Research Review NS Vbl.4 No.1, 1968 ECONOMIC DIFFERENTIATION AMONG GHANAIAN MIGRANT COCOA FARMERS Kwama Arhin I. INTRODUCTION This paper examines the econcmic basis of differentiation among migrant cocoa farmers in the Central and Western Regions of Ghana. Polly Hill, the pioneer student of migrant fanrers on the Gold Coast (now Ghana), saw the earlier migrant fanrers as uniformly capitalistic, embattled errtrepreneurs who deserved serious study because they were capitalists. (Hill: 1963a 203-223; 1963b; 1970). But certainly, in the first two decades of cocoa produc- tion what Gumarsson (1978: 107) identi- fies as the socio-economic factors of differentiation could not have existed. Hopkins (1973) believes in the emergence of Kulak-type fanrers at a point in the production for export, and writes: "On the Gold Coast a small group of wealthy farmers, who were responsible for a dis- proportionately large amount of cocoa shipped in the colony, had already appeared by 1930" (1973: 238-239). Gumarsson (op. cit.) goes at sore length into the question of "socio-economic stratification" of fanners but in the older cocoa growing areas. He describes the factors of stratification as the commercialization of land and the marketing system based on cash financing, which led to the concentration of wealth in the hands of a group of landlords, brokers and factors, and the corresponding impoverishTent of the large majority of farmers who took advances and pledged their crops to the brokers and factors (pp. 110-121). Bernstein (1979: 430) has suggested two bases of differentiation within African peasantries, among whan must be included the migrant cocoa farmers (Howard, 1980: 61-77; Arhin 1983b: 471-475). Bernstein's two bases of differentiation are those in the "materialist sense" that is, in terms of relations to the means of production, and in the "sociological sense" or in terms of life-style. This paper examines differentiation among the migrant cocoa farmers in the 'materialist sense", and not the "sociological sense". There are no visible bases for differentiation in the sociological sense. The migrant settlements generally consist of simply constructed isolated homesteads, which do not form carmunities, so that differentiation, even in terms of traditional ranking (Arhin, 1983a: 2-22), does not exist. The living conditions of the farmers ane those of frontier farmers with inconsiderable material culture. At the same time it was clear fran obser- vation and questioning among the migrant cocoa fanners that there were substantial econcmic differences among them; and that on the basis of these differences it is possible to distinguish groups with diffe- rential relations to the means of produc- tion and who, consequently, correspond to the groups Bernstein distinguishes as "poor", "middle" and "rich" peasants. wattle-and-daub Before examining the basis of this differentiation, I point out some of the characteristics of the migrants in the Central and Western Regions. 10 II. THE MIGRANTS Unlike the migrant cocoa farmers of the late nineteenth century, who were Akwapem, and the Ga-Adangne of the present Eastern and Greater Accra Ragions (Hill, op. cit.), the migrant cocoa fanners in the Central and Western Regions are drawn from all the ethnic groups and the regions of Ghana and beyond. Responses from 809 migrants interviewed show the ethnic composition of the resDondents as follows: Akan Ga-Adangre Eue forthem/Upper Ragion Nationals Other Ghanaians Aliens Total 670 or 82$ 70 or 9% 32 or 4% 6 or \% 9 or \% 22 or 3% 809 100% The Akan, who are predominant in the Ashanti, Brong-Ahafo, Central, Eastern and Western Regions, consist of the Agona, Akwapem, Akyen, Asante, Assin, Bnong, Denkyira, Effutu, Ekunfi, Gomoa, Ktoahu, Nzema and Senya, and collectively consti- tute about 43.5$ of the peoples of Ghana, according to the 1970 Census. They inha- bit the forests of central and southern Ghana and following the lead of the Akwa- pen pioneers (Hill, 1956; 1963) have, since the beginning of this century, been the major producers of cocoa. The Akwapem established the tradition of migrant cocoa farming in the late nineteenth century, and migrant cocoa fanning has, since then, been a continuation of the fanner's occupation in virgin forest lands. The Central and Western Regions are the latest such lands. The Ga-Adangne (Ga and Krobo) and the Eoe of Southeastern and Eastern Ghana, have also been long associated with cocoa cultivation either as cocoa farm owners - for the Krobo on land purchased from the people of Akyem Abuakwa - or as labourers (Hill, 1963). The various peoples of the Northern and Upper Regions of Ghana and of Burkina Faso have been connected with 11 cocoa production but mainly as labourers on the farms (Beckett 1944: 60-61; Mstcalfe 1964: 65). The average age of the migrants was 45 years and 85% of them were over 30. 7 1* were illiterate and 28% literate. 41% had previously been cocoa farmers; 35$ had turned from food crop and vegetable fanning to cocoa fanning. The rest (24 %) had engaged in multiple occupations, including the cultivation of citrus/orange, coconut, oil palm, coffee or cola. Same had been custodians of family property, land and buildings. The migrants included traditional office-holders such as stool occupants (sub-chiefs) .lineage heads, and palace functionaries. The migrants, then, had not all been cocoa fanners. Just over 40$ of then had gone to farm cocoa in new areas. About 60% had turned to cocoa fanning frcm other occupations in response to the high cocoa prices of the nineteen-fifties. Investment in cocoa then seemed worthwhile. The data shows that while one or two migrants had moved to the new cocoa produc- tion areas before 1950 the westward migra- tion really took off in the fifties. The presence of literates in cocoa farming is of some interest as a new development: in the territories that now constitute Ghana, literates were not expected to go into fanning of any sort. (See Addo, 1974: 205). In the context of this paper, the signi- ficant point relates to property ownership among migrants. A migrant's carmand over the means of production in the area of settlement depended on whether or not he owned property and the kind of property he owned. This is because farming in the mid-twentieth century requires considerable outlay. The size of the farm depends on the farmer's ability to purchase land (See Ckali, Cwusu-Ansah and Rourke, 1974: 175-179) labour and farm inputs such as matchets, weedicides and pesticides; some fanners even need to invest in a cocoa shed or storage building. Responses from 814 respondents to ques- tions on property (cash and equipment) showed the following: in the Assin area the average amount farmers took to the farms was 076.70; in Wasa 0148.57; and in Sehwi 8227.70. The figures in Wasa and Sehwi probably reflect the predominance of Asante migrant farmers in the two areas. Farmers generally had amounts ranging from under 010 (about £5 sterling at the time) to 01,000.00. 91 % of the respondents had between 01 and 0400; anc1 9% between 0400 and 01,000.00. There were a few outstanding individuals, such as an erstwhile cocoa farmer who arrived with 02,000 and complained that it was inadequate. Besides cash, fanners also brought with them some equipment. This included farm tools such as matchets, hoes, axes and sharpening stones; personal effects, household goods such as cooking utensils, mats and beddings; and "business" items, including guns, sewing machines, smithing tools, looms and lorries. The amount of cash and equipment a farmer brought into the field depended on his occupational background, and deter- mined his standing as a farmer. His equipment facilitated his settlement in the new area and the cash gave him access to the means of production, land, labour and farm inputs. I shall now go on to examine how diffe- rentiation among the fanners is related to access to the means of production. II (a) F W E RS *O LAND The size of farm land a farmer acquires and the terms of its acquisition depend on the amount of cash he has on arrival in the farming area (Okali et al. op. cit.). The Central and western Regions are inhabited by the Assin, Sehwi and Wasa peoples, all Akan sub-groupings with centralized political and corresponding land tenure systems. Along the various Akan sub-groups, land sales may have been rare before the late nineteenth century for a variety of economic and religious reasons.2 But since the /fcuapem moved into the Akyen Abuabe area in the last decade of the nineteenth century, land sales have become a firmly established aspect of cocoa production in areas other than the producer's own tirth-place, where he has usufructuary rights in land as part of his bundle of socio-political rights. 3 The survey showed that in the Central and Western Regions, land was available for outright purchase (trama) in all the three areas; landowners included (i) chiefs, stool occupants and citizens; the primary owners of the land, and (ii) migrants, relatives of migrants and spouses of migrants; secondary owners of land.^ This means that all the three areas had reached the point where land had become a carmodity; whoever acquired 5 or purchased land could sell it (See also Bermeh, 1970). It was not possible to establish the cost of land in any of the units of size mentioned by the farmers; poles, "arm- lengths" (abasa), or ropes (ahama). But in relation to the estimated amount of money migrants brought with them to the field, it was high, so that seme indivi- dual farmers had to form companies^ in order to be able to afford outright purchases. Unspecified areas of land were estimated to cost between 071.14 and 076.86 in the three areas. But these suns must be set against the estimated amount of 0147.09 that migrants had on arrival. The data shows that not many farmers acquired land through outright purchase, so that outright purchasers of land them- selves became landlords and land specu- lators, and formed a group apart from the migrant farmers who obtained land on the following terms: abusa: an arrangement by which a cultiva- tor divided his cultivated farm or harvest proceeds into three with the landowner and took two/thirds or a third; 12 abunu: an arrangement by which farms or proceeds were divided into two, and the labourer took a half; nto (ground rent): an arrangement by which the landowner claimed varying proportions of proceeds as tribute. These arrangerrents, other than outright purchase, turned the migrant farmer into a dependent labourer of the landlord, and gave him uncertain tenure on two grounds. The landlord could vary the terms of holding to an extent unacceptable to the tenant so that either he quit the farm or took to litigation;7 and it was never clear to a tenant that his heirs could inherit his farms. Landlords were very much inclined to take advantage of migrants so that in the three areas the mast usual arrangement for' land acquisi- tion were in the following order:- Percent 41 25 9 8 8 5 3 1 Terms outright purchase abusa abunu free (gift) any contract ground rent nto share proceeds 1) abusa 2) abunu 3) nto 4) trama Accordingly, the survey showed the following order of preference among the migrant farmers for the various terms of land acquisition: Rank 1 2 3 4 4 5 6 7 Two obstacles stood in the way of outright land purchases: lack of cash and land- lords' increasing unwillingness to sell land outright. A man who arrived in the field with a sizeable amount was able to purchase land, often in enough quantity to enable him to speculate in land, and join the group of landlords. He was then freed frcm the burdens of tenancy, uncer- tain tenure, varying tenurial terms and possible litigation. On the other hand, those who obtained land on abusa, abunu and nto terms belonged to the class of tenant or depen- dent farmers, whose status was only slightly above that of labourers. They differed among themselves only on the basis of the different terms of tenancy; a two-thirds share in cultivated farms or harvest proceeds was more profitable than a third share or indeterminate ground-rent or tribute exacted by the landlord as he pleased. Migrant farmers might then be differen- tiated economically on the basis of the terms on which they acquired land for cultivation. The two broad groupings were outright purchasers of land and tenant farmers. The first group consisted of previous cocoa farmers and others who went to the field with cash and equipment to continue the "business" of cocoa farming in new areas. The second group may or may not have been previously asso- ciated with cocoa farming but in either case they did not have enough cash to enable them to set themselves up as inde- pendent cocoa farmers. The first group can be identified as among Polly Hill's capitalist farmers while the second group constitutes the rural working class. II (b) BWMERS ft€ LffiOUR As with land so has labour also become in part a ccrrmodity that a farmer could obtain through purchase. Labour is identified as, in part, a cormodity because the data shows that farmers used three kinds of labour on the farms. These were family, hired and co-operative (flcan, rroboa) labour. Family labour was nominally unpaid service of a wife, child, sister's son or daughter (as among the matrilineal Akan) or of any other relation. Hired labour was of three kinds: (a) casual or 'by day1 labour; (b) permanent or full-tirre labour and (c) contract labour. Casual or 'by day' labour meant that a farmer recruited the labourer for a day's work on the farm, paid him the basic wage prevailing in the area and, in some cases, provided him with food and drink, as well, in the course of his work. Permanent or full-time labour consisted of (i) annual and (ii) seasonal 13 labour. An annual labourer was recruited for the whole year and received remunera- tion in cash plus other incentives such as food, clothing and shelter. A seasonal labourer was hired to undertake certain farming duties at the peak season, October to January, and for as long as the need for which he was employed existed. With regard to contract labour, agreement is reached with a labourer or a group of labourers to do a specified piece of work for a specified cash payment. Co-operative labour is mutual aid labour, an adaptation of an arrangement of the past by which young men mutually satisfied family and affinal demands in the farm and for other household tasks. The responses from 814 questionnaires showed that:- 53% used family labour; 33% hired labour; and 14% used mctoa labour:8 Hill says of labour on the cocoa farms: And while on the subject of farm labourers, it must be noted as a most important point of principle (though, again, one that has gone unnoticed) that farmers never "wasted" their savings on farm labour employment; they rarely employed labourers until with the help of their families, they had established sufficiently bearing cocoa farms to reward their labour with a share of the cocoa income (HU1, 1970:2). Killick also states: These figures suggest that the cocoa industry is essentially run by those who run the farms with the help of their families (Killick, 1966: 238-239). Hill and Killick are right only to a limited extent. The average cocoa farmer is usually not in a position to employ annual or seasonal labour although larger farmers employ all the categories of hired labour. Even in the late 1930's the Comnission on the marketing of Wast African cocoa remarked on labour on the cocoa farms:- The original conception of the Gold Coast farmer ... is of a peasant culti- vator who, with his own labour and the help of his family, grows his food and tends an acre or two of cocoa trees. This picture is no longer true of more than a minority of farms, and these of the smallest size. The enployment of labour has become a regular feature of cocoa growing, even where the cvmer resides on the farm (Metcalfe, 1964:653). The comnission probably exaggerated the extent to which paid labour had replaced family labour, but it was clearly the case that the use of paid labour parti- cularly of migrants on the cocoa farms was increasing. Addo (1974:206) has written of labour on the cocoa farms: of those who did not require permanent labour the main reason given was that the farm(s) was (were) not large, enough to justify employing such personnel. Many felt that they could always deploy family labour and therefore did not require permanent labour. The use of hired labour is more urgent in the migrant farming areas where, owing to the competition for land, a fanner has to clear his acquired land within reasonable periods in order to avoid litigation arising from counter-claims (Okali et al. 1974). And the bigger the farm, the more hired labour is needed to supplement family labour. Thus it was found that78£ of all hired labour was used for weeding and clearing, while 66% of all family labour was used in weeding and planting. (/VJdo 1974:206) 27% used only one kind of labour, that is, family labour, to esta- blish and maintain their cocoa farms, as caipared to 5% who used only hired labour for this;- and 5 8% of the sample combined their own services with those of family, casual and co-operative labour in cocoa cultivation (See also Addo 1974:206). Farmers employed various kinds of labour on their farms, and the use of hired labour was a matter of the capacity to pay. This becomes clear when labour costs are exa- mined in the light of the average amount of money that migrants claimed to have brought with than. The following were our estimated costs of labour: Casual or 'by day1 labour - 82 pesewas per person, per day, Annual labour - 98.92 pesewas per person per year, Contract labour - 34.15 pesewas per contract, per year. It was noted above that farmers had an average of 0147.09 on arrival. The migrant could barely subsist on this amount, much less use it for hiring labour. In addition, the acquisition of land on any condition involved the use of money and purchased drinks as consideration.9 After the acqui- sition of land, not much money was left for hiring labour. In surmary, a migrant cocoa fanner with enough cash was likely to employ hired labour on his farm; and, in fact, he had to employ hired labour for clearing purpo- ses in order to validate his claim to his acquired land. Land left uncleared for a year or more after acquisition was likely to attract counterclaims and become an object of litigation. It is more than likely that those who hired labour were those who made outright purchases of land. They not only owned land, the basic means of cocoa production; they also bought the labour of others. II (c) F/WERS fH) FARM INPUTS In order to safeguard his farm against pests, and in other ways maintain high productivity, the fanner requires such farming inputs as matchets, spraying equip- mant and sprays, and fertilizers. Before 1973, a Division of the Ministry of Agricul- ture, Cocoa Services, administered the supply of these inputs to the farmers. Since 1973, the Cocoa Services Division has become part of the cocoa Marketing Board. It was found fran the 807 responses to the questionnaires that 84% of the farmers regarded the administration of supplies as highly unsatisfactory. The average dis- tance frcm farms to the supply centres was nineteen miles, and the extremely poor network of roads prevented access to lorries, so that supplies had to be head- loaded to the farms. Of the 84% of the respondents who considered the adminis- tration of supplies unsatisfactory, 9 out of the 10 thought the cost of the services was too high, and made such statements as: "CWing to the high prices, I have not been able to buy a spraying machine frcm them"; "I am unable to pay for their services. I cannot buy garrmalin"; and "their supplies core in at a time that I have not the money to buy any". The high costs meant that only the rela- tively rich farmers could afford than. Sane of the fanners were aware of this and said "they feel my farm is not big enough to warrant their working on it". The rich had alternative sources. Seme of the farmers stated "I buy these things frcm KLmasi" or "I buy my DDT from a buying agency". In seme areas, supplies of the spraying material were linked up with possession of the spraying machine. Fanners complained that officials refused to sell garrmalin to them because they did not possess chit coupons issued to owners of spraying machines, who were considered to be the only people entitled to buy garrmalin; that garrmalin was supplied on rationing basis to non-holders of "blue cards"; and that these restrictions made it impossible for the farmers to benefit fnem the services. Fanners said "I cannot buy DDT from them because I have no permit to do so" and "they attend only to those with certifi- cates, but I have not got one". "Blue cards" and "certificates" were issued to the owners of spraying machines, the big farm owners.'0 As is corrmon experience in Ghana in respect of scarce carrnodities, the rela- tively rich farmers, in control of hired labour, also had access to most of the farm inputs. 15 III. TJ£ PHBL96 GF The possession of "capital" in the fonr. of both cash and equipment enabled a migrant to deal satisfactorily with the early problems of settlement; the absence of capital induced a recourse to solutions that generally curtailed the independence of the farmer and gave him a labourer's status, frcm which it is usually difficult to recover. The respondents to our ques- tionnaires stated their problems, other than the lack of capital as: 45% said it was difficult to maintain their family owing to; insufficient food (253), lack of acccmrodation (16%), and ill-health of members of tiie family (39, or the general inability to meet the bare necessities of life: 31% referred to financial problems which were related to difficulties in obtaining farm land, lack of capital for farm deve- lopment (lack of farming implements and other inputs) and poor conditions of the farm, poor soils, and threats posed by pests and crop diseases; 14% stated their problems as social: they had difficulties in adapting to the new envirorment; the non-Akan among them had language difficulties; the host- ccrmunities were unfriendly; there was a general absence of such amenities as markets, medical and transport facilities and good drinking water; or the settlements were isolated. The problems were basically financial. With adequate money, a migrant could deal effectively with the problems. Farmers were then asked how they dealt with their problems, and answered as follows: 61% hired themselves out;l9# stayed with friends and relatives; 10% borrowed from friends, relatives and cithers; and 2% fell bade on resources frcm hare such as proceeds from farms, especially cocoa, or rents fran buildings or savings, or the sale of personal property; and 8% did not reply. The 2% who had resources at heme must be included in the group of the well-to-do fanners. A great majority of the rest riust be seen as dependent farmers, whose farming career could be imperilled through lack of adequate attention to their own farms, while working for their living, insecure acccmrodation, or indebtedness. The future of migrants with inadequate capital is fairly predictable. They either fall victim to the demands of a landlord, to whem a tenant farmer is little more than a labourer, or they get into perpetual indebtedness because their output is small and they have inter-seasonal monetary needs. IV. CONCLUSION Migrant cocoa farmers constitute a well stratified peasant group, with their ini- tial capital for farming operations as the major determinant of stratification. The initial capital outlay determines the migrant's relationship to the means of production, land, labour and farm inputs. It makes a crucial difference that a migrant is able to (a) purchase his farm land, (p) hire labour, and (c) secure farming imple- ments and services. A man who purchases land (often more than one plot) attracts labour for hire, and also through well- placed gifts is able to attract more than average attention from those who achiinister the supply of farm inputs. He is also a full-time farmer who devotes full attention to the "management" of his farm business, "ftenagement" includes supervising labour and ensuring supplies of inplements and other inputs. His business includes conti- nuous land purchases, and speculation in land provides him with additional funds for the farm fall, 1963a, 1970). The "business" or capitalist farmers, perhaps only 5% of the migrant farmers, must be distinguished frcm the tenant or dependent farmers, many of whan are both farmers and labourers. But it is possible to make distinctions within this broad second group, on the basis of relationship 16 to the maans of production and the extent to which a farmer works for others in order to make a living. Fran the response given to questions relating to problems of settlement, it is possible to suggest that about 37% of the migrants belong to a middle rank o farmers who are only semi- dependent, and 67/' to a rank of dependent. farmers who may also be classified as semi- labnurers and form the pool of rural labour. The ultimate basis of the differentiation is the occupational background of the migrants. Rural socio-eccncmic organization in the older cocoa growing areas is trans- ferred to the migrant areas. Capitalist farmers in the older cocoa growing areas and seme townsmen become capitalist farmers in the migrant areas, where there are also the same strata of "upper", "middle" and "lower" ranked farmers as are found in the older cocoa growing areas. MTES 1. This paper Is based on i soclo-econoaic survey of aigrant cocos faraers in Ctntral tnd Western Regions conducted in 1972-73. The rt«eirch t tm combined tilt methods of observation and ques- tionnaire administration through inter- views of farmrs scattarad in ssttle- •ants in Assin, Sehwi and Wasa. Res- ponst to tht Interviews was good. The teea was to It to interview as many as 8T» raspottdants in tht thrat araas in tht araas in tht two ragions. This paptr rafltcts on som of tht Mtarial presented (aimographtd) Ltgon, 1986, Tk* Expaasita tf C«c«« »rt«actie«: Vm tnrkina cwMtttiaas tf aiara*t ctcta farmrs it taw Caatral 8*4 Wtstera ttgioas of Gaaaa. As statad in that volum, ay inforaation froa Asantt Migrant farmrs shows that tht conditions in which thay work have not changtd to any significant extant so that the findings retain their value. Miss Karen Legge, University of Arhi* in Liverpool, read a draft of the paper and aado soat valuable suggestions. 2. For exaaplt as amng the Asanta, see Rattray 1929; and Benneh's survey of 20th in land tenure, lenneh 1970: *3-61. century developmnts 5. On the principles of Aksft land tenure, see Sarbah 1897, (F. Cass, 1968), lattray 1929, Buslt 1951, Ollenu 1962, 1967. •». The tens "priaary" and "secondary" are used Instead of "native" and "non- native" to signify original and non- original occupancy. 5. Land was somtims obtained as a gift after a long sojourn in an area, parti- cularly if doaor and recipient were fellow members of any of the Akan elans; see Arhin 1986: 22. 6. According to Hill, land-purchasing companies art organizations for land purchasing purposes associated with patrilineal peoples, Mill 1963:8. But a "coapany" could be a group of autually acquainted pool their resources together in order to obtain land of the deaired size. farmrs who ?. Answers to interviews showed that 17% (i.e. 1 out of 6) of the Migrant farmrs had been involved in disputes of two kinds; 59% were disputes over ownership of land and 3*t% over breaches of con- tract, and the average cost of dispute settlement was #M).Q0. There was a nuabtr of unreported cases, the farmrs preferring to leave well alone to entangling themelves in litigation. Unrtporttd cases arose froa arbitrary variations in teras of land grants, arbitrary increases in rent, att, and the grant of the mi land to two persons. 8. On labour see also Beckett 19M»: 60. 9. Set Sarbah (1897) 1966: 86-87. 10. Killick 1966: ZV7 gives a short account of the various cocoa diseases, and the aeasures taken to coabat thea. 17 Ghanaian Cocoa Farmers, in Biebuyck ed. African Agrarian Systems. Inter- national Institute/Oxford University Press. African 1970. Studies in Rural Capita- lism in West Africa. Cambridge Univer- sity Press. Hopkins, A.G. 1973. An Economic History of Vest Africa. Longman. Howard, R. 1980. Formation and Stratifica- tion of the Peasantry in Colonial, Ghana, in Journal of Peasant Stufie*' 8, 1. Killick, T. 1966. Cocoa, in W. Birmingham, I. Neudstadt and E.N. Oraaboe eds. A Study of Contemporary Ghama, Vol.1. The Economy. Evanston: Northwestern University Press. Kotey, R.A., C. Okali and B. Rourke, eds. 1974. The Economics of Cocoa Produc- tion and Marketing, Legon: ISSER. Metcalfe, G.E. 1964. Great Britain and Ghana: Documents in Ghana History 1807-1957. T.H. Nelson and Sons Ltd. " Okali, C, Owusu Ansah, M. and Rourke, B.E. 1974. The development pattern of some large cocoa holdings In Ghana, in Kotey, Okali and Rourke eds. Ollenu, N.A. 1962. Principles of Customary Land Law in Ghana. London: Sweet and Maxwell. 1967. Aspects of land tenure, in Biraingiian et. al. op cit, Vol.ii. Some Aspects of Social Structure, pp. 251-263. Rattray, R.S. 1929. Ashanti Law and Constitution, Oxford University Press, London. Sarbah, J.M. 1897. Fanti Customary Laws. Cass. reprinted by Frank London. 1966. REFERENCES Addo, N.O. 197^. $orae employment and labour conditions on Ghana's cocoa farms, in R.A. Kotey, C. Okali and B.E. Rourke editors, Tht Economics of Cocoa Production and Marketing, (pp. 204-218). Arhin, K. 1983a. Rank and class among the Asante and Fante in the nineteenth century, in Africa, Vol.53. 1, pp.2-22. 1983b. Peasants in 19th Century Asante, in Current Anthropology, Vol.4, No.4, August to October. 1986. The Expansion of Cocoa Production: the working conditions of •igrant cocea farmers in the central and western regions of d m. Mimeo- graphed, Legon. Beckett, W.H. 1944. Akokoaso: A Survey of Percy Lund, a Gold Coast Village: Humphries and Co. Ltd. Benneh, G. 1970. The impact of cocoa cultivation on the traditional land tenure of the Akan of Ghana. Ghana Journal of Sociology, Vol.6, No.1. Bernstein, J.H. 1979. African Peasantries: a theoretical framework, Journal of Peasant Studies, 6, 4. Busia, K.A. 1951. The Position of the Chief in the Memera Political System of Ashaati. Oxford University Press. Gunnarsson, C. 1978. The Sold Ceast Cocoa Industry 1900-1939: Production, Prices and Structural Change. Lund. Hill, Polly. 1956. The Gold Coast Cocoa London: Oxford University Farmer. Press. 1963a. Migrant Cocoa Farmers of Southern Ghana. Cambridge Univer- sity Press. 1963b. Three types of Southern 18