension contact farming attitudes and knowledge farm location transport of produce from farm to Wage crop storage and marketing. The second questionnaire was administered somewhat later and coflected more information on farming practice, agricultural transport, migration, c~dren’s education, but its main emphasis was concerned with accessto social facfities and passenger trip frequency and purpose. Additiond data sheets were completed by enumerators covering field sizes, crop mixtures and crop yields, this data having been coflected for the usual Ministry of Agriculture Survey. 3.6.2 me village survey. h addition to data co~ected by Ministry enumerators project staff visited every vfiage to administer a flage survey questionnaire and to coflect son samples. hforrnation for the Wage survey questionnaire was provided by knowledgeable people in the vfiage. This information was further cross checked by visualinspection and reconnaissance during the visit. The tiage survey couected information on fie different occupations found in the tiage, public utfities, social facdities, schools and churches and travel charges to district centres and Kumasi. 3.6.3 Soil samples. Soti samples were co~ected from three farms belonging to sm~olders Uvingin the Wage. On each farm sofl samples were taken from four locations and put together to form a composite.sample. These samples were further tested at the Sod Research Institute, Kwadaso. Tests were carried out for acidity, organic matter content and avaflable phosphorous and potassium in the sofl. 3.6.4 me survey of extension o~anisations. A separate questionnaire was distributed to the eight main extension organisations working in Manti Region. The purpose of the questionnaire was to identify the major transport constraints of these organisations. Questions were asked on the structure of each organisation, methods of farmer contact, materials to be distributed, purchases to be made, transport vehicles avaflable and organisational constraints. 3.6.5 Other data. The Ministry of Agriculture Statistics Department supphed past data on crop yields for the survey Wages and &ta on market prices and transport charges for Ashanti Region. Other data and information was co~ected from the Meteorological Services Department, Ghana Hghway Authority and the Central Bureau of Statistics. 11 3.7 Data analysis The survey data from the holders was aggregated to provide statistics for each individual vflage; the statistics for the 33 vUages are presented in the Appendix. Using this Wage data parameters of accessibility were tested as explanatory variables of the parameters of agricultural development using regression analysis. 4. SURWY MSULTS This section examines in some detafl the results of the analysis of the data from the agricultural surveys. Particular emphasis is given to farm inputs, food production and cocoa. A table of the survey data used in the analysis is shown in the Appendix. The results clearly indicate the etistence of considerable variability in farming patterns between vi~ages.This has added some difficulty in drawing precise conclusions from the analysis. In common with other socioeconomic field studies the bounds of samphg error were wide and it was not possible to control perfectly ford extraneous influences. 4.1 Population and soils in the suney area Regression analysis indicated that district population density and vfllagepopulation are unrelated to accessibtity (Table 1). The apparent significance of Regression No. 1 is only because of the presence of Pankrono amongst the survey vi~ages.Pankrono lies within the Kumasi area and as a result has the highest accessibility of all vfiages and a population density that is over ten times that of any other area. If Pankrono is removed from the data set the regression relationship is shown to have no significance at dl. Dependent Variable district population density (J13) Wage population (J51) soil characteristics: PH level (J33) % organic matter (J34) p205 ppm (J35) K20 ppm (J36) Reg. No. 1 2 3 4 5 6 7 8 4 10 11 12 T~~ 1 Survey area characteristics Regression equation J13 = 313.7 – 2.88J15 J13 = 243.7 –3.08J17 J51 = 1591.7 – 2.26J15 J51 = 1955.5 – 12.78J17 J33 = 5.79 t 0.0037J15 J33 = 5.82 t 0.0053J17 J34 = 3.79 – 0.0052JI 5 J34 = 3.33 t 0.0027J17 J35 = 62.16 –0.021J15 J35 = 84.26 –0.581J17 J36 =405.5 – I.107J15 J36 = 375.1 – 1.09J17 R2 0.089 0.052 0.003 0.045 0.05 0.054 0.02 0.003 0 0.174 0.022 0.011 Independent Variables: J15 headload costs vWageto Kumasi, in units of @+ J17 headoad costs vtiage to district centre in units of @+ F Value 3.01 1.69 0.086 1.47 1.27 1.37 0.504 0.067 0.01 5.065 0.536 0.265 Observations 33 33 33 33 26 26 26 26 26 26 26 26 Significance level 10%1 Not sig. Not sig. Not sig. Not sig. Not sig. Not sig. Not sig. Not sig. 5% Not sig. Not sig. 1. If data from Pankrono is omitted in this equation R2 = 0.007 and F value = 0.223 making JI 5 an insignificant explanatory variable. 12 The sofl fertfity measures dso show little relationship with accessibility, although phosphorus pentoxide is statisticrdly associated in the sample with proximity to district centres. h later analysis these sofl fertflity measures were found to be largely unrelated to maize and cocoa yields or to crop des although organic matter content was found to be a significant explanatory variable of cassavasales. The variation in cassava sales may be explained by changes in the level of soti nitrogen that are indicated by the sofltest for organic matter content. It is interesting to note that the two vi~agesin the savanna sotis area, Dromankuma and Sekodumasi had the expectedly lower organic matter content in their sod samples than the other vi~ageswhich lay in the forest sofls area. 4.2 General holder ctiracteristics h total 491 holders were surveyed in 33 Wages. The holders lived in villageslying between 8 and 102 km from Kumasi; the average distance to Kumasi for the average holder was 61 km. 48 per cent of holders were mde and 58 per cent of au holders were over 40 years old. The average household size was found to be 4.7 people including one ctid. 72 per cent of holders had no schooling at dl. The average total farm size was 4.2 acres. 59 per cent of holders said their major source of livelihood was their food farm while 28 per cent claimed that this was provided by their cocoa farm. A further 9 per cent said that a non farming job provided their major source of livelihood. 36 per cent of the holders grew cocoa. The relationship between general holder characteristics and accessibility is shown in Table 2. The table shows that the proportion of holders that were over 40 years old increased with distance or travel costs from Kumasi. Similarly the proportion of holders that were mde also increased with inaccessibdity, Athough the more accessible vi~ageswere shown to have a greater proportion of holders with elementary school certificate and above, this was not statistically significant. It may be thought fiat accessibility would influence economic opportunities and hence household size, but no statistical relationship was found between accessibdity and the number of people in each household. The average total farm area was found to increase with inaccessibility,this being particularly marked in terms of travel costs to district centre. Average non cocoa farm area (as we~ as cocoa farm area) dso increased with inaccessibility. Table 2 shows that cocoa was reported to be a more dominant source of livelihood the more remote the farming location. By contrast the proportion of holders reporting that their food farms or a non farming job were more important sources of livelihood increased with accessibility. Accessibility thus appears to be lessimportant to cocoa farming than to food farming. The table shows that the proportion of vtiage population over 8 years with regular jobs dso increased with accessibtiity reflecting the greater job opportunities in the more accessible locations. 4.3 Labour input hbour input into farming is shown in TaMe 3. Household labour input per person and per holder was found to rise with inaccessibility. There is however some evidence to suggestthat household labour input per acre declined with increased transport costs to the district centre but the relationship was only significant at the 10 per cent level. This latter relationship may reflect the smaller labour demands of cocoa farms per acre because no significant relationship could be found for the two thirds of the holders that grew no cocoa. 13 TAB~ 2 General holder characteristics Dependent Variable % holders more than 40 years old (J2) %holder’s mde (J3) % holders with no education (J8) % holders with elementary school cert. and above (J12) average No. of people in holder’s house (Jl 1) average total farm area (J20) average non cocoa farm area (J25) % holders with cocoa as first source of tive~ood (J4) %holders with non farming job as first source of tive~ood (J5) % holders with food farm as first source of UveWood (J6) %Wage pop. over 8 years in regular jobs (J9) Zeg.No. 13 14 15 16 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Regression equation J2 = 38.9 +0.247 J15 J2 = 51.3 +0.107 J1,7 J3 = 23.8 +0.377 J15 J3 = 30.8 +0.455 J17 J8 = 65.3 + 0.0362 J15 J8 = 70.9 – 0.077 J17 J12= 21.4 –0.132J15 J12=17.4– O.12J17 Jll = 3.98 t 0.0067 J15 Jll = 4.56 –0.0032 J17 J20 = 1.42 t 0.0366 J15 J20 = 1.21 + 0.0664 J17 J25 = 0.914 t 0.0143 J15 J25=1.18+0.0174J17 J4 = 0.607 + 0.405,J15 J4=3.18+0.613J17 J5= 19 –0.128J15 J5 = 20.3 – 0.025 J17 J6=81.6– O.312J15 J6 = 77.2 –0.412 J17 J9=11.5– O.063J15 J9=12.6– O.132J17 R2 0.174 0.017 0.323 0.24 0.003 0.006 0.10 0.04 0.016 0.002 0.162 0.272 0.228 ‘0.173 0.286 0.335 0,063 0.118 0.159 0.142 0.096 0.216 F Value “6.53 0.525 14.78 9.81 0.08 0.18 3.42 1.36 0.508 0.058 5.61 10.85 8.56 6.07 12.4 15.6 2.08 4.15 5.87 5.13 3.29 8.54 Observations 33 33 33 33 33 33 33 33 33 33 31 31 31 31 33 33 33 33 33 33 33 33 Significance level 2.5% Not sig. 1% 1% Not sig. Not sig. 1o% Not sig. Not sig. Not sig. 5% 1% 1% 5% 1% 1% Not sig. 10% 5% 5% 1o% 1% hdependent variables: JI 5 headoad costs vtiage to Kumasi in units of @~ J17 headoad costs wage to district centre in units of@ ~ The average wee~y household labour input into each holder’s farm was estimated at 8.7 days. By contrast hired part-time labour contributed on average about 32 man days of effort for each holder for the whole farming year. Approtiately 20 per cent of holders claimed that a caretaker looked after some of their land, but caretakers were only recorded for cocoa farms. 14 No significant simple relationship could be found between averageWage wage rates and accessibfity. However, a multiple regression showed that average vWagewage rates rose with Wage population, transport costs to district centre and population density. This suggeststhat large isolated Wages must pay more for farm labour than sm~er more accessible Wages. Tm~ 3 hbour input Upendent Variable household days worked on farm per person over 8 years (J1O) total household days worked on farm per holder (J21) household days worked on farm per acre (J23) non cocoa holders standardised days worked per acre (J40) average Wage wage rate (J58) Reg. No. 36 37 38 39 40 41 42 43 44 43 46 Regression equation J1O = 2.27 t 0.011 J15 J1O = 2.37 t 0.016 J17 J21 = 7.03 t 0.04 J15 J21 = 7.53 + 0.054 J17 J23 = 5.7 –0.021 J15 J23 = 5.8 – 0.039 J17 J40 = 8.31 – 0.005 J15 J40 = 8.16 – 0.005 J17 J58 = 5.54 t 0.0078 J15 J58 = 5.46 t 0.014 J17 J58 = 4.25 t 0.00044J51) t 0.025 J17 ) +0.001 J13 ) R2 0.21 0.22 0.19 0.18 0.066 0.109 0.001 0.0005 0.04 0.068 0.366 F Value 7.69 8.32 6.74 6.28 2.05 3.54 0.036 0.017 1.16 2.04 8.54) 7.22) 3.6 ) Observations 31 31 31 31 31 31 33 33 30 30 30 Significance level 1% 1% 5% 5% Not sig. 1o% Not sig. Not sig. Not sig. Not sig. 1% 5% 10% hdependent variables: J13 district population density . J15 headoad costs tilage to Kumasi in units of Q~ J17 headoad costs Wage to district centre in unit~of @~ J51 Wage population 4.4 Modem inputs The farm surveys showed no evidence that inaccessibtity prevented the use of fertfiisers, tractors, or insecticide or that ‘it prevented contact with extension workers. However, the issue is somewhat comphcated by the two agricultural zones covered by the survey. Machinery hire and fertfiser use are more suited to savanna sofls which are fighter and easier to plough and often less fertfle than forest sofls. One of the two remotely located savanna vfiages, ~omankuma, alone recorded 32 per cent of total extension contact, 65 per cent of total machinery hire and 75 per cent of total incidence of fertdiser use of the whole su,wey. Table 4 shows that no direct significant relationship was found between extension contact or the use of cocoa insecticide and accessibility. This is unaltered even if data relating to the two savanna vtiages are excluded from the analysis. Npendent Variable %holders with extension contact (J7) % cocoa holders with cocoa sprayed (J52) Reg. No. 47 48 49 50 Modern inputs ~ J7 = 5.29 t 0.066 J15 0.014 J7 = 12.3 –0.064 J17 0.007 J52=42.5 tO.19 J15 0.041 J52 = 55.05 + 0.0037 J17 o 1 F Value 0.434 0.207 0.98 0 Observations 33 33 25 25 Significance level Not sig. Not sig. Not sig. Not sig. kdependent variables: J15 headoad costs tiage to Kumasi in units of @+ JI 7 headoad costs Wage to district centre in units of Q~ Out of the 65 holders that reported extension contact 44 holders mentioned the Crops Extension ~vision, 17 holders mentioned Cocoa Production Mvision, and 2 holders mentioned Veterinary Services. h answer to a specific question on the Ashanti Cocoa Project, 16 holders (out of a total of 179 holders that grew cocoa) said they were members of the Project. Overd it appears that the pattern of extension contact is more dependent on the local management and enthusiasm of individud extension workers than on the problems posed by inaccessibility even though the latter may we~ hinder directly or indirectly the overw efficiency of each extension organisation. 4.5 Holder fi~nce Holder finance is examined in Table 5. Nthough a simple positive relationship was found between transport costs from Kumasi and the proportion of holders that applied for fmancid assistance this relationship became insignificant once account was taken of holder age. A very strong positive relationship was found to exist between the proportion of holders who applied for fmancid help and the proportion of holders greater than 40 years old. A different picture emerges with the successin obtaining loans. Athough no significant relationship was found between accessibfity and gaintig a proportion of the money requested it does appear that money loaned per holder that applied was positively related to accessibdity. A simple regression relationship significant at the 5 per cent level shows that cedis loaned per holder applying was positively related to accessibility toKumasi(R2=0.18). Further examination by multiple regression showed}hat this relationship was strengthened (to 1 per cent significance) once the average number of people in the holder’s household and the transport costs to district centre were taken into account (R2 = 0.47). bans from both ‘official’institutional sources (eg. the commercial banks) and non official sources (money lenders, friends and famfly) were more difficult to obtain in the more remote locations. The comparative lack of successfaced by holders in the more remote vdlagesin obtaining institutional loans may relate to the communications problem of getting the holder’s field measured (a necessary part of the process) and the difficulty and expense of making follow-up trips to chase the progress of the loan. The greater difficulty in obtaining loans from tinofficial’ non institutional sources may reflect the greater scarcity of the latter sources of assistance in the more remote locations. On average 15 per cent of afl holders belonged to a loan cooperative and 22 per cent of holders had applied for fmancid assistance. Of those that applied for help an average of@191 was obtained from official sources and@ 127 from unofficial sources. (The average part time labour wage rate at the time of this survey was about Q6.5 per man 16 day.) Mostloans were for the duration of the agriculturrd crop season. h common with other surveys of this kind it was not possible to check effectively on how the loan was spent. The high rates of inflation coupled with the much lower ‘officird’interest charges would have provided an undoubted incentive to spend some of the low interest institutional loans on domestic consumption goods. However, it was stti profitable to use the loan to expand food production. TAB~ 5 Holder finance Dependent Variable % holders that apphed for finance (J41) cedis loaned per holder applying (J42) official aid as a % of total (J43) % of requested aid given (J44) Reg. No. 51 52 53 54 55 56 57 58 59 60 Regression equation J41 = 8.6 +0.225 J15 J41 = 19.4 tO.108J17 J41 = –17 t 0.734 J2 J42 = 679 – 5.26 J15 J42 = 480 – 3.84 J17 J42=195–11.6J15 ) tl18.3Jll) +g.61 J17 ) J43 = 54.3 + 0.03 J15 J43 = 52.3 t 0.1 J17 J44=62.1 +0.15J15 J44 = 72.6 – 0.006 J17 R2 0.131 0.015 0.489 0.185 0.05 0.466 0 0.004 0.02 0 ,F Value 4.66 0.49 29.7 4.75 1.12 10.3 8.47 3.67 0.02 0.09 0.51 0 Observations 33 33 33 23 23 23 23 23 24 24 Significance level 5% Not sig. 1%1 5% Not sig. 1% 1% 1o% Not sig. Not sig. Not sig. Not sig. hdependent variables: J2 per cent holders more than 40 years old JI 1 average number of people in holder’s house JI 5 headoad costs vtiage to Kumasi in units of Q~ J17 headoad costs Wage to district centre in units of @~ 1 BothJ15 andJ17 were found to be insignificant exploratory variables in a multiple regression with the other variables in this equation. 4.6 Cocoa production Of the 491 holders interviewed in 33 Wages, 179 holders grew cocoa in 23 Wages. The largest cocoa production per holder was recorded at Mpasaso which had both the largest cocoa and non cocoa farms. At 74 km from Kumasi, Mpasaso was in the midde ranges of accessibility of vtiages in the sample. Table 6 shows that the proportion of holders gro~ng cocoa significantly increased with inaccessibility. Both the average cocoa area per holder and the proportion of total farmed area devoted to cocoa significantly increased with transport costs to the district centres. No significant, simple regression relationship was found between accessibility and cocoa sales per grower or cocoa sales per acre. However, multiple regressions show that the average cocoa sales per cocoa grower were stron#y positively related to average cocoa area but negatively related to the average number of people in the holder’s household (R2 =0.55). The latter may reflect the domestic food needs of the holder’s household. 17 Average tilage cocoa sales per acre of cocoa farm were apparently stron~y related to the balance of sexes of holders in each Wage. The regression suggeststhat women holders are relatively more successfulin maintaining a higher level of productivity per acre. Cocoa yields W decline if the trees are ne~ected and disease is Mowed to spread. New trees need to be planted as the older trees decline in yield and die. It is interesting to note that plenty of evidence was found of new cocoa planting in the South Eastern area of the Region, an area previously affected by swo~en shoot disease. Of the 179 holders growing cocoa moderate or poor yield was attributed to poor sofl by 9 holders in 6 Wages, to capsid “ attack by 15 holders in 10 tiages, to black pod by 4 holders in 3 Wages and to weather by 16 holders in 6 Wages. No holders mentioned swo~en shoot disease as a contributing factor to poor cocoa yields. TABW 6 Cocoa production Dependent Variable % holders growing cocoa (J55) cocoa sales per cocoa grower (J37) cocoa sales per acre of cocoa farm (J38) cocoa area as % of total farmed area (J24) average cocoa area per holder (J22) Reg. No. 61 62 63 64 65 66 67 68 69 70 71 72 Regression equation J55 = 4.23 + 0.454 J15 J55 = 7.2 t 0.685 J17 J37= 1870 t4.31 J15 J37 = 1220 t23.2J17 J37= 356 t72 J22 ) –64 Jll ) J38 = 474 – 0.81 J15 J38=440-O.49J17 J38= 1023 – 13.5J3 ) t g.8 J7 ) – 62.7 Jll ) t8.7 J17 ) – 53.8 J22 ) J24= 20.9 +0.19J15 J24= 13.7 tO.5 J17 J22 = 0.502 t 0.022 J15 J22 = 0.031 t 0.049 J17 R2 0.279 0.325 0.004 0.064 0.55 0.008 0.001 0.90 0.054 0.188 0.105 0.259 F Value 12.02 14.94 0.074 1.17 17.79 ) 8.3 ) 0.13 0.03 80.0 ) 37.6 ) 11.4 ) 31.2 ) 11.6 ) 1.66 6.73 3.41 10.11 Observations 33 33 19 19 19 19 19 19 31 31, 31 31 Significance level 1% 1% Not sig. Not si . 1%f 1% Not sig. Not sig. 1% 1% 1% 1% 1% Not sig. 5% 1% 1% hdependent variables: J3 per cent holders that are male J7 per cent holders with extension contact J11 average number of people in holder’s house J15 headoad costs Wage to Kumasi in units of @A J17 headoad costs Wage to district centre in unit~of @~ J22 average cocoa area per holder 1 BothJ15 andJ17 were found to be insignificant explanatory variables in a multiple regression with the other variables in this equation. 18 4.7 Animal husband~ Smd numbers of poultry were kept by a very large proportion of the holders in the survey. Poultry farming on a commercial scale was reported in Koben and Mpasatia, two Wages close to Kumasi. It is generally believed that commercial scale poultry farming is fairly concentrated in and nearby the major towns in the Region. The major towns provide both a market for the poultry, and are a major source of the distribution of chicken feed. Because of the shortage of chicken feed concentrate the poultry farmer is usu~y keen to maintain good contact with the ‘Animal Husbandry Department to help maintain his supphes of the concentrate. h these circumstances a remote location wotid put the commercial poultry farmer at a distinct disadvantage. h total 420 sheep were reportedy kept in 15 Wages and 172 goats were kept in 7 Wages. Table 7 shows that there was no evidence of any relationship between accessibfity and ownership of sheep and goats although 25 per cent of d sheep and goats were kept in one Wage, Ofoase in the south east of the Region which had the longest road distance to Kumasi. This suggeststhat animal production is not particularly dependent on good accessibility. As expected virtu~y no other anirnds besides chicken, sheep and goats were reportedy kept by holders in the suNey. TAB~ 7 Dependent variable Reg. No. Regression equation R2 F Observations Significance Value level No. of sheep and 73 J53 = 0.87 t 0.007 J15 0.01 0.304 33 Not sig. goats per holder (J53) 74 J53 = 0.83 + 0.128 J17 0.016 0.51 33 Not sig. hdependent variables: J15 headoad costs vWageto Kumasi in units of Q~ J17 headoad costs Wage to district centre in units of@ ~ 4.8 Food production yields h Section 2.4 it was suggested that the overwhehing majority of food farmers in Ashanti Region practise mixed cropping. It was found in the course of the survey that the Ministry of Agriculture enumerators measured plant yields of ody one food crop from the crop mixture. The combination of these two factors made it extremely difficult to estimate total food production of farmers in the Region. Data sufficient for statistical tests was co~ected ordy on maize yields. Table 8 shows that no significant relationship was found between maize yields (either per yield plot or per plant) and accessibfity. It should be remembered that many other crops were grown in the same yield plots as the harvested maize plants. No statistical relationship was found between the maize yields and the sofl fertfity characteristics reported eartier. h order to test the hypothesis that more accessible land, being more valuable, would be planted more intensively a separate survey on planting density was carried out. 60 locations were visited in 16 dages and plant composition and plant populations recorded in randotiy placed plots. It was recognised that different plants tend to take up different amounts of land area. Even after allowing for a range of different combinations of ground area weighings for the different crops, no significant relationship was found between accessibtity and overd plant density. h none of the sixty plot locations was a completely pure stand of any food crop grown. 19 Dependent Variable maize yield per plot (J31) maize yield per plant (J32) Maize yields Reg. No. Regression equation R2 F Observations Significance Value level 75 J31 = 17.2 – 0.027 J15 0.014 0.166 14 Not sig. 76 J31 =18.2–0.07 J17 0.048 0.6 14 Not sig. 77 J32 = 0.45 – 0.000003 J15 o 0 14 Not sig. 78 J32 = 0.37 – 0.0018 J17 0.06 0.785 14 Not sig. hdependent variables: J15 headoad costs Wage to Kumasi in units of @~ J17 headoad costs vtiage to district centre in units of Q~ 4.9 Food ~op Sales h the survey 55 per cent of holders reported selling maize which was more widely sold than cocoa. Ody 17 per cent of holders reported se~g cassavaand 13 per cent reported sefling plantain. Ody one per cent of the holders reported se~ing tomatoes, cocoyam or rice. Srdesof otier crops were insignificant. Table 9 shows no apparent significant relationship between maize sales and accessibihty. However, cassava was sold relatively more frequently in the more accessible vtiages. The multiple regression shows a significant relationship between the organic content of soil, accessibility to district centre and cassava sales. By contrast the less accessible vflages reported selling more plantain. However, multiple regression analysis shows that this may be because plantain is grown on the larger mixed cocoa and food fafis. Hantain is frequently interplanted with cocoa especially when the cocoa is relatively young. OveraUit appears that accessibdity does not easdy explain the proportion of farmers in a Wage se~ing food crops. This may more easfly be explained by other factors such as the influence of household size which was found to significantly reduce the proportion of holders selling over 70 per cent of any food crop grown, Nevertheless, the proportion of farmers selling more than 30 per cent of any crop including cocoa, does apparently increase with inaccessibility although this may reflect no more than factors such as the increase in farm size and the rise in labour input per farm with inaccessibility. 4.10 Rotten produce and accessibility Only 16 per cent of the survey holders could recall personal experience of their produce becoming rotten before they could se~. Three vi~ages,Mpasatia, Mpatoam and Nyinahin (all of which had lower than average transport costs to Kumasi) accounted for 45 per cent of the reported cases. h total less than five per cent of holders fincluding all those givingmultiple reasons) identified road condition as a cause for concern in this respect. Because farmers were referring to particular instances over the last few years they remembered it appears that overall only a minute proportion of produce was effectively lost because of poor road condition. 20 Tm~ 9 Reduce sales Dependent Variable % holders se~ing over 30% of maize crop (J27) % holders se~ing over 3W0 of cassava crop (J28) % holders seWng over 3W0 of au food crops (J29) % holders sefling over 70% of au food crops (J30) % holders seUng over 30% of au crops inc. cocoa (J54) %holders sefling over 30% of plantain crop (J57) Reg. No. 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 Regression equation J27 =45.6 tO.116 J15 J27 =48 t 0.136 J17 J28= 23.2 –0.143 J15 J28 = 21.8 – 0.203 J17 J28 = 1.744 t 0.6 J34 ) – 0.22 J17 ) J29 = 70.7 + 0.14 J15 J29 = 73.8 +0.16 J17 J30 = 51.4 – 0.024 J15 J30 = 52.8 – 0.075 J17 J30 = 78.8 – 6.57 Jll J54 = 75 t 0.59 J15 J54= 81 t 0.84J17 J57= 0.81 +0.1 J15 J57=0.51 tO.17J17 J57 = –8.4 t 22.6 J26 ) tO.136 J55 ) R2 0.026 0,018 0.116 0.121 0.39 0.073 0.048 0.001 0.005 0.204 0.307 0.317 0.06 0.09 0.316 F Value 0.825 0.573 4.08 4.26 10.1 ) 5.29 ) 2.44 1.56 0.03 0.15 7.94 13.76 14.39 1.97 3.2 6.03 ) 3.15 ) Observations 33 33 33 33 26 33 33 33 33 33 33 33 33 33 31 Significance level Not sig. Not sig. 5% 5% 5% 5% Not sig. Not sig. Not sig. Not si 1%f 1% 1% Not sig. 1o% 5%1 1o% hdependent variables: J11 average number of people in holder’s house J15 head load costs vfllage to Kumasi in units of Q~ J17 head load costs tilage to district centre in units of@ ~ J26 non cocoa growing farm area per person J34 percentage of organic matter content in sod J55 percentage of farmers growing cocoa 1 J15 andJ17 were found to be insignificant explanatory variables in a multiple regression with the other variables in this equation. 4.11 Factors affecting th$ expansion of production The survey provided an opportunity to ask holders their opinions on the factors they could identify which tended to limit the expansion of production of their food and cocoa farms. 13 per cent of holders were not interested in expanding production, often they said they were too dl or too old to do any more work. Those holders that were interested in ~creasing production identified three key factors which limited their farming. Financial assistance was mentioned as a,constraint by 58 per cent of food farmers and by 46 per cent of cocoa farmers. Avaflable land was mentioned as a constraint for food farming by 20 per cent of holders and for cocoa by 27 per cent of cocoa farmers. bbour was mentioned as a constraint by 16 per cent of food farmers and by 22 per cent of cocoa farmers. 21 5. ACCESSIBILITY TMSPORT COSTS N PNCES Section 2.2 identified how a rise in farm gate prices may represent a major stimulus to increasing agricultural production. Unfortunately, it was not possible to assessaccurately farm gate prices in the survey in view of the lack of records kept by farmers, the high rates of inflation, the imperfect nature of the market, and the sensitive nature of the subject. However, in order to show how prices maybe expected to vary with distance the graphs in Fig. 2 give an estimate of maize prices for farms located at different distances from Kumasi. The relationships shown were derived by subtracting the wholesale transport charge and the headoading charge from the Kumasi retati market price after an a~owance had been made for wholesale and retafl margins. Data relating to market prices and wholesale transport charges for August 1978 were collected from the Ministry of Agriculture Statistics Division. Headoad costs were derived direcdy from the field survey. Fo~owing Gorel 7 wholesale and retafl margins were assumed to be one third of the find market price. Kumasi less wholesale and givescombined farm gate Market Price retd margins and transport charge Q91 — (4x market price) = Q60.17 From this can be subtracted the wholesale transport charge to give the farm gate price at different road distances from Kumasi and the headoading charge to give prices of headoading done is used. The fo~owing relationship was found between wholesale transport charges and travel distance by road. Charge per @ = 0.485 t 0.036 km R2 = 0.88 220 lb bag of maize F value= 132.4 22 observations By contrast the charge for carrying a standard 88 lb load by headoading was @0.5 per tiometre. It can be seen that at 100 km from Kumasi maize prices would be just over 6% per cent lower for Wages located on the road. A much steeper dectine in price is shown for Wages which can ody move produce by headoad. The relative change in farm gate prices with distance for any agricultural commodity dso depends on their ratios of value to weight and value to volume. The percentage change in price for the more bulky commodities may be expected to be greater than that shown for maize. However, an analysis of yam and plantain transport charges and prices showed relatively a much higher constant component and a relatively sm~er variable component with distance giving a much sm~er proportionate decline in price with distance. The percentage decline in price at 100 krns was little different from maize at 6.5 per cent for yam and 5.2 per cent for plantain. The regressions for yarn and plantain had sm~er R2 values and were much lesssignificant. These figures give some guidance in helping to evaluate the practical significance of the range of accessibfity measured in the survey. (31 of the 33 Wages of the survey had direct accessto a road or track and they were located between 8 and 102 km from Kumasi.) Athough changes in farm gate prices are important they cannot be thought to represent the total impact of different levels of accessibility, changes in input prices and the other factors already mentioned such as extension, accessto credit etc. must dso be taken into account. The cost to the farmer of buying industrial products WMdso rise with inaccessibility. A more detafled analysis of transport costs and prices is includ~d in SR 8093. 22 6. DISCUSSION The study has identified a range of factors which can influence agricultural development and it has dso shown that the measurements of development is complex in Ashanti Region at the present time. It is self+vident that in the extreme, agricultural development is dependent on accessibfity. If the costs of taking produce to market are too high then produce W not be grown profitably for sale. However, within the range of accessibfity considered the study found that the more remote Wages which were connected by roads and tracks for up to 100 km from Kumasi did not appear to have their agriculture adversely affected by their relatively higher cost of transport. Had the range of accessibdity studied been much greater (for example up to 400 km from a major market but with vehicle access or up to 25 km from a road for tiages without any vehicle access) then it seems reasonable to suppose that poor accessibility would have been seen to adversely affect agricultural performance, as the higher transport costs would have lead to poorer profitabtity of market agriculture. The policy of purchasing cocoa at a uniform price at CMBbuying posts widely located throughout the region ordy helped to minimise the adverse effects of inaccessibility on smfiolder cocoa farming. Within the range of accessibility considered, if anything, the least accessibleWages appear to be more agriculturdy developed than the most accessible tiages. The least accessibleWages had larger farms, grew more cocoa and sold a greater proportion of the crops they produced. They dso devoted more labour to farming @er member of household) than the more accessible flages. However, the overa~ strength of the relationships found were generdy weak. No evidence was found to suggestthat the less accessibleWages suffered any disadvantages in obtaining insecticide, fertfiser, using tractors or gaining extension advice. However, poor accessibfity might adversely affect agriculture in an important way, through the inabflity to obtain finance. Viflageswith better accessibility appear to be more dependent on non agricultural activities for their Evelihood. The development of non agricultural activities such as rural industry and more particularly the provision of rural services are, at first sight, more likely to be dependent on good accessibility for their success.Services are very dependent on a constant turnover of new clientele and clearly could not thrive in a smd remotely placed Wage. The study supports the conclusion that where a road investment induces ody a relatively sm~ change in transport costs and market prices (such as would arise, for example, from the upgrading of an existing track or earth road) then correspondin@y little impact on agricultural development may be expected. 7. ACKNOWLEDGEMENTS The work in this report forms part of a joint research programme between the OverseasUnit @nit Head: Mr J S YerreH) of the Transport and Road Research bboratory and the Buflding and Road Research Institute, Kumasi, Ghana (Acting Director: Dr Gidigasu). The work was undertaken for the Ghana Highway Authority as part of their Highway Research Programme and supported by the World Bank. 8. REFERENCES 1. Buflding and Road Research hstitute, Feeder Roads Study, Ghana Highway Authority. Kumasi, 1981. 2. ~NE J L. Road planning for rural development in developing countries: A review of current practice. Department of the Environment Department of Transport TML Report LR 1046, Crowthorne, 1981 (Transport and Road Research hboratory). 23 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. H~ J L, J D N ~WON and E A KWAKYE. Accessibdity, transport costs and food marketing in the Ashanti Region of Ghana. Department of the Environment Department of Transport T~ Report SR 809 Crowthorne, 1981 (Transport and Road Research Laboratory). BATEMANM J. Cocoa prospects 1975–1 985. hternationd Bank for Reconstruction and Development. Mission to Ghana, 1971. STERN R M. The determinants of cocoa supply in West Africa. Seminar on African Primary Products and International Trade. University of Edinburgh, September 1964. OURY B. A review of past efforts at deriving agricultural supply models for developing countries. WorldBank Economics Department Working Paper (18), 1968. CARNEMARK C, J BIDERMAN and D BOWT. The economic analysis of rural road projects, World Bank Staff Working Paper (241), 1976 world Bank). BOSERUP E. The conditions of agricultural growth, hndon, 1965. NORM~ D W. Rationalizing mixed cropping under indigenous conditions. The example of Northern Nigeria. Institute for Agricultural Research, Samara Zaria, 1965 (Ahamadu Bello University). OTOO J. Shifting cultivation and sofl conservation in Ghana and efforts to improve them. &ops Research Institute. Kumasi 1973. MINISTRY OF ECONOMICPLANNING, Five Year Development Han 1975/76–1979/80. Accra, 1977. M~ISTRY OF AGMCULTURE. Ghana Sample Census of Agriculture, 1970. Accra 1971. HILL P. The migrant cocoa farmers of Southern Ghana. A study of rural capitalism. Cambridge 1973. FINANCIAL TIMES, LONDON. July 28th 1979. FINANCIAL TIMES, LONDON. January 30th 1982. ATSU S Y. The focus and concentrate programme in the Kpandu and Ho Districts of Ghana. An evaluation of one agricultural extension programme. Institute of Statistical Social and Economic Research, University of Ghana Technical Publication Series. No 34 kgon. GORE C G. Food marketing and rural underdevelopment: A study of an urban supply system in Ghana. PhD Thesis. Pennsylvania State University, Department of Geography. 1978. 24 T 1–37 Survey Villages & Regional Capital D District Centre — Roads ‘-=- Regional Boundary Fig. 1 A map of Ashanti Region showing Io@tion of suwey villages — u a c n =.-: 8 0m ~ 8 0m 0 0 (JalesaloqM ol 6u!llas) (~) Peel 6?o0 L Jad aa!~daz!ew ,sJawJeJ palew!lsq . 9. APPENDIX WLLAGE S~VEY DATA The data listed here is used in the regression analysis of the main report. The data relating to each Wage is grouped into 8 geographic zones for the s~e of convenience. Each tiage is identified by its number in Figure 1. NOTE: Variables J29, J30 and J54 are expressed as percentages but some observations are above ‘100’. This is because a holder is listed each time he or she seflsmore than 30 per cent in J29, and J54 (or 70 per cent in J30) of any crop. Hence a holder W be fisted twice in J29 if he se~smore than 30 per cent of two crops. 27 % Holders % Holders % Holders % Holders % with cocoa with non with food farming farm as more than Holders as first ~stance 40 years job as first first mde source of liveWood source of source of Zone Vfllage VMagename No. of ‘o Kumasi hvelihood hve~ood No. Holders ~ J2 J3 J4 J5 J6 Zone 5 1 Pankrono 29 8 41 34 0 14 83 2 Atimitim 8 9 25 38 13 0 100 15 Maase 7 13 29 14 20 0 86 16 Edjunase 5 14 40 20 0 0 100 Zone 2 3 Tease 10 25 30 60 0 40 60 4 Koben 8 35 63 50 25 0 63 5 Mpasatia 28 29 46 50 18 11 71 6 Winiso-Sekyikrom 26 45 50 38 4 12 85 9 Mpatoam 18 48 61 22 61 17 22 Zone 3A 7 Anwia-Nkwanta 3 31 33 33 0 0 100 8 Huntado 14 36 50 21 7 0 93 10 Kensere 17 49 59 53 0 6 94 11 Akrokerri 12 51 83 42 0 42 58 12 Brobriasi 13 52 46 38 0 39 54 13 Kyenaboso 5 53 80 40 20 80 20 14 Edubiasi 9 54 67 44 11 a 56 Zone 1 19 Nyinahin 12 59 33 33 58 .0 42 Zone 3B 17 ~echewere 7 53 71 57 0 29 0 23 Kente 8 71 75 38 75 0 25 35 Ntebene 1 73 0 100 0 0 100 36 Kokoben 13 69 85 62 54 0 46 37 Obenebeng 9 76 22 100 67 0 33 Zone 6 21 Mpasaso 22 74 50 41 41 0 55 22 Tepa 27 72 70 15 52 15 30 24 Abonsuaso 35 89 54 63 40 6 20 28 Hwibaa 17 54 77 24 35 0 59 29 Nyambekyere 6 89 83 83 17 0 83 Zone 7 25 Sekodumasi 20 77 50 45 10 5 85 27 Dromankuma 26 84 85 92 0 8 92 Zone 4 26 Kyempo 15 90 64 53 53 0 47 31 Odubi 9 82 67 100 89 0 11 33 Ofoase 30 102 63 60 23 3 73 34 Dwendwenase 22 96 82 64 73 0 27 28 1 2 15 16 3 4 5 6 9 7 8 10 11 12 13 14 19 17 23 35 36 37 21 22 24 28 29 25 27 26 31 33 34 % Holder! with extension contact J7 o 0 0 0 40 20 7 6 0 0 0 0 0 0 0 0 0 0 5 0 8 0 0 41 30 15 0 53 92 0 0 5 0 % Holders with no education J8 79.3 37.5 0.0 60.0 40.0 87.5 60.7 73.9 ?7.8 100.0 92.9 82.4 33.3 81.8 80.0 87.5 66.7 85.7 87.5 0.0 84.6 55.6 81.0 88.9 25.0 100.0 83.3 80.0 52.0 66.7 87.5 63.3 53.3 % Vfiag( pop. ove 8 years u regular jobs J9 13.8 11.1 13.3 25.0 17.2 8.5 11.1 3.0 2.8 0.0 8.3 4.8 10.5 0.0 0.0 2.4 7.5 6.6 0.0 0.0 5.6 0.0 0.0 24.5 0.0 7.7 21.5 11.1 1.0 0.0 13.3 8.7 1.4 Householc days worked on farm per person ove 8 years JIO 2.83 2.45 2.34 2.50 2.07 1.86 2.70 2.95 2.40 4.17 3.14 2.67 2.42 1.97 2.11 2.17 3.19 3.73 3.54 6.00 3.35 4.41 2.8 — 4.02 — 3.61 2.34 3.00 3.85 3.13 2.36 3.85 AverageNo of people in holders house J11 2.62 2.25 2.57 1.60 3.70 7.75 5.71 4.12 6.28 2.00 3.21 5.88 3.58 5.46 5.80 4.88 5.58 6.57 5.13 1.00 7.08 2.66 4.05 9.74 2.23 5.41 3.71 4.35 4.69 3.46 4.66 4.80 4.18 % Holders with elementa~ school cert and above J12 3.4 50.0 57.1 20.0 10.0 0.0 14.3 26.9 11.1 0.0 7.1 5.9 33.3 18.2 0.0 12.5 16.7 0.0 12.5 0.0 7.7 0.0 4.8 7.4 50.0 0.0 0.0 10.0 0.0 8.3 12.5 6.7 6.7 Mstrict population density ‘op/sq. Kn J13 2054 56 56 56 67 67 67 100 57 57 57 198 45 45 45 30 24 57 42 42 72 44 57 57 57 49 49 30 30 67 67 67 67” Hea~oaf costs Wage t( Kumasi Q J15 0.9 1.5 1.9 1.9 1.9 2.8 3.7 3.7 5.6 3.7 5.6 5.6 5.6 5.6 5.6 5.6 5.6 5.6 8.3 9.8 13.3 10.5 7.4 7.4 9.3 8.3 9.3 9.3 9.3 9.3 11.1 13.0 14.8 Headoad costs farm to Kumasi $ J16 — 3.5 .6.9 — 3.2 4.2 5.2 8.7 7.1 4.6 6.3 8.6 — 10.6 7.6 7.9 8.3 7,0 9.3 10.8 17.0 11.3 9.3 9.4 11.0 12.1 — 15.2 12.7 11.9 16.9 15.4 18.7 29 Vi~age No. 1 2 15 16 3 4 5 6 9 7 8 10 11 12 13 14 19 17 23 35 36 37 21 22 24 28 29 2s 27 26 31 33 34 30 Headoad osts tiage to district centre c J17 0.9 1.5 1.9 1.9 1.9 2.8 0.7 3.7 5.6 1.9 1.9 1.9 1.9 1.9 2.8 2.8 5.6 3.7 5.6 7.1 10.1 7.7 7.4 0.0 5.6 3.7 3.7 5.6 5.6 5.6 5.6 .7.4 9.3 Average otal farm area acres J20 0.49 2.58 1.56 0.78 2.92 5.70 7.24 1.64 3.63 1.03 1.56 1.60 1.02 1.39 2.58 1.77 7.69 8.25 1.50 0.60 7.32 2.06 12.91 — 4.17 — 2,12 3.01 2.86 7.30 8.49 3.88 10.91 Total lousehold days workedon farm per holder J21 6.34 5.51 5.00 4.00 6.00 14.40 10.43 7.48 9.47 8.33 8.50 ~ 13.04 7.66 9.25 11.40 9.86 14.08 16.00 13.28 6.00 18.31 10.28 8.27 — 7.35 — 8.40 8.44 10.73 10.29 10.43 .10.00 12.77 Average ocoa area Ierholder acres J22 0.0 1.88 0.0 / 0.0 2.12, 3.90 3.53 0.05 2.44 0.0 0.46 0.0 0.0 0.60 0.0 0.71 5.08 ;.82 0.95 0.0 4.54 0.84 8.85 — 2.71 — 0.0 0.23 0.0 3.43 5.04 1.32 7.74 Household days worked on farm per acre J23 12.9 2.1 3.2 5.1 2.1 2.5 1.4 4.6 2.6 8.1 5.5 8.2 7.5 6.7 4.4 5.6 1.8 1.9 8.9 10.0 2.5 5.0 0.6 — 1.8 — 3.9 2.8 3.7 1.4 1.2 2.6 1.2 Cocoa area as percentage of total farmed area J24 o 73 0 0 76 44 19 3 67 0 29 0 0 43 0 40 66 70 63 0 62 41 69 — 65 — o 8 0 47 59 34 71 Average non cocoa farm area acres J25 0.49 0.70 1.56 0.78 0.80 1.80 3.71 1.59 1.19 1.03 1.10 1.60 1.02 0.79 2.58 1.06 2.61 2.43 0.55 0.60 2.78 1.22 4.06 — 1.46 — 2.12 2.78 2.87 3.87 3.45 2.56 3.17 Non cocoa farm area per person acres J26 0.19 0.31 0.61 0.49 0.22 0.23 0.65 0.39 0.19 0.52 0.34 0.27 0.29 0.15 0.45 0.22 0.47 0.37 0.11 0.60 0.96 0.46 1.00 — 0.66 — 0.67 0.64 0.61 1.12 0.74 0.53 0.76 6 Holders se~ing over 3070 of maize crop J27 70 38 29 0 80 75 14 46 72 100 71 53 25 , 23 60 33 33 71 50 100 39 11 64 56 46 65 33 75 92 20 78 67 77 % Holders % Holders % Holders Matie Make Cocoa se~ing se~ng se~ng sod % sod sod yield yield SofiPH sales per over 30% over 3W0 over 7W0 organic P205 k20 per plot per plant level cocoa Vfiage of cassava of au food of& food lbs lbs matter ppm ppm grower No. crop crops crops @ J28 J29 J30 J31 J32 J33 J34 J35 J36 J37 1 31 101 93 8.5 0.280 6.5 2.7 34 166 – 2 13 50 25 – – – – – – – 15 14 43 29 – – – – – – – 16 40 40 40 – – – — — — — 3 10 100 60 12.5 0.263 5.8 3.1 31 329 1,100 4 0 88 63 – – – – – – 700 5 39 89 93 30.8 0.299 5.5 1.9 65 846 1,700 6 15 81 15 12.3 0.480 7.1 4.4 73 144 700 9 0 78 28 24.5 0.851 6.0 4.2 36 86 2,900 7 0 100 100 – – 5.4 3.3 47 645 – 8 7 93 86 – – 6.5 3.7 72 710 – 10 35 88 41 9.4 0.430 5.7 3.0 86 660 – 11 33 58 17 – – 5.6 3.9 107 154 – 12 46 69 39 – – 6.1 5.7 58 84 – 13 0 60 20 – – – — — — 1,300 14 11 44 22 – – – – – – 600 19 50 92 25 19.9 0.671 5.4 7.2 23 137 2,000 17 0 86 20 7.0 0.650 5.4 3.1 60 697 2,000 23 0 75 40 – – – — — — 1,700 35 0 100 100 – – 6.7 4.6 88 563 – 36 8 62 23 13.6 0.436 5.9 2.7 39 733 1,100 37 0 67 56 – – 5.7 2.9 64 700 200 21 5 82 64 – – 5.3 3.1 39 287 10,800 22 7 63 19 – – 5.5 3.0 204 280 – 24 20 86 89 – – 6.9 4.1 32 288 5,100 28 24 94 47 19.2 0.205 6.9 5.4 52 503 – 29 0 83 83 – – 5.2 2.4 43 130 – 25 5 90 80 7.9 0.372 5.9 2.0 48 74 2,400 27 0 96 6S 29.4 0.467 6.1 0.9 86 75 – 26 7 100 53 – – 6.6 2.4 44 61 1,500 31 11 100 44 – – S.8 3.2 48 68 1,300 33 7 87 47 1s.2 0.590 6.3 2.8 64 82 1,400 34 9 100 18 5.3 0.286 7.1 3.8 37 99 2,500 31 ems Ioanea UIIICl~ . . . Percent of aid as VWage cocoa —--------- requested holders d @ven population with cocoa Cocoa Non cocoa ~’>,., -...> fire . , Percent of sales per holders % Holders acre of standardised that apptied per n“!aer VWage cocoa farm days worked for fmmce applying percerltage tic No. @ per acre c of total sprayed J38 J40 J41 J42 J43 J44 J51 J52 1 – 11.67 0 — — — 2,080 – 2 – 7.03 0 — — — 836 100 15 – 6.17 0 — — — 1,203 0 16 – 5,29 0 — — — 290 – 3 105 7.85 0 — — — — 50 4 107 8.48 50 .-2,284 56 100 1,714 60 5 185 10.47 30 190 14 57 608 18 6 1,029 4.54 0 — — — 903 9 9 885 9.89 4 280 54 10 1,800 69 7 – 8.68 0 — — — 576 – 8 – 7.84 15 100 0 2 810 100 10 – 8.39 6 400 100 100 1,177 — 11 – 8.34 50 250 73 44 3,167 — 12 – 6.97 46 433 81 81 903 — 13 427 5.60 40 350 57 100 136 100 14 427 5.56 33 400 100 67 3,093 100 19 228 19.96 17 75 100 36 4,816 14 17 276 12.03 43 360 0 10 301 0 23 637 5.25 25 290 14 100 145 50 35 – 8.87 0 — — — 112 — 36 66 24.67 38 76 48 100 461 50 37 73 4.88 11 150 100 100 115 29 21 497 3.09 0 — — — 1,305 56 22 – 21.18 30 161 59 124 6,696 71 24 761 4.36 3 290 14 100 834 79 28 – 7.45 59 100 0 100 900 60 , 29 – 4.66 67 300 25 92 253 50 25 1,190 3.29 0 — — — 5,075 0 27 – 1.61 15 225 100 36 787 — 26 284 3.53 33 112 100 35 410 67 31 194 3.50 56 180 0 100 716 71 33 313 9.34 10 253 26 95 2,038 87 34 295 8.87 64 178 100 47 1,166 90 32 ) Vtiag( No. 1 2 15 16 3 4 5 6 9 7 8 10 11 12 13 14 19 17 23 “35 36. 37 21 22 24 28 29 /25 27 26 31 33 34 No. of sheeI and goats per holder J53 0.0 0.0 0.0 0.0 0.0 2.0 1.4 2.5 2.5 0.0 0.4 1.5 0.3 0.2 11.0 0.0 1.1 0.0 9.6 0.0 0.0 3.4 0.3 0.3 0.2 0.0 0.0 1.6 0.0 0.0 1.3 4.9 0.0 % Holders setig over 3W0 of A crops (inc. cocoa) J54 101 63 57 40 120 151 125 89 150 100 100 88 58 69 80 55 150 157 150 100 139 145 123 130 126 159 100 110 96 167 178 134 186 Percent of holders growing cocoa J55 o 13 14 0 20 63 36 8 72 0 7 0 0 0 20 11 58 71 75 0 77 78 41 67 40 65 17 20 0 67 78 47 86 % Holders se~ing over 3W0 of plantain crop J57 o 0 0 0 0 13 32 14 0 0 0 0 0 0 0 0 8 0 25 0 15 56 14 0 9 0 0 0 0 73 0 13 14 Average tiage wage rate $ J58 8.00 5.00 5.00 5.00 5.86 6.50 4.12 5.86 6.00 5.00 4.75 8.00 — — — 6.82 5.33 6.50 5.00 4.89 5.00 9.50 5.91 5.04 6.13 4.33 9.08 6.00 8.00 5.00 8.00 8.00 33 (1898) Dd8041338 1,200 8/83 HPLtd So’ton G1915 PRINTED IN ENGLAND ABSTWCT ACCESSIBILITY AND AGWCULTURAL DE~LOP~NT ~ THE ASHANTI REGION OF G~A. J L Hine, JD NRiverson Qnd E A KwQ~e: Department of the Environment Department of Transport, TRRL Supplementary Report 791: Crowthorne 1983 (Transport and Road Research hboratory). The report examines the relationship between agricultural development and accessibtity in the &hanti Region of Ghana. A wide variety of factors are identified that can influence agricultural development in the Region and some of the problems of its measurement are hi@ghted. Using a cross sectional framework of analysis, data was co~ected from 33 tilages (M but two with vehicle access) in the Aahanti Region of Ghana located between 8 and 102 km from the Regional Capital, Kumasi. By comparing a number of development parameters and the transport costs of moving farm produce between each Wage and Kumasi (and dso between each Wage and its respective district centre) me fink between accessibility and agricultural development was investigated. Within the range of accessibility considered littie evidence was found to indicate that market agriculture was promoted directiy by accessibtity. However, loan finance was easier to obtain the nearer the farmer lived to Kumasi. Overd there is evidence to suggestthat the most accessibleWages tended to concentrate more on non agricultural activites (such as rurrd industry and the provision of services,including marketing) wtie the less accessibleWages concentrated rather more on agriculture. The study supports the view that where road investment can induce ody a smrdl change in transport costs then little impact on agriculturrd development may be expected. ISSN 0305-1315 ABSTWCT ACCESSIBILITY AND AGNCULTURAL DEWLOP~NT ~ THE ASUTI REGION OF GHANA. J L Hine, JD NRiverson Qnd E A Kwa&e: Department of the Environment Department of Transport, TRRL Supplementary Report 791: Crowthorne 1983 (Transport and Road Research bboratory). The report examines the relationship between agricultural development and accessibtity in the Ashanti Region of Ghana. A wide variety of factors are identified that can influence agricultural development in the Region and some of the problems of its measurement are hi~ighted. Using a cross sectional framework of analysis, data was coflected from 33 Wages (d but two with vehicle access) in the Ashanti Region of Ghana located between 8 and 102 km from the Regional Capital, Kumasi. By comparing a number of development parameters and the transport costs of moving farm produce between each Wage and Kumasi (and dso between each Wage and its respective district centre) the link between accessibility and agricultural development was investigated. Within the range of accessibility considered little evidence was found to indicate that market agricdture was promoted directiy by accessibility. However, loan finance was easier to obtain the nearer the farmer fived to Kumasi. OverM there is evidence to suggestthat the most accessibleWages tended to concentrate more on non agricultural activites (such as rural industry and the provision of setices, including marketing) wtie the less accessible Wages concentrated rather more on agriculture. The study supports the view that where road investment can induce ordy a smd change in transport costs then litfle impact on agricdturd development may be expected. ISSN 0305-1315