4. The structure of economic opportunity

As will have been seen from the previous chapters, it is impossible to isolate completely the 'rural' from the 'urban' or the 'agricultural' from the 'nonagricultural' sectors of the economy among the Yoruba. The settlement pattern and the relationship between the towns and the farm villages mean that many farmers see themselves as town residents, and many have dual occupations. Migration is very common and a large proportion of Yoruba farmers move away from their home areas and into other forms of work at some point during their working lives. Profits made from farming in one generation are likely to be invested in the education of the next generation, and there is a complex network of relationships between the wage- and salary earners in the urban centres and their relatives elsewhere. The towns and the farms are also linked by the complex marketing system and the vast number of petty traders involved in it, many of whom operate over long distances, linking the Yoruba economy to that of West Africa as a whole.

The purpose of this chapter is to examine the major forms of economic opportunity available, and the ways in which they have developed. These are the result of the interaction of a number of major factors: the distribution of population and natural resources; the location of cash crops; the development of the transport system; the growth of education; and the economic policy of successive governments. The 19th century created a belt of high population density across the middle of Yorubaland. When the British arrived, the pattern of trade and administration tended to follow this line. Generally, these were the areas in which the most rapid economic changes took place at first.

A second factor was the growth of the cash-crop economy. In the 19th century, the growth of the trade in palm oil led to a reorganisation of production in the interior and the development of slave estates owned by the powerful warchiefs. It also meant the development of a Saro trading elite in Lagos. After 1880, African merchants were affected by a trade recession (Hopkins, 1973: 154ó7; Berry, 1975: 25ó8), and some of them looked for alternative investments. Cocoa had come to West Africa from Brazil, via the Spanish and Portuguese African colonies, and the crop was adopted quickly in Ibadan and Abeokuta. With the exception of kola, alternatives to cocoa were less successful. A brief rubber boom did not last and attempts to develop the cotton industry were soon abandoned. Cocoa, which involves few economies of scale, was well adapted to production on the small holdings of the Yoruba farmers. The spread of the crop was helped by the success of planters at Agege and Ota, including prominent members of the CMS and African churches (Webster, 1961; Berry, 1975: 41). They made use of labourers from towns in the interior who took the crop back with them when they returned home. In Ijesa and Ibadan, the rapid spread of the crop may have owed something to the search for new opportunities by the former 'war-boys' of the opposing armies after the end of the war (Berry, 1975:49 53).

The growth of the industry brought new patterns of migration. The farmers in Ife, who had adopted cocoa before 1939 were joined in the following decade by Egba and Ibadan migrants looking for new land. Since the 1950s, most migration in search of cocoa land has been to Ijesa and Ondo (ibid: 67ó71). Later migrants were often able to work for established cocoa-farmers from their own home areas in return for help in starting their own farms. Cocoa prices fell drastically after 1929, but the prices of other commodities fared even worse, and people continued to move into cocoa farming through the depression (ibid: 85).

The implications of the rise of the cocoa industry have been profound: innovations in land-tenure patterns and labour organisation; the intensification of inequalities between the forest and savanna areas; innovations in marketing, and the creation of the marketing-board system with its important political repercussions.

The earliest major achievement of the colonial administration was the construction of the railway from Lagos to Ibadan. This was completed by 1900, and its extension to Kano was finished by 1912. The construction work drew large numbers of migrant workers from all over Yorubaland, and some of these later moved to Ghana to work on the railways there (Oyemakinde, 1974). They provided one of the starting points for the later large-scale migration of Yoruba to Ghana. Other wage-employment opportunities developed, especially in Lagos, and there was large-scale recruitment for the army. The social effects of these new openings, together with other opportunities for wage labour in agriculture, were far-reaching: increased amounts of money in circulation, increasing social mobility, and the decline of domestic slavery (cf. Agiri, 1972: Ch. 6). With their proximity to Lagos, Yoruba traders were in a good position to act as middlemen in the trade in imported goods, and they became a familiar sight in the urban centres in the north of the county: Kano, Jos, Kaduna and Zaria.

In the early days of colonial rule, British administrators toured the interior on horseback, or in hammocks carried by local porters. Vehicles had begun to appear in Lagos by 1914, and the first British road in the interior, built in 1905, covered the 53 km between Ibadan and Oyo. After 1918, the road system was gradually expanded. By 1957, 60 per cent of the tarred roads in the country were in Lagos and the Western Region, and 57 per cent of the motor vehicles were registered there as well (Hawkins, 1958). From the start, most of the commercial vehicles belonged to local entrepreneurs. Transport became one of the most popular investments open to Nigerians.

The main arteries in the road network run from south to north: from Lagos to Ibadan, via Abeokuta, Sagamu or Ijebu Ode, and then on to Oyo, Ogbomoso and Ilorin. Major offshoots run from west to east: the short route to Benin through Ijebu Ode, the older route through Ife, Ondo, Akure, and Owo, and the main road through Kwara State, from Kisi to Lokoja. Most of the major roads have now been tarred. Off them lie the numerous laterite feeder roads which fan out into the rural areas, and off these again lie the bush roads and tracks, narrow and unsurfaced, with precarious wooden bridges over which ancient lorries, laden with firewood or yams, make their way. Many of the minor roads are still impassable during the rains from July to November.

There has been a complex interaction between the development of education, changing patterns of urban employment, and levels of migration to the major urban centres. The missions were responsible for the early development of schools and generally the areas in which Christianity spread most rapidly are those which produced the highest numbers of school-leavers. The northern areas where Islam had spread before the colonial period ó Ibadan, Oyo and Ilorin ó had fewer schools, lower rates of literacy and less representation in the administration. Expansion of primary education in the 1950s increased outmigration from the rural areas, particularly to Lagos. The number of migrants has grown much faster than the number of jobs, and this remains a major problem.

What sort of occupational structure has resulted from the interaction of these factors? Here again one is hampered by a dearth of accurate census information. In the 1953 census, 67 per cent of the adult men were recorded as farming, together with 69 per cent of the women (Oluwasanmi, 1967: 31). These figures concealed the number of people with dual occupations, and probably underestimated the proportion of women engaged in some form of trade. Since the 1950s, with increased levels of outmigration from rural areas to the commercial, administrative and industrial centres, the proportion of the population in agriculture must have declined, particularly in the younger age-groups, and the average age of those remaining in farming has increased. There are, however, considerable variations in occupational distribution, even in the rural villages. In Berry's sample of cocoa-farming villages, the proportion of men farming varied from only 60 per cent in Araromi-Aperin, near Ibadan, to over 90 per cent in Orotedo, near Ondo. Araromi-Aperin was an older settlement, and while good cocoa land was scarce, there had been an influx of migrants with other occupations. Few women did any farm work, apart from food-processing like making palm oil or breaking-open cocoa pods at harvest time (Berry, 1975: 174ó 5). In Orotedo the great majority of women helped on the farm, and a smaller percentage were traders.
The division of labour in the towns is more complex. Barbara Lloyd found in Oje, in Ibadan, that only 10 per cent of the adult men were farmers. The majority were craftsmen (50 per cent), traders (18 per cent), and professional and technical workers (17 per cent) (1967: 71). Marris (1961: 68) and Baker (1974: 41) reflect a similar situation in Lagos, though here the category of farmers almost disappears. 84 per cent of the women in Oje, and 70 per cent of the Lagos women were traders.

The sexual division of labour

While the majority of Yoruba men are farmers or craftsmen, women are more likely to be involved in trade. On the farm, the women sometimes grow vegetables or help with transporting the produce. In the border area between Ife and Ondo, Clarke reports that women receive an annual sum of money from their husbands in return for their work on the farm (1979: 364, 371). Their main role, however, is usually in processing or selling foodstuffs. Though the husband is responsible for his wife's debts, he is expected to give her the means for her to work for herself, and many women are provided with trading capital by their husbands on marriage. After some years of marriage, the wife will probably want to pursue her own enterprise to further the interests of her children. If she is a trader, this may involve increasing periods of separation from the husband, who may acquire a second wife to take over the domestic chores. The ideal husband is one who lets his wife get on with her own career, with no obstacles. As a woman's children grow up, one of her main concerns becomes financing their education, and defending their interests against those of the children of her husband's other wives. Thus for most Yoruba women the roles of wife, mother and trader are closely connected (Sudarkasa, 1973).

The household budget is divided between the spouses, according to the income of each. The husband usually provides the house and repairs it if necessary. He also usually provides staple foods and some money for education and children's clothing. The wife provides her own clothes, the rest of the children's clothes, and other items of food. When the husband is away on the farm, his wife may be almost entirely responsible for feeding herself and her children. At some times of the year, women are more likely to have ready cash than the men. Many Yoruba wives go to great lengths to pay for their children's education if their husbands are unable to do so. Normally the father tries to treat the wives equally by sending an equal number of children by each of his wives to school. If any more go, it is because the mother has been able to find the money. It is common for a child's career at school to be interrupted for months or even years if the parents cannot find the money, or if either parent dies.

For the first few years of marriage, then, a woman is mainly concerned with bearing children, and is most likely to accompany her husband to the farm, to cook for him and increase her own chances of conception. When the children are still young, there is little pressure on her to increase her own level of trade. When the children are at school, more money is needed.

Polygyny fits in well with this cycle. High on the list of priorities from many Yoruba men is a second wife who can take over some of the household duties, leaving the senior wife to trade. If the senior wife continues to bear children, she will probably need extra help, and this is provided by child fostering. Children of school age can be sent to live with their parents' relatives, while the mother looks for a junior relative old enough to help her in her work or in looking after the younger children. (Failing this she may hire a housemaid, as is common among women in salaried jobs.) Finally, once a woman has grandchildren, she may help look after them, but the older she grows the more dependent on her children for support she becomes (Sudarkasa, 1973: 132ó44). Hence the importance of many of the remittances which find their way back to the home towns from migrants abroad (Adepoju,1974).

Van den Dreisen found a relationship between polygyny and farm size in Ife, (1972). In 1952, Galletti et al. had found that nearly three-quarters of the household heads interviewed had two or more wives (1956: 73). By the time of Van den Dreisen's survey a decade later, the proportion had fallen to just over half. Men with over 3 hectares (7.5 acres) of land were generally polygynous, while those with less than 3 hectares were generally monogamous. However, even among those with less than 3 hectares of land, men with holdings separated by 32 km or more were usually polygynous. When a man's holdings of land are sufficiently scattered, the pressure on him to set up a second household apparently increases.

Agriculture

Most Yoruba farms are small, and the size is limited by the available labour and the level of technology. The main tools are the hoe and bushknife, and manpower is usually the only energy input. The use of draught animals is impossible because of tsetse fly, and mechanisation is also difficult because of the small size of plots and the pattern of shifting cultivation. The use of fertilisers and chemicals is for the most part restricted to the cultivation of cash crops like cocoa and tobacco. The fertility of farm land was traditionally maintained by a long period of fallow after only two or three years' cultivation, though in many areas this has been modified because of increasing pressure on land.

The Yoruba method of reckoning farm size is not in terms of area, but in terms of 'heaps', the mounds of earth prepared for the cultivation of yams and other food crops. The density of cultivation varies from area to area depending on the fertility. The number of heaps is often reckoned in multiples of 200, or around 1/15th or 1/20th of an acre (Berry, 1975: 135ó6; Guyer, 1972: 45ó6). Guyer found in Ibarapa that most farms consisted of 10ó20 plots, usually adjacent or in 3ó4 separate groups. A plot of 8ó10 units of 200 heaps was considered large. She calculates that a farmer working on his own can farm a maximum area of 12 000 heaps (about 1.2ó1.8 hectares or 3ó4.5 acres). In Berry's sample of four villages in the cocoa belt, 47.5 per cent of the farmers farmed 2 hectares (5 acres) or less (1975: 180). Coupled with the small size of the plots is the fact that farmers often have widely scattered holdings. This is well documented for Egba and Ife, where some farmers' plots are separated by over 30 km.

Labour units are also typically small. Most adult men farm independently with help from their wives and children, though they may hire labourers when necessary. According to Ojo, patrilineal group farming in which all the men in a lineage worked together under the direction of the oldest man used to be common, particularly in Ekiti, but now it has almost vanished. So have the institutions of aro and owe. Aro consisted of groups of kin or age-group members who helped each other on a rotational basis, especially to clear new land in the dry season or to help with weeding during the rains. The owe groups were larger, involving a hundred or more agnates and affines, who worked in return for food, palm wine and kola (Ojo, 1966b: 59ó61). With the growth of education, it is becoming increasingly difficult for farmers to retain the labour of their children on the farm, and this has increased their reliance on hired labour. Even by the time of Galletti's study, 40 per cent of the labour on cocoa farms was hired (1956: 668).

Substantial numbers of labourers come from other parts of the country, particularly from the Niger-Benue confluence area, and, since the end of the civil war, from the Igbo areas. But much of the labour used in many areas is local. According to Berry, many farmers in the cocoa areas also work part of the time for other farmers (1975: 131). Hired labourers work either on an annual basis, in return for food, lodgings, and a lump sum at the end of the season, on monthly or daily rates, or on a piece-work basis. The farmers who needed regular help preferred to hire workers on an annual basis as the rates of pay were much loweró£15 to £36 (N30 to N72) per annumócompared with the five shillings (50 k) or more which a daily labourer could make (Berry, 1975: 131 4). In a tight labourmarket, the workers may refuse to work on an annual basis. In the study by Adelabu and Cook (1975) rates of pay for all hired labour vary widely, both between areas and between different jobs. Probably the importance of non-Yoruba labourers has increased since the 1930s as many of the Yoruba migrant labourers in the cocoa belt have either become established as farmers or moved into other occupations (ibid: 147). Hired labour is probably less important in the savanna, though some of the wealthier Igbeti farmers did use it to expand their food-crop production for the market.

The major food crops are yams, maize, cassava, beans, cocoyam and guineacorn. Rice cultivation is spreading in some areas like Egba and northern Oyo, while plantain and bananas are important in the forest areas. A wide variety of other vegatables is grown on a smaller scale. Yams are the major crop in Ilorin, Kabba and Oyo, all mainly savanna areas of sparser population. Cassava is more important in Ijebu, Ibadan and Abeokuta. Guineacorn and millet are grown only in the savanna.

Yams, cassava and maize provide the basis of the diet of most Yoruba. Yams grow best in rich soils and are normally planted on freshly cleared or fallowed land. In the forest areas, land suitable for yams is often used for cocoa instead. Yam cultivation is thus in steady decline, and the forest areas are becoming more dependent for their yam supplies on savanna areas to the north (cf. Clarke, 1979: 28). A number of different varieties are used which mature at different times and which are suitable for different types of soil and preparation. Yams can be eaten roasted or boiled, but they are usually boiled and pounded in a mortar to make iyan, probably the most popular staple dish. They can be made into yam flour (elubo) which is much easier to store and transport. Pieces of dried yam are ground and mixed with boiling water to form a thick, grey porridge, amala, which is eaten with a stew.

Cassava contains less nutrients and its taste is liked less, but it thrives well in poorer soils and demands less attention during cultivation. As a result cassava has increasingly replaced yam as the major food crop in the forest. It can also be left in the ground for a long period after maturation, so there are fewer storage problems than with yams. It is normally eaten in the form of eba, which is made by grating, fermenting and drying the tubers.

Recent official figures suggest that maize is becoming steadily more important. According to one estimate, 40 per cent of the acreage planted in 1975 was under maize, compared with 23 per cent for yams and 17 per cent for cassava, while in 1972ó3 the total production of maize was greater than that of either yams or cassava. A rise in maize production might be expected, as it requires less labour than yam cultivation, and is more suitable for mechanisation.

A variety of crop-rotation patterns are followed. A common pattern is to plant yams on newly cleared land and to follow this after a year or two with maize or cassava, before abandoning the land to fallow. Intercropping is common. It reduces the labour input and increases the productivity of the land in cases where the different crops require different nutrients. Maize can be planted between the yam ridges, and in Igbeti it was common to plant the slower-maturing guineacorn along with the first of the two crops of maize. Beans, cotton and gourds could be interspersed with the second maize crop. Gourds are particularly useful. The dry outer rind can be made into bowls, bottles and other containers, while the dried seeds (egusi) form the basis of one of the most popular types of stew.

Yoruba food is extremely hot, even by West African standards. The main meal of the day, eaten in the afternoon or early evening, consists of one or other of the main staples together with a vegetable-based stew, prepared with palm oil, and a small amount of meat or fish fried with oil, tomatoes, onions and peppers. The men and women eat separately. A housewife who also trades may buy most of her meals from cooked-food sellers, including the evening meal if she does not have to cook for her husband. Many of the most important Yoruba recipes are laborious to prepare, and the cooked-food sellers specialise in producing one or other of themófor instance akara, the beancakes eaten for breakfast, and eko, a white solid pap made out of ground maize. Most towns of any size have at lease one 'chop bar' consisting of a lean-to shed with tables and benches and a proprietress who presides over pans of stew simmering on the wood fires. In the evenings, the variety of cooked food available usually increases as other women appear by the sides of the street selling fried plantain or yams, fish, beancakes and roast corn-cobs.

Cash crops

In one sense the distinction between food crops and cash crops is irrelevant in the Yoruba case, as most of the farmers dispose of at least some of their crops on the market. In Igbeti, where land was plentiful, those who could afford it were expanding their yam production for the market using hired labour. In some northern areas tobacco production for the cigarette companies has spread rapidly in the last twenty years, mainly because of very successful extension work by the Nigerian Tobacco Company (Harrison, 1969). Cotton is grown in many areas, for local use rather than for export, but the two major cash crops, cocoa and kola, are only produced in the forest areas.

Kola is grown almost entirely for the Nigerian market. Much of the crop is bought by Hausa buyers stationed in the Yoruba towns and villages, for sale in the northern states of the country (Cohen, 1966: Agiri, 1972) where it is the major stimulant permitted by Islam. There are two varieties, nitida, known in Yoruba as abata, which is indigenous to the area, and acuminata, known by the Yoruba as gbanja, which is the main variety exported. Before the colonial period, most of the gbanja kola sold in northern Nigeria came from southern Ghana. It was brought by long-distance trading caravans via the Gonja kingdom, from which its name derives. It was introduced into Yorubaland as a cash crop in the 1890s and by 1930 most of the supplies for the north came from the Yoruba areas. Abata is more important in Yoruba ritual. Kola has proved a useful crop in some of the forest areas like Agege, Ota and Ijebu where cocoa was introduced very early on but did not perform well.

Western Nigeria produces substantial amounts of palm oil, though in recent years the amount exported has dwindled. The most efficient way to extract the oil from the palm fruit is to use a mechanical press, but many women still use the traditional labour-intensive technique. The fruits are boiled and then pounded, and the resulting mash is boiled again. The oil rises to the top and is scraped off. The equipment is cheap, opportunity costs for the women are low, and the residue can be used for lighting fires. The kernels are cracked and used separately to make palm-kernel oil. In some areas, the palm-oil industry has been taken over by Urhobo and other migrants from Bendel State, who lease the right to reap the wild fruits from the local farmers. Some of the migrants have now moved into farming, land-owning and money-lending (Adegbola, 1972).

The cocoa industry developed first in the western areas: llaro, Agege, Abeokuta, Ibadan and Ijebu. As the trees in these areas have aged, the centre of production has gradually moved east to Ondo, Akure and Owo, and in some western areas kola has replaced cocoa entirely. Cocoa grows best in loamy soils and on freshly cleared forest land. The trees take about seven years before they start to produce, and food crops are usually planted to provide shade until the saplings are established. The average life of the trees is around forty years, but productivity declines and many of the trees die before this. If the tree canopy is broken, the rest of the trees deteriorate more rapidly. Thus a continual search for new land suitable for cocoa cultivation is necessary, and migration, first to Ife, and later to Ondo, has been the result. Even so, cocoa represents a much more permanent investment than do the staple food crops, and the land on which new plantings can be made is becoming increasingly scarce. Both these factors have implications for the developments in patterns of land tenure.

Land tenure

Land tenure is a question of great complexity and only a very schematic account can be given here. Three important points should be made at the outset. First a distinction has to be made between the right to use land and rights of full ownership, particularly the right to alienate it. In many cases throughout Yorubaland, the two do not coincide. Second, as land becomes more valuable, either because of its increasing scarcity or its potential for cash crops, conflict over access to, and control of, land will increase, and an increasing quantum of rights will be asserted over it. As Lloyd remarks of Ondo land tenure (1962: 131), 'while land has little scarcity or commercial value it will be described as communal: but as soon as it becomes valuable the descent groups currently using it will begin to claim rights amounting to full ownership'. In different areas of Yorubaland, ownership of land is variously thought of as being vested in the ruler on behalf of his community, as being vested in descent groups, or as being vested in individuals. In Ondo, the position, according to Lloyd, is that the ruler claims that all land belongs to him on behalf of the community. 'An Ondo man may farm anywhere in the kingdom provided that if he farms within one hour's walk (3ó4 miles) of a subordinate town he must get the permission of the oloja (ruler) of that town; he may not be disturbed in his possession of land, but should he abandon it the land reverts to the community' (1962: 110). The Ondo rules may reflect Benin influence: in the Benin kingdom, land is vested in the community rather than in descent groups. However, it appears that Ondo descent groups have asserted rights of ownership over particular tracts of land, particularly on the perimeter of the capital and in areas suitable for cocoaplanting, where they are demanding annual payments (isakole) for the use of the land.

Third, a sharp distinction is to be drawn between the rights that a member of a kingdom can have in its land and the rights which can be acquired by an outsider. In many cases, outsiders can become tenants, but cannot claim rights of ownership over land, and as the scarcity of land increases, the more rigidly this rule may be applied.

Descent group control over land is more usual. This is the pattern in Ibadan, Ijebu and Ekiti. Within the descent group, land is allocated according to need. A farmer can use land allocated to him and can pass it on to his children, but usually he cannot alienate it without the permission of the descent group as a whole. In the case of large descent groups a process of partition has often taken place: the land is divided between segments which can dispose of it without reference to the other segments, and this process of fragmentation has reached its fullest extent in Egba, where it is common to have land rights vested in individual farmers (Lloyd, 1962: 84-5, 241-2).

In Ibadan descent-group control of land has remained strong, despite the early introduction of cocoa. In the l9th century the leading warriors and their followers claimed large tracts of land, and they maintained control over them into the colonial period, establishing a tradition of corporate ownership. On the introduction of cocoa, hunters who acted as guides to Ibadan farmers looking for land in the forest suitable for cocoa began to claim rights of ownership over it themselves. Their ownership was marked by the initial gifts they received from the farmers, and by annual payments of isakole thereafter. At first these were nominal payments recognising ownership, rather than an economic rental. The tenants could remain permanently, and pass their usufructuary rights on to their children who continued to pay isakole, but they had no right to alienate the land itself.

Land rights took a different form in Egba, where a free market in land had developed before 1880, before the arrival of cocoa. European concepts of ownership had been introduced by the Saro repatriates to Abeokuta. After 1880 recolonisation of land abandoned during the 1820s was stimulated by the introduction of the new crop, and the elders of the abandoned towns, now living in Abeokuta, began to reallocate land to individual farmers in return for gifts. These transactions came to be regarded as sales. Land sales became so common that steps had to be taken to restrict them to members of the kingdom. With the division of Abeokuta into townships and the complex political institutions cross-cutting lines of descent, powerful descent groups on the Ibadan model did not emerge. Partitioning of land in each generation has meant that some land has been inherited through women as men without sons have passed on their holdings to their daughters. Taken together, these factors have led to a situation in which farmers have widely scattered plots separated in some cases by 30 km or more. Plots which a farmer is unable to use may be let to tenants, for foodcrop cultivation rather than cocoa-planting.

Individual ownership may result from other processes. In a recent study of the Ife-Ondo border area, Clarke (1979) found that individual tenure had developed through individual colonisation and appropriation of forest land from the 1930s onwards, without a stage of descent group or communal control. As all the available land in this area has now been claimed, so migrants coming in to establish cocoa farms have been forced to enter into tenancy agreements with established farmers. A similar process has been taking place in Ife, where the proportion of cocoa-farmers who were tenants rose from 15 per cent to 39 per cent between 1952 and 1968. As the available land has been used up, so immigration has slowed down, and in some areas stopped completely (Van den Dreisen, 1971).

At the present time, therefore, access to land, both for food crops and tree crops, can be gained in a number of ways. In areas with low population densities and abundant supplies of land, members of the community (or less often outsiders) can make free use of it, and descent groups and individuals have not yet established claims to unoccupied land. This is still the situation in the more sparsely populated savanna areas of northern Oyo. It was also the position in the Ife-Ondo border area in the 1930s, and was legally still the position in Ondo in the 1950s, though descent groups were beginning to assert more control. The situation here is moving towards that found in other areas, where land is obtained from a descent group or by payment of isakole as a tenant.

A variation on the theme of descent-group ownership is found in Ijebu, where the cognatic elements in the kinship system are more pronounced. By claiming membership of more than one land-owning group, a farmer can gain access to land in different areas. The Ondo data suggest that rights in land can be passed on through women there as well, but that the situation is changing with the strengthening of descent-group control. Lloyd (1970: 312) has suggested that, with the increasing scarcity of valuable categories of land, sons are trying to claim an exclusive right to it at the expense of those related through women. This may be a fairly widespread trend. In Ibarapa, Guyer found (1972: 69ó74) that while land for food crops was passed on freely through men and women, the scarce cocoa land was kept more strictly within the agnatic descent group.

The major alternative to obtaining land from one's own descent group in many areas is to 'beg' it (toro) from another group, often in return for initial gifts (isaigi) and annual payment of isakole. The categories of people required to pay isakole vary. In Ibadan, all tenants, whether from Ibadan or not, were required to pay it at an early date. In Ife, isakole was only demanded in the 1930s, as the number of migrant farmers from elsewhere increased. In both northern Ijebu and Ondo, it seems that migrants from other parts of the kingdom do not have to pay it, but migrants from other kingdoms increasingly do (Lloyd, 1962: 182; Berry, 1975: 98ó9). Isakole payments initially bore little relation to the area of land cultivated. In Ibadan, early migrants paid it in the form of gifts of foodstuffs or services to the land owner. The payments tended to remain small, though isaigi payments rose rapidly. In Ife, the isaigi payments were smaller, but since the 1930s isakole payments have been made in the form of a fixed weight of cocoa beansóusually 0.5 cwt or 1 cwt (25.5 kg or 51 kg). Thus they have increased along with the price of cocoa (Berry, 1975: 104ó11).

Rights in land and cocoa farms may also be purchased, though the extent to which an actual land market has developed varies. In Egba, sales of land were recognised very early on. In Ibadan and Ife actual land sales are much less common. Where farms are sold, it is usually rights over the trees which are transferred, not rights over the land: the new owner simply continues to pay isakole. In Ife, sales of cocoa farms have been recognised since the 1930s, but in Ibadan many deny that cocoa farms may be sold at all. The difference appears to arise from the different origins of the tenant farmers in each case. Those in Ibadan are often Ibadan people themselves, and could claim ownership of the land. The fear is that if rights in the trees are transferred, the new owner may claim he has purchased rights over land as well. In Ife, where most of the tenants are migrants, this does not arise (Berry, 1975: 112ó13). In Ondo, where in theory the land was vested in the ruler, sales of cocoa farms between members of the kingdom developed without restriction. With the arrival of large numbers of migrant farmers, the rights claimed both by individual local farmers and descent groups over land are likely to become more comprehensive and more jealously guarded, with increasing demands for isakole payments.

Thus Yoruba land-tenure patterns in some areas have increasing political implications. First, their existence reinforces the attachment of individuals to their home towns. Secondly, they make it difffcult for strangers to become assimilated in the areas in which they have settled. Yoruba migrants and their descendants in the cocoa belt tend to remain 'strangers' (alejo) if they come from another kingdom, even when they speak the same dialect of Yoruba, and even when, as in the case of Modakeke in Ife, they have been there for a century or more (Berry, 1975: 113ó16; Oyediran, 1974).

A final means of access to land is through sharecropping. This is becoming increasingly common as owners or inheritors of cocoa farms are unwilling or unable to manage them themselves; as wealthy entrepreneurs see cocoa holdings as a good investment and require labour to run them; and as other opportunities for migrant farmers to find land are restricted. Sharecroppers take over the trees and land on behalf of the owners, many of whom are traders, civil servants and teachers (Abaelu and Cook, 1975). The owner often provides the seed, chemicals and accommodation as well. In the past, the proceeds were split between the sharecropper and the owner. Now it is usual for them to be split three ways: the owner and sharecropper each take a share, and the third goes on chemicals and pesticides. In some areas farm-owners are commuting their third of the crop into a cash payment agreed in advance (Berry, 1975: 133)óa system which assures them of a regular income, reduces their responsibilities and provides the sharecropper with greater incentive to raise productivity. Anyone in a position to buy, or be granted, large tracts of land on which to install sharecroppers in this way is assured of a large regular return, for which he has to do very little after the initial clearing and planting. It is also the wealthier land-owners who are in the best position to capitalise on the recent cocoa-replanting programmes, the introduction of higheryielding varieties, and the higher costs of chemical inputs which they require.

Problems of the agricultural sector

Despite the appearance of affluence in many cocoa-producing areas, the agricultural sector of the Yoruba economy is largely stagnant. Although in places like northern Oyo the population density is still low, elsewhere there is some pressure on land. Galletti noticed in 1952 that there was no 'free' land in the cocoa areas where the population was densest, and that food farms were not being given sufficient rest in fallow (1956: 279ó80). Callaway found in Ola, near Ogbomoso, that the period of fallow had been reduced to between three and five years by the 1960s, and farmers were aware of the declining size of their new yams. Storage houses were becoming smaller, and for many people there was an acute shortage of food before the next harvest (1969). This is unusual. The Yoruba generally have an adequate diet, and a rich and varied one by African standards. In Olusanya's study of three villages (1969) 80 per cent of his sample felt that they had insufficient land. Van den Dreisen found a similar shortage of land in Ife, coupled with increasing inequality in its distribution (1971). Despite the high rates of outmigration in many areas, rapid population growth and a net increase in the rural population mean that the problem is likely to become more acute unless productivity is increased or new methods of land rehabilitation are adopted. However, the increasing age of the rural population, the declining cocoa yields in many areas, the fragmentation of holdings and the difficulty of obtaining credit are some of the factors preventing innovation.

Credit institutions in the traditional Yoruba economy were well developed. They included the use of pawns, iwofa, and the esusu rotating credit associations. In the iwofa system, a loan was secured in return for the labour of the debtor or one of his relatives. The labour served as the interest on the loan until it was repaid (Fadipe,1970: 189ó33). The system was abolished by the British in 1926 (Agiri, 1972: 162), but esusu groups still flourish throughout Yorubaland. The basic idea is that a group of people make regular contributions to a central fund. After each person has contributed once, the entire amount is given to one of the members as a lump sum. The process is repeated over and over again, until each member of the group has received his share (Bascom,1952). Esusu are organised within compounds and descent groups and other types of associations. But they are most useful to those who can make regular contributions, and craftsmen and traders are more likely to do so than farmers. For many farmers, the only sources of credit are the money-lenders, and the only security they can offer is rights over land or tree crops.

It is difficult to assess the extent of rural indebtedness, though there is some evidence that it has increased since the 1950s (Galletti et al., 1956: Ch. 15; Essang, 1970; Van den Dreisen, 1971). In 1956, Galletti concluded that the level of indebtedness was tolerable, but this was in a period of high cocoa prices and increasing production. A major source of the money owed (63 per cent) consisted of interest-free loans from friends and relatives. Other sources were produce-buyers (19 per cent) and money-lenders (12 per cent). Interest rates otherwise were high. Money-lenders often charged between 5 and 12 per cent per month for secured loans and up to 25 per cent per month for unsecured loans, compared with legal rates of 1.25 and 3.75 per cent respectively. The effective rate of interest on loans from produce-buyers was also high, between 60 and 166 per cent per annum, as the loans were often repaid in produce at low prices agreed in advance.

In the late 1960s, the situation for many farmers had probably become more difficult. Cocoa prices had been low for some years, and the civil war imposed a further financial burden. Van den Dreisen thinks that the percentage of 'hardcore' debtors, unable to clear their debts in the year, had increased, and debts were taking longer to pay off (1971: 43ó4). Essang's material suggests similar conclusions. For many the producebuyers were now the most important source of credit, despite the high rate of interest. Many of the licensed buyingagents had invested in cocoa farms themselves. Their average annual cash income from cocoa-farming was over six times as large as that of the full-time cocoa-farmers (1970: 43). It is against this increasing inequality in the rural areas that the unrest of 1968ó70 has to be seen.

An important factor influencing the level of indebtedness has been the reduction of producer incomes through the activities of the commodity marketing boards, which have operated in the export market since 1939 (Bauer, 1954; Essang, 1972; ,Olatunbosun, 1975: Ch. 2). For the Yoruba, cocoa-marketing is by far the most important. In 1939-40, the British Ministry of Food agreed to purchase the entire crop from the large expatriate firms who bought it from the Nigerian producers. In 1942 the system was institutionalised, with the creation of the West Africa Produce Control Board, and in 1947 the Nigerian Cocoa Marketing Board was set up. In 1954, the commodity boards were restructured and put under the control of the regional governments, and the Western Region Marketing Board was created. This was the crucial change which provided the regional governments with major sources of funds in the 1950s. With the breakup of the regions, there have been a number of other reorganisations since then.

The early rationale given for the operation of the boards was the stabilisation of producer incomes, to be achieved by insulating them from fluctuations in the world market price. After 1947, the boards started to build up large surpluses by systematically paying the producers less than the world price. Generally these surpluses were not used to keep up the producer price in bad years: almost invariably if the world price fell, the producer prices were also lowered. The net effect was to transfer a large proportion of producer incomes to the government: 30 per cent between 1959 and 1965, and over 50 per cent between 1965 and 1969. The main function had changed from price stabilisation to the mobilisation of capital for the government.

There has been considerable debate over the economic, political and social effects of taxing the producers in this way. Bauer argued in the 1950s that the system was depriving the economy of a valuable stimulus. On the other hand, in the 1960s Helleiner concluded that the boards had been relatively successful. While agreeing that there was a problem of equity in taxing producers in this way, he argued that taxation through the boards had been cheap and easy to administer, and development capital had been mobilised which might have otherwise been dissipated in increased producer consumption (Helleiner, 1970). The debate has continued, and the 1970s have seen further reorganisation of the boards.

The case for the boards' pricing policy was stronger when the agricultural sector was still the major source of government revenue. With the growth of the oil industry it has become much weaker, especially in view of widening income differentials and periodic rural unrest. It is, though, unlikely that these long-established government institutions will be eliminated completely, whatever changes are made in pricing.

Cocoa-marketing has had other implications. It is based on a large number of middlemen collecting and bulking the produce for the larger enterprises. The main change over the years has been at the top of this hierarchy, with local entrepreneurs replacing the expatriate firms as the licensed buying-agents for the boards. A small but significant percentage of the crop is purchased by farmers' cooperatives which have developed, with varied success, in the cocoa areas. Below the licensed buying-agents come a group of intermediate buyers, and below them sub-buyers and 'pan buyers', buying in small quantities at the local level. The lower levels of the hierarchy in some areas have been eliminated with the improvement in communications. The system works on a series of cash advances made by higherlevel traders to the lower-level traders, and, ultimately, to the producers. The rewards to the larger buyers are considerable. For the smaller buyers, however, competition at the farm level may be intensive. Essang found that in his sample, the licensed buying-agents in Ondo had incomes nearly fifteen times as high as those of the full-time cocoa-farmers (1970: 43).

Despite the importance of the rural sector in the Nigerian economy, it has generally received low priority in development planning. Olatunbosun has calculated that between 1960 and 1974, it received only 20 per cent of total government expenditure, while it produced 50 per cent of the national output, and employed 80 per cent of the total population. In the 1946ó56 development plan, only 6.5 per cent of the funds were allocated to primary production: this increased to 13 per cent in the 1962ó8 and 1968ó74 plans. Agricultural production between 1959 and 1970 remained virtually static (Olatunbosun, 1975: 15,49ó69).

In the Western Region, the funds that were allocated were often spent on capitalintensive schemes, such as the Western Region Farm Settlements, which were expensive and involved only a small number of farmers (Wells, 1974: Ch. 10; Olatunbosun,1971). This scheme appeared attractive as a possible solution to the unemployment problem which was already apparent in the 1950s. The model used was the Israeli moshav, a group of independent producers, selling their produce and buying supplies through a cooperative. Plans were made to start thirteen settlements, each with 200 farmers, farming mixed plots of tree crops and food crops, and running small-scale poultry units. The government would furnish community facilities, roads and houses, and pay the farmers a daily allowance until their tree crops were established. The settlers were to have primary education plus two years of training in a farm institute. It was hoped that after six years, the settlers would be able to start repaying the costs of setting them up. The number of settlements was increased to nineteen in 1964.

The projected returns to the farmers turned out to be wildly optimistic, partly because of the slump in the cocoa price. The problems and cost of setting up the settlements had been underestimated. By 1968, only 3 per cent of the costs had been repaid. The settlers regarded themselves as government employees rather than independent producers, and morale was low. Only 38 per cent of the initial recruits were left in the scheme by 1971. Needless to say, the 'demonstration effects' of the scheme on the practices of other Yoruba farmers were nil.

The general picture of the rural economy painted so far has been a gloomy one, the more obvious features being declining productivity, rising pressure on land, increasing outmigration, an ageing labour force, lack of investment and increasing social inequality. The opportunities open to the majority of Yoruba farmers are limited. For a small number of entrepreneurs, farming large areas of land with hired labour, mechanisation or sharecropping can be a profitable investment, but for the great mass of the rural population, these possibilities do not exist.

Trade and marketing

The Yoruba have a wide reputation for their skill in trade, both throughout Nigeria and elsewhere in West Africa. Some of the earliest accounts we have of the area speak of thriving craft industries and a complex division of labour. The development of urban centres produced a marketing system in which agricultural produce, craft goods and imported goods changed hands, and much of this trade, especially in the daily markets of the towns, has traditionally been handled by women.

As well as the daily markets, there were the periodic markets which served wider areas. Trade was an important issue in international relations. Some towns were termini on the long-distance trade routes that linked the Yoruba kingdoms with the Akan to the west, and with the Bariba, Hausa and Nupe to the north. Trade on these routes was well-organised. Roads were often wide and well maintained, and caravanseries were established outside the main towns. Trade and tolls provided a major source of revenue for the political authorities along the route (Mabogunje, 1968: 79ó90).

Some features of the marketing system have survived to the present. Daily markets in the towns and periodic markets in the rural areas are still the basis of the distributive system (Hodder, 1969). The pattern of long-distance trade in the 19th century has given way to a Yoruba diaspora in the 20th. Yoruba traders have settled in large numbers throughout West Africa. The large daily markets of the major commercial centres supply goods not only to the local consumers, but also to traders from other towns. Good examples are the major Ibadan markets, some of which are quite specialised, such as Gege and Orita Merin in the trade in foodstuffs, and Oja Iba in the trade in kola (Hodder, 1969: 104ó9). There are also some specialised urban periodic markets like Oje in Ibadan, where sales of Yoruba cloth and locally made soap alternate. Ibadan, like most Yoruba towns, also has a number of small night markets, scattered through the town, selling mainly foodstuffs and cooked food.

Outside the large towns are the 'rings' of village markets organised into four- or eight-day cycles, with a different market being held on each day. The best documented of these is the Akinyele ring, an eight-day cycle to the north of Ibadan (Hodder, 1968: 58ó93), but four-day cycles are more common. Over 80 per cent of the traders in the Akinyele markets were women, and between 50 and 60 per cent of the traders sold foodstuffs. The markets in a ring are evenly spread out, and most rural settlements are within walking distance of one or other of them. Hodder found that much of the produce was still brought in by headload. The existence of market cycles presupposes a relatively dense population, and they are less important in the more sparsely populated savanna areas.

Marketing of this type is very labour-intensive. The commodities involved are of two main types: manufactured goods moving outwards from the major urban centres, and farm produce moving in the other direction. Manufactured goods mostly originate from the large expatriate and Lebanese firms in Lagos and Ibadan. The middle level of Yoruba wholesalers are usually men and women who buy in bulk on regular accounts from the larger firms, and who sell goods in smaller units to the network of retail traders. At the retail level, trading capital is often very limited. Turnover is rapid, and the quantities sold are very small. Many women sell individual cigarettes, matches in bundles of ten, and sugar cubes in piles of two or three at a time.

The Yoruba market traders have much in common with petty traders in other peasant economies, both in their problems and in the strategies which they use to cope with them. Yoruba markets are highly competitive, and traders attempt to establish a more permanent clientele through extending credit. Keeping capital working is vital, and those with sufficient capital prefer to trade at the wholesale level if they can.

In the trade in manufactured goods, one of the most important factors is access to supplies. These are often erratic. With the end of the civil war and the growth of the oil industry in Nigeria, demand has grown faster than the capacity of the main Nigerian ports, and imported goods are often held up at Lagos for months. Shortages of raw materials and the activities of the Price Control Board have meant that supplies of locally manufactured goods are no less erratic: in some cases the imposition of price controls has meant reduction in output and the disappearance of the goods from the open market. In such a situation, good contacts with the storekeepers in the large firms are essential, and the goods tend to go to the wealthier traders who can afford the 'dashes' necessary to secure them.

While the traders in manufactured goods are important in bulk-breaking, the traders in foodstuffs are also important in collection and bulking. Foodprocessing is mainly done by the women, including making eba, elubo and palm oil. The women in Igbeti bought yams on market day, and after two or three days of work they had produced baskets of elubo pieces which they sold on the following market day. The price was determined by the traders who came from Ibadan, Ogbomoso, and Ilorin. If they were offered a price which did not cover the costs of production, the Igbeti women were left with the choice of selling at a loss, or waiting to see if the price would rise again the following market day. By then the pieces of elubo might have shrunk and more would have to be added to the baskets. If the price stayed down, the women had no option but to sell at a loss and try their luck in some other trade, like weaving or making beancakes. As a result of these individual decisions, the amount of yam flour reaching Igbeti market fluctuated markedly from week to week.

In many areas, women are responsible for transporting produce to market, as well as selling it. Some women buyers station themselves on the paths to the farms in order to make a quick deal before the goods even reach the market (Hodder, 1969: 3). Where the farms are nearer to the roads, the farmer himself may make arrangements for the sale of the produce, and the wife may not be involved. In Igbeti in 1970 a farmer wanting to make a bulk sale would usually get in touch with one of the foodstuffs traders in the town. They were often farmers themselves, but they also had the contacts and the storage space to trade on market days, and they could arrange a deal with one of the visiting buyers in return for a commission. If the farmer needed small amounts of money between his larger sales, he could bring a few yams back to his wife to sell along with her other trade goods, but otherwise she would concentrate on her own business. Sudarkasa comments on the down-to-earth economic relations which exist between husband and wife. Where trade is concerned, they do not give each other preferential treatment, and even if the wife is dealing in foodstuffs, if the husband can get a fairer price elsewhere he will do so (1973: 119-20).

The traders who can accumulate larger amounts of capital move either into wholesale or long-distance trade. The dividing line between retail and wholesale trade is by no means clear-cut: many operate at both levels. In the foodstuffs trade, the chain of middlemen may be long. Typically, a buyer at the farm gate sells to a long-distance trader, sometimes through an agent. The goods then go to a wholesaler in one of the towns, who passes them on through one or two more intermediaries to the consumer (Anthonio, 1970). While the majority of traders in Yorubaland are women, men predominate at the higher levels of the trading hierarchy. A man who is a trader may have several years before he marries in which to build up his capital, free from domestic responsibilities. A woman marries younger, and is likely to produce a child soon after marriage. Thus, right at the start of her trading career, domestic responsibilities make it difficult to save and move up the trading hierarchy. In Awe, Sudarkasa found that only a handful of women had the capital to trade between Awe and Lagos or Ibadan, compared with the hundreds who either brought in goods from the surrounding farms or who sold in the town itself. Even for this trading elite, the returns she describes were very small. The woman selling grain in Ibadan who made three or five shillings on sales of £7 or £9 every nine days was typical in the early 1960s (1973: 88ó9). Other studies suggest that profit margins are low, though there are wide variations between commodities (Olayemi, 1974: 41ó4). In general it appears that wholesale traders have higher profit margins than retailers (Anthonio, 1967).

Entry to the wholesale trade is restricted not only by the scarcity of capital, but also the scarcity of storage facilities. In the yam trade, for instance, the harvesting season is long, and some of the produce may be stored for a time on the farm before selling. But if the farmer is short of cash (and many of them are by the time of the harvest) he may have to sell the bulk of his crop at once. The wealthier town trader with storage facilities plays a key role. He has information on the market over a wide area, and it is at this level that the prices are most likely to be determined (Anthonio, 1967: 40;1970; Ilori, 1971).

Another factor which may affect prices, though how far is not entirely clear, is the trade association. In some cases, the association acts as a closed shop, and tries to channel all supplies through its members. Traders coming from outside have to deal with members of the association rather than direct with the consumer. In the case of some perishable commodities brought over long distancesócattle are the most obvious exampleóthe trade is in the hands of a single ethnic group (Cohen, 1965). A monopoly of this sort is easiest to organise when goods have to come from another region of the country, or when there are particular technical problems connected with their packing and transport. It is more difficult in the case of foodstuffs produced within the Yoruba area. Here, the studies from the 1960s suggest that, despite the number of middlemen involved, profit margins are generally low, and the producer gets between two-thirds and four-fifths of the final selling price (Olayemi, 1974: 414ó44). Traders' associations certainly function as pressure-groups to demand better market facilities, but their role in regulating prices needs further research.

Craft production

Many Yoruba occupations were traditionally organised within particular compounds or descent groups, including weaving, smithing, woodcarving, leatherwork, drumming and medicine. Many of these specialisations persist. In Igbeti the best drummers in the town still come from lle Onilu, and facial scars are still made by members of lle Olola.[2] These occupations are mainly confined to men, but others such as pottery, indigo-dyeing and weaving on the upright loom are carried on only by women.

Some of the crafts have survived better than others. There are still Yoruba carvers who produce work of exceptional quality in response to modern commissions (Carroll, 1967) but the craft has declined along with the traditional religion for which most carvings were made. Some palace crafts like leatherwork or calabash-carving in Oyo have been reorganised around the tourist market. Pottery has survived competition from imported enamelware and locally cast aluminium, and is still made in large quantities in Ilorin. But the craft which has perhaps adapted best to the changing conditions is weaving (Bray, 1968).

As with farming or trade, many children help their parents in the crafts and have mastered the skills by the age of 16 or so. Parents were traditionally expected to set up their children in an occupation, provide them with the necessary tools, and arrange their marriages. Until then, they could keep the profits from the children's work, but the child could keep the income >from work done in his spare time and on his own account. The head of the craft in a town was normally the bale of one of the compounds involved in it. Members of the main crafts held regular meetings to discuss prices, sort out disputes and share information on techniques and markets. Taxes were paid to the political authorities in craft goods (Lloyd, 1953).

Kinship is less important in the crafts introduced more recently. Training is usually through apprenticeship, and children are less likely to follow their parents' occupations. Only 9 per cent did so in Koll's sample in Ibadan (1969). Some idea of the extent of the craft and small-scale industrial sector of the Yoruba towns can also be got from this work. About 42,000 of the population of some 900,000 were employed in the craft sector, including some 14,000 qualified craftsmen and 28,000 employees and apprentices. The industrial labour force at this time numbered 3784. The largest category of craft workers were the tailors, followed by the carpenters, weavers, goldsmiths, pepper-grinders and barbers, with smaller numbers of tinkers, motor mechanics, shoemakers, electricians, watch-repairers, printers and photographers. Similar rankings have been found in other studies (Callaway, 1973; Aluko, 1973; Lewis, 1972).

The apprenticeship system is relatively formalised. A contract is usually drawn up by a local letter-writer, with details of the conditions of service and the payments to be made. Apprenticeship lasts on average from three to five years, though this varies between trades. An apprentice short of funds may continue to work for his master for a period after he is fully trained, in lieu of fees. The amounts paid also vary. Callaway in the 1960s found rates of £5 a year for carpenters, goldsmiths and leatherworkers, and up to £10 or £15 a year for photographers (1964: 68-9). The fees are usually paid by the close relatives of the apprentice, though some older men finance their own training out of their savings. The end of the training period is often marked by a 'freedom' ceremony and the presentation of a certificate or diploma of competence. The apprentice can then set up in business on his own and join the local trade association.

The result is a proliferation of independent, small-scale businesses, but patterns of wider cooperation occasionally develop. Nearly all Yoruba towns have their mechanics, working in open spaces littered with the rusting remains of dismembered cars. Koll describes a cluster of mechanics, sprayers, welders and electricians working on one of these sites in Ibadan. The group included 12 craftsmen who, between them, had trained numerous apprentices and who were in the process of training several more. Between them the group offered a comprehensive range of repair services, at prices much lower than those of the commercial firms. Many of the employees of the larger firms operate flourishing workshops at home, where they work after hours, assisted by their own apprentices.

Like the traders, most of the craftsmen are organised into formal associations which attempt to regulate prices and quality and to act as pressure-groups in relations with the authorities. In Ibadan, the associations were apparently strongest when they were recognised by the local government and had a role in tax collection. Some associations, like the motor transporters' unions, can still effectively represent their members' interests, but they are in any case a wealthy and successful group. Other associations tend to be weaker (Williams, 1974: 116). However, Koll argues that these indigenous associations deserve more attention from the government than they have received. The government has usually set up its own organisations and cooperatives to assist small-scale industry, and the programmes have usually been both ineffective and expensive (1973).

The categories of craftsman and trader shade off into those of transportowner, small-scale industrialist and building contractor. Among the most popular enterprises are sawmilling, baking, and printing. Nearly all towns have at least one printer, producing mainly visiting cards, wedding invitations and programmes, bread-labels and almanacs. A town the size of Igbeti could support three small bakeries, each with three employees, and each producing about 200 loaves a day. There is a small group of very wealthy Yoruba industrialists, though in general the Nigerian industrial scene is dominated by government and expatriate capital. Some areas, involving mainly small firms, were reserved for Nigerians by the Enterprises Promotions Decree of 1972.[3] Do the owners of these small businesses represent the nucleus of a group of future industrial entrepreneurs? One problem is that wealthy men tend not to be committed to a single type of business, but prefer to spread their risks. Rather than expanding production in one area, they are more likely to invest in houses, transport or trade where the technical problems are less and where the returns are very high.

Migration

Since the 1950s, migration has become an increasingly important factor in Nigerian life. Green has calculated that between 1952 and 1967, 644.000 people migrated to metropolitan Lagos, including 510,000 people from Western Region (1974: 289), mainly in the 15ó30 age category.

Four main types of migration among the Yoruba can be distinguished. Firstly, there were the unskilled labour migrants of the colonial period, looking for work on the cocoa farms or in the larger towns. Secondly, there were the migrant farmers looking for suitable land, especially for planting cocoa. Many of these included former labourers who eventually became established as tenant farmers in their own right.

Thirdly, there were the long-distance migrants, many of them traders. Trade migration was particularly common from the savanna towns. From Saki, for instance, there has been a massive diaspora all over West Africa, from Guinea to the Cameroun Republic (Mabogunje, 1972: 134ó5), thanks to the lack of opportunities at home (Mabogunje and Oyawoye,1961). Many early migrants were labourers or artisans who had accumulated the capital to move into trade. Others accumulated capital from farming, selling the produce, and buying trade goods with the proceeds.

The best-documented of these migrations is the one to Ghana (Hundsalz, 1972; Sudarkasa, 1975; Eades, 1975a). There are mentions of Yoruba traders in Ghana from the middle of the 19th century onwards, and Yoruba were already established in Accra at the turn of the century. Nigerian migrants were taken to Ghana by the British as troops and railway workers. Some of these moved into trade with their savings. In northern Ghana, there was a flourishing trade in Yoruba cloth, particularly in the periods after 1918 and 1945 when supplies from Europe were scarce. The migrants came mainly from the savanna ó Ilorin, Ogbomoso, northern Oyo and the Ofa area. By 1969 there were probably 200,000 Yoruba in the country. Ilorin migrants were concentrated in Accra, the Ogbomoso in Kumasi, and those from the Oyo towns in the north of the country. By the 1960s, the Yoruba had come to act as middlemen between the large firms and the consumer, dealing mainly in manufactured goods. They provided a large percentage of the traders in the major urban markets, but they had also spread to the most remote rural areas of the country (Hill,1970: 1).

Kinship links were important in the recruitment of the migrants, as the successful traders brought in junior relatives to help them, and eventually set them up in business on their own (Eades, 1975b). The pattern of migration in northern Nigeria is rather similar. The Yoruba migrants here are also from the savanna towns, and they deal in the same sorts of goods. Ije,bu migrants have tended to settle nearer home, in Lagos and Ibadan (Mabogunje, 1967; Aronson, 1971; Akeredolu-Ale, 1973).

The implications of this pattern for the migrants' home towns are rather different than those of other forms of migration. Remittances are often invested at home in houses, transport etc., in the education of other relatives, or to support the elderly. The branches of the town improvement unions in other parts of West Africa contribute heavily to development projects at home, including markets, dispensaries, hospitals, schools, town halls, churches and mosques. Many of the migrants eventually return home to retire in the houses they have built, and Adegbola's study of Osun Division (1972) suggests that they often bring capital or new techniques with them. In general, the areas with large numbers of long-distance migrants appear to be more receptive to change than those without. This type of migration is more likely to involve complete families than is labour migration, and does not result in the same concentration of unproductive dependants in the home town.

Finally, there is the migration of the younger educated people to the larger urban centres, especially since the rapid expansion of education in the 1950s (Callaway, 1967). To some extent this has followed the lines already established by the older migrants. School-leavers from a particular town or village tend to go to the urban centres where their friends and relatives are already established (Callaway, 1969: 52ó3). The extent of the problems produced by this movement was already apparent by the 1960s. In his Ibadan survey, Callaway found that 28 per cent of the male and 15 per cent of the female labour force were unemployed, and that 78 per cent of the unemployed were school-leavers (1967: 201ó4). Interestingly, 78 per cent of them said that they were involved in 'further studies', the most popular of these being typing-lessons: Ibadan had 327 typing-schools at this point! Increasingly, unemployment is affecting secondary-school-leavers as well. In a 1970 survey, Callaway found that of a sample which had left school between 1967 and 1969, a third had been unemployed for over six months at some point since leaving school, and 18 per cent were still unemployed (Callaway, 1975).

Obviously, migration on the scale suggested by Green's figures is an immense drain on manpower in the agricultural sector, and this lies behind the poor levels of productivity in agricultural production. What therefore is the attitude of the farmers themselves towards this exodus of the younger generation? Olusanya in 1969 found that rural household heads appeared overwhelmingly in favour of the boys and young men going to live in large towns like Lagos and Ibadan, though fewer approved of the girls going (1969: 91, 93). Nearly two-thirds of the sample said they would not recommend farming as an occupation to their sons, the main reasons being that it was tedious and financially unrewarding. Callaway commented that the farmers in Ola saw their hope for the future not in expanding their farms, but in the success of their educated sons (1969: 25). Waiting for something to turn up in Lagos or Ibadan is for many preferable to staying home to farm. It means finding a friend or relative willing to locate a job and provide the bribe necessary to secure it in many cases. Peace's work in Agege shows clearly the importance of these personal networks in job-hunting (1979: 33), but even without the necessary contacts, many prefer to stay in the towns.

Conclusion: the opportunity structure

The question still has to be asked: what are the possibilities of mobility between the different niches in the economy, and how have these altered over time? The majority of Yoruba men are still farmers. All Yoruba have access to land at home if they need it. If a farmer is unable to get suitable land from his own descent group, he may be able to 'beg' it from another, but he may decide to migrate elsewhere, especially if he is interested in planting cocoa. He is most likely to follow previous migrants from the same area: this leads to clusters of migrants from the same home town throughout the cocoa belt. However, the chances of becoming established as an independent cocoa-farmer are becoming lower: the costs of access to land are rising, and the migrant may have to make do with sharecropping. For the farmer who has only his land and the labour of himself and his family to rely on, there is the constant problem of finding the cash needed for taxes, rites of passage and education, or to deal with sickness and other emergencies. In the 1950s, friends and neighbours were probably more often in a position to help out than they are at present. Remittances from relatives elsewhere are sometimes important, but they cannot always be relied on. In many cases, they are more than compensated for by cash flows in the opposite direction (Essang and Mabawonku, 1974).

A cash shortage puts pressure on the farmer to dispose of his crops quickly after the harvest, when prices are usually lowest, and after that he may be at the mercy of money-lenders. Tree crops can be pledged as a last resort, but this reduces income and makes solvency even more difficult to achieve. Some farmers can increase their income by planting cocoa, kola or tobacco where the conditions are right. But reliance on cocoa makes the farmers' terms of trade dependent upon the world price and the profit margins of the marketing boards, and the result is severe hardship if the price falls.

What, then, are the possibilities of moving out of agriculture? Until the 1960s, some farmers found it relatively easy. Many of the migrants to Ghana had farmed for a season, sold their entire crop, bought Yoruba cloth, and walked to Ghana with friends and relatives to sell it. After several trips, many had accumulated the capital to buy a bicycle, hire porters or travel by lorry or steamer along the coast. Since independence, opportunities in the market have become more difficult to find. A man whose relatives are already wealthy traders is much more likely to be successful in trade than the farmer attempting to start on his own. For many farmers now, the best strategy is to learn a craft, such as tailoring, cycle-repairing or washing clothes. If a man can persuade a friend or relative to teach him for nothing, so much the better. In a small town, such a skill supplements income from famming. In a larger town, it may provide full-time work, and it can also open up the possibility of long-distance migration to join relatives already established.

But there is a limit to the amount craftsmen can earn through practising a craft. They may try to increase their turnover by taking on apprentices, but apprentices' work is often of a lower quality and loses customers. A few craftsmen may be able to find regular jobs. The mechanic who works for a firm or the government in the morning and then retums to run his own workshop in the evening is an obvious example. He may be able to use his employer's equipment as well. The other major alternative is to move into trade. A cycle-repairer can start to deal in spare parts, or even bicycles, and many tailors also sell cloth or provisions.

Traders range from the woman selling a few kola nuts in front of her house in a farm village to the owner of a large store in Lagos. The amount of starting capital that can be mobilised is crucial, and this can vary from a few kobo to N1000 or more. The smaller the capital, the more rapid the turnover must be, and the more difficult it is to move up the trading hierarchy into the wholesale trade. Competition in the retail trade in the large towns is fierce. The centre of Lagos or Ibadan is full of small children touting ballpoint pens, writing-pads, sunglasses, watch-straps and combs from trays which they carry on their heads. Their parents are often selling in the large markets, where bribes are often necessary to get a stall at all, and where a lot of very profitable, if illegal, subletting goes on.

Success in trade also depends on securing supplies. In the trade in manufactured goods, the main prize is an agency or credit account with one of the large firms. It is often in the interests of the firms to restrict their wholesale customers to a few wealthy traders. Certainly there are many opportunities for extra income for the employees who actually control the movement of goods. Opportunities for profiteering are immense, if one has the capital, in a situation where supplies can dry up at any moment. The further up the trading hierarchy one advances, the more often these opportunities present themselves. The ambition of all traders, both men and women, is to move away from retail trade at the local level and into the more profitable long-distance and wholesale trades. Very few succeed. Family obligations, taxation, stall rentals, and bribes to officials and storekeepers make capital accumulation difficult, and the need to keep capital in circulation reduces opportunities of hoarding and speculation for all but the wealthiest traders. A wide range of other investments are open to the wealthy: housing for rent, transport, sawmills, licensed buying agencies, and tendering for govemment contracts. Wealthy entrepreneurs spend lavishly on houses in their home towns, on rites of passage, on cars and other status symbols, and in acquiring more wives. Many enterprises do not survive the death of their founder. Large-scale industrial investments are few, and houses and vehicles can easily be divided between the heirs.

A final, major, form of investment is education. Children who get through university are likely to get lucrative salaried positions in the professions, the large firms or the civil service, with all the security and fringe benefits which go with them.

At the lower levels of the job market, however, getting a job depends largely on personal networks, and the educational qualifications being demanded for particular jobs are steadily increasing. For many wage earners with limited education, self-employment provides the best opportunities for increased income, and the purpose of a factory job is to accumulate enough capital to move out of it. At the higher levels of the job market, civil servants, professionals and university teachers have substantial salaries and generous loans and allowances which can be invested in housing, farming, transport or businesses in their wife's name. This is frequently supplemented by the types of informal income which officials in strategic positions can often command. The next chapter considers the distribution of political power in Yoruba society, and the nature of the resources available for exploitation.