Centre for Social Policy Studies, University Of Ghana, Legon
seminar presentation by David Korboe, 30th January 1998
Introduction
Over the last few years, a small number of Africanist scholars have begun to draw attention to a curious tenure group common to the towns and cities of anglophone West Africa. Tenure stratification, in the global housing literature, tends to be limited to the two nodal groups - owners and tenants. In the West African sub-region however, there exists a third category comprising rent-free housing consumers - a contingent too large to be ignored by policy makers.
The culture of rent-free housing per se is not new. Across the various regions of world, rural homes have traditionally provided more hospitable access to not-paying patrons. What is of interest here, however is the scale of practice in West African cities. Where the fact of even remote kinship has proved able to invoke free accommodation for the poor. It is the future of this category of housing, known in anglophone West African as the family-house, that this paper attempts to address and preview.
The discussion is in six sections, beginning with a sketch of low-income housing in urban Ghana. The second section describes the Ghanaian family and introduces the concept of family-housing. This provides the framework upon which the main discussion of social costs and benefits is developed in the third section. The fourth section presents the impact of legislation and policy trends on the family-house system, and is followed by an attempt to predict the future of this form of tenure. Specifically, interest centres on establishing - for each factor addressed - who the key interest groups and beneficiaries are. In the concluding section, I suggest that the active interest of academics and policy-makers in this subject is imperative if homelessness is to be preempted.
Housing occupied by the Ghanaian poor is mostly of the traditional compound form. Typically, a completed compound - known in Kumasi as effiehyia - takes the form of a single-story structure with a series of single-banked rooms surrounding a square unroofed courtyard. The compound normally grows by accretion, each of its four wings constituting a phase in the ideal situation. Unlike the structurally, it is occupied once the first few rooms have been built.
At the median, completed compound accommodates some eight households in thirteen rooms on a plot area of approximately 1000 square metres. Most of the rooms are rented out, the owner and other resident households sharing between them what limited services there are. A full half of Kumasi's population live in this type of property, the multi-storey variation, with even more rooms, accommodates another quarter of the city's households (Korboe, 1992)
Entry to all rooms is typically through the unroofed quadrangle which provides a semi-private and climatically sensitive setting for most daytime activities. Sociologists and urban anthropologist - eg. Clark (1984), Goody (1975), Sanjek (1982) and Schildkrout (1975) - document the courtyards as the spatial arena in which social interaction and inter-household cooperation are typically nurtured. On the whole, the poor and the more traditionally minded tend to be associated with communal lifestyles and simultaneously with the more indigenous, multi- nuclear lifestyle and - choice permitting - opt for uni-nuclear, villa-type accommodation.
The rate of housing starts has been slow and access for the urban poor is more constrained than ever before. At the mean, a room is occupied by some four persons. Three in four households occupy a single room, and a rental search which takes less than six moths is exceptional rather than normal.
In order to appreciate the intricacies of rent-free tenure, an overview of Ghanaian family relationships is necessary.
At the hub of rent-free housing practice is the resilient extended family, known variously for its supportiveness and dependency (Clark, 1984; Ellis, 1978; Fortes, 1971; Oppong, 1981). Technically, the extended family and lineage are not coterminous in Ghanaian ethnography in that the former commonly includes one's in-laws (relations derived solely through marriage) whereas the latter is necessarily consanguineal. Similarly, the lineage differs from the clan: in the case of the matrilineal Akans, Lineage's derive from maternal chains (composed solely of uterine kin) whereas clans are constituted through paternal links. However, the expressions have mostly been used interchangeably in Ghanaian sociology and anthropology and, for many Ghanaian, the differences are of quite minor significance in everyday life. Indeed so fluid is the practical position that colloquially, most English-speaking Ghanaians simply bundle these ethnographic entities into a broad and rather amorphous "family". For the purposes of this discussion the "extended family", "lineage" and "sub-clan" are of similar significance except that what the blood relation gains by right, the in-law only possesses by privilege.
The term family, as used here, is thus all-embracing. Traditionally the lineage resided together in a compact homestead and operated as an economic unit. Bonds of marriage were considered as being more frail than those of nativity. Consequently, man and wife did not inherit each other (Bentsi-Enchill, 1975; Ollennu, 1960, 1962). While lineage homesteads are still common in rural parts of the country, residential separation has been more difficult to avoid in the cities. In many of today's rental compounds, tenant households are fulfilling the role of kinsmen as lineage's experience physical separation under the combined impact of urbanisation and migration.
Trends towards urbanisation and modernization have been accompanied by serious tensions in family relationships as obligations and expectations begin to be reappraised. Indeed 45 per cent of the sample on which this paper is based described the extended family as "parasitic". Fortes (1971) highlights the innateness of conflict between the modern conjugal family and the traditional extended family. This conflict of loyalties is perhaps best captured by Oppong (1981, pp.2, 6, 65) who submits that the contemporary Ghanaian man not only has " responsibilities... as husband and father" to his conjugal family on the one hand, but also "obligations...as son, brother and uncle" to his lineage on the other:
After five centuries of interaction with we cultures, joint investment between couples remains uncommon except within a minority of elite households. In a study of 59 households of westenised public officials, Oppong (1981, pp. 90, 148) found that even with such meddle-class urban couples, less than 10 per cent reported having joint bank accounts. Between blood relations, however, joint holdings - especially in landed properties - are common. Clearly, in spite of the changes that Ghanaian family structures continue to undergo, traditional norms remain surprisingly active even in urban life.
To clarify the admittedly complex pattern of obligations and expectations, one final point must be made. While it is widely accepted that the most noticeable benefits do indeed accrue to the poorer (often rural) members of the extended family, it is important to stress that there are gains on the other side as well. Urban households commonly look forward to receiving gifts of agricultural produce from their rural kinsfolk. The receipt by an urban member of such a gift easily results in a life-time indebtedness not to the grantor only, but also to the entire extended family. In general, the acceptance of the initial gift or favour is the key event that triggers the unending stream of reciprocal kinship obligations.
Family-House Variations
For the purposes of this discussion, family-houses may be grouped into two categories. Classes One and Two. The former comprises all that housing which, upon inheritance, becomes and remains jointly owned property. The precise mode of housing inheritance varies across tribe. In the case of the matrilineal Akans, inheritance mostly corporate under the overall custodianship of the abusuapanyin, the head of the larger lineage: with the other tribes though, inherited housing is more commonly assigned on a room-by-room basis.
In contrast to the usual practice in other parts of the world, inherited interests are not unilaterally tradable (Meek, 1952) and bequests remain jointly owned by the deceased's successors. Because inherited interest are not usually alienated, individual rights become rather minute with successive inheritances. In comparison, Hamnett et al (1991) indicate that liquidation is the statistical norm in the more "regular" UK market. Williams (1991) reports similarly - that three out of four inheritances are sold and 40 per cent of all housing equity outflow in Britain is attributable to inheritance liquidation alone. For Sri Lanka too, it is reported that the common practice would be for the entire house to be assigned to one heir. This is not unlike the concept of primogeniture practiced by Europe's aristocracy. If the deceased owned a substantial parcel of land, the undeveloped land could be divided up between the other heirs, or else one heir would buy out the interest of the others.
Even such forms of inter-family alienation are rare in Ghana. Those who prefer not to reside locally and those who - due to the inadequacy of rooms - find alternative accommodation still have inalienable rights in the property and will often use that house for important social functions such as marriage, outdooring and funeral ceremonies. The combined product of the inevitability of death and non-tradability of inherited housing is that virtually all Ghanaian house convert, at some stage, to this mode. Every Ghanaian, by right, has membership status in one such house at least; this category of family-house will be referred to as Class One in the remainder of this text.
The second form of family-house, Class Two, is the house which is built by an individual who then assigns rights or privileges therein to poorer kinsfolk. 23 per cent of Kumasi's first-generation compounds are, through a variety of generous provisions on the part of the producer, jointly owned. More commonly, though, access is granted as a privilege (Oppong, 1981, p. 8), but while the recipient's position is only a favour in law, an owner would not easily rescind this decision even if he genuinely required more space for his own household. Provided the beneficiary maintains a record of satisfactory conduct, therefore, such access once gained, is reasonably secure over the life-time of the producer. Overall, class Two access has been granted in an overwhelming 71 per cent of first-generation compounds. Households owning accommodation outside Kumasi were asked about the identity of current occupants. Some six in ten of such rooms also are occupied rent-free by non nuclear kin. Quite clearly, Class Two family-house is no fringe practice.
In a minority of cases, prosperous house-owning family members may make outright inter vivos assignments to individual relatives or groups of relatives. These would be immediate or distant relations who the producer wishes to honour. Possession of an assignment so made is traditionally effective when the beneficiary responds with a token thanks-offering known as aseda. Any inter vivos housing transfer effected in this manner is irrevocable if either the grant or acceptance proceedings are performed before witnesses. Eight per cent of the rent-free households in the survey received such outright grants which they in turn can pass on to their own heirs.
A variation of Class Two family-housing arises where relatives are granted permission to append rooms onto an uncompleted compound. This is made possible rent-free occupiers had received such limited site rights.
Simplistically, rent-free consumption of Class One family-housing is by right whereas such access is only a privilege in the case of the second Class. This compressed classification differs from that used by Amole et al (1993) who have more "types" overall. The practice employed here is intended to stress more clearly the distinction between rent-free accommodation which one gains by right and that which is obtained through the owner's generosity. Thus, the lesser "type" distinguished by Amole et al (1993) become derivatives of these two distinct forms.
Costs and BenefitsBenefits
a. In a very real sense, both forms of family-housing constitute an invaluable social security instrument. In faithfully providing accommodation for the most vulnerable social groups family-houses mitigate the social costs of poverty and old-age. This is particularly valuable in a region which not only claims the highest house production costs known world-wide (world Bank, 1984). But also lacks a well-developed social security system (Ekpenyong and Peil, 1985 Peil and Sada, 1984). It is interesting to note that true homelessness as is known in other parts of the world (developing and developed alike) is extremely rare in Ghana. In contrast, even major industrialized nation capitals like London - with some 50,000 homeless - usually have their inhospitable Regent Streets and Embankments where the destitute will be found huddled up in their cardboard boxes on the chilliest winter nights.
b. The form of family-housing which arises through outright gifting is especially equitable in economic terms. Since transfers are invariably from wealthier citizens to the more disadvantaged sections of society, this form would score highly on distributional efficiency.
c. In the variation of Class Two family-housing where a kinsman receives accommodation as a privilege rather than by outright gifting, security of tenure theoretically terminates on the death of the benefactor as successors are under no obligation to extend favours granted by their predecessors. Still, an honest and well-mannered kinsman is less likely to be thrown out on the death of his benefactor. In this respect then, this category of family-housing is valuable in promoting good behaviour as beneficiaries have much to gain by proper conduct.
d. * As has been indicated above, standard plot sizes are large and land cost can be disproportionately high for those who can only afford to erect a cheap room. The option of adding a room to an existing house is therefore particularly attractive to the less prosperous householder who has received a small windfall perhaps from back-pay or other unexpected income from formal employment. A bonus is unlikely to be large enough to warrant considering the option of regular full-scale construction. The cheapest plot would cost several times more than the single earth room he wishes to build and such a house would also be especially wasteful of land. If, however, he can obtain limited site rights on a relative's plot, then he is better able to build cheaply. In this way, he gains free access to whatever services there may already be on site and consolidates his links with the larger family. This also minimizes the extent of land under-utilization and costly urban sprawl. Whilst such qualities are undoubtedly impressive, family-housing has its costs as well. Family-house occupants tend to be among the poorest in the city and housing conditions therein are among the worst. As a result, family-house residence is also interpreted as representing failure. The position really is something of a chicken-and-egg case in the sense that the perceived stigma attached to family-house residence is simultaneously the cause and product of the unwillingness of the more prosperous to remain in residence. It could be said that partial result of the outward movement of upwardly mobile members is the loss of maintenance resources and a consequent deterioration in the physical fabric of such buildings. This then contributes to stigmatizing the house even further, so that other prosperous members are less willing to continue living therein. Another reason why the more prosperous are unlikely to remain in family-housing is because these houses are - in cost-sharing terms - a major liability to those members who are perceived as being "comfortable". Those who are economically more stable often have, whether by design or default, to bear more than their due share of service bills and other common cost. If they can, they will find more personal rental accommodation in other locations where they will, for the most part, be free from the burden of financing the service bills of free-rider households. Whilst they may continue to help in some way with maintaining the family-house - when for example face-lifting is required for a funeral - this kind of maintenance is largely superficial: in the form of painting for instance rather than the provision of any meaningful services.
Legislation
The impact of legislation of family-housing is diverse, if often indirect. Potentially, the nature of current consumption and the future of the entire systems are both affected by Ghanaian law.
For the dominant Akan ethnic group, traditional inheritance practice still excludes a man's own progeny from his inheritance (Fortes, 1971;Ollennu, 1960; Woodsman, 1971). Though children would normally enjoy use of their father's privately-built home over his life-time, continued residence in the house after the father's death is by no means assured. This insecurity arises because Akan men are traditionally succeeded by their nephews rather than by their own progeny (Oppong, 1981). In variations of this inheritance system, ascendants as well as descendants may be entitled to stakes in inheritance fortunes (Fortes, 1971). Overall, Akan are the dominant tribe in Ghana, comprising five out of Ghana's ten administrative regions.
After a prolonged period of public outcry regarding the inequities of matrilineal inheritance (Fortes, 1971: Oppong, 1981, pp.43-4), the government eventually intervened with legislation designed to redress this asserted anomaly. This law - the Intestate Succession Law (Ghana, 1981) was also made necessary by the continued prevalence of intestacy. In the event of intestate death, the spouse and children now have control over the deceased's house would the non-nuclear lineage be entitled to shares in subsequent houses. Local observation indicates, however, that the disenfranchised lineage is not surrendering its customary interest without a fight. Indeed, intra-lineage litigation continues to be common not only with intestate inheritance but even in testate situations.
In spite also of the succession law, 43 per cent of owners interviewed in Kumasi still expect their housese to be inherited by the lineage. A small proportion of these owners consider themselves indebted in some way to the extended family others are simply trapped into this position as it were having built what were intended as personal homes on family-owned land. Case-law precedent suggests that where an individual builds his house on "family land" - a minority but by no means uncommon practice - the usufruct legally reverts to the larger family on the death of the builder. In Ansah vs Sackey for example, it was ruled that "a house which a person builds out of his private means on family land .. is family property" (Ollennu, 1962, p. 40). In a subsequent Court of Appeal ruling - in the case of Biney vs. Biney - this interpretation was confirmed. with the increasing commercialization of ancestral land, however, such private construction on family land may well be on the decline.
Rent control laws (Ghana, 1986a, 1986b) have also had some effect on the volume of rent-free consumption. In their work investigating the impact of rent control legislation on the Ghanaian housing market, Malpezzi et al (1990) and Tipple and Willis (1991) found that controls have encouraged landlords to withdraw rooms from rental occupation for the rent-free use of their kinsfolk. National policy intentions and proposals by development consultants tend to lean in favour of uni-nuclear housing. Logically too, the high standards of construction demanded by regulation encourage the production of smaller, single-household forms rather than larger, multiple-household compounds. Discussions with municipal administrators also reveals a marked anti-compound stance. If there are fewer rooms in a house, then obviously it is less likely that remote kin will find easy access to [Class Two] accommodation. Official housing policy is otherwise silent on family-housing however.
The future
This section presents some arguments in a review of the present and a simultaneous preview of forces that are most likely to clash in determining the fate of family-housing in the earlier half of the twenty-first century.
Routine economic hardship has also been identified as aiding the process of individuation as remote kin are neglected by default. In spite of this generally valid trend, however, kinship networks remain quite vibrant even in the urban arena. In times of extreme adversity especially, such relationships are reactivated and even strengthened to the benefit of all partners in the extended union. During the droughts of 1983-84 for example, fading ties were promptly restored with rural kin who undoubtedly, were still better laced to provide some life-saving staples. It is therefore difficult to conclude that continuing hardship will necessarily speed the decline of the extended family system.
b. There continue to be calls (from albeit minor sections of Ghanaian society) for further action to suppress the authority of the non-nuclear lineage. Reasons vary whereas some claim justifiable disgust at the history of poor treatment meted out to children and widows on the intestate death of their Akan fathers and husbands, others would appear to be driven almost exclusively by the desire for modernisation.
The interstate Succession Law, enacted in response to such calls, was expected to resolve the question conclusively. On the surface, it may seem that observance of this law would result in the hastened death of the family-house. Paradoxically however, this law gives legal support to the contempt associated with the trading of inheritances. It specifically stipulates that spouses and children so inheriting do so as tenants-in-common. This clause would appear to entrench the corporate status of family-house, making physical subdivision and unilateral alienation of individual shares difficult. It may be argued that the law only redefines the "family", assigning to the conjugal family an advantaged status previously enjoyed by the extended family. To the extent that inheritance remains corporate, however, the future of /class One family-housing appears by this law to be guaranteed for some time to come.
While also this law undoubtedly favours the conjugal family, the de facto position is much more complex owing to the plurality of legal systems still operating in Ghana. Customary law has never actually been repealed, only made subservient to statutory law. Even then, one simply needs to observe the ubiquitous weekend funeral ceremony to recognise that the "average" inheritance continues to be determined through customary law. "Church law" - ie. The Marriage ordinance inherited from England - and statutory law carry significantly less weight in practice. Even in the unlikely event that the conjugal family is legally protected by will, the extended family will often litigate. The more "modern" bases of inheritance determination apply mainly in the formal law courts rather than the traditional courts presided over attending western-style litigation in Ghana, all but the most determined litigants are deterred from seeking redress in these formal courtrooms. Most would rather seek justice through the traditional hearings routinely held at domestic level. In all but the small fraction of litigation which reach the statutory law courts, therefore, it is customary law that takes precedence.
c. In an elaborate monograph, Gilbert (1981) illustrates the seeming disproportionate importance accorded to death and funerals in Ghana, and Ghanaians are all too aware of the extensive power wielded by the lineage and sub-clan on the death of a member. The traditionally approved venue for funerals and other such functions is the Class One family-house.
Traditionally too, decisions regarding details such as the date and nature of burial and funeral ceremonies are the prerogative of the deceased's siblings, parents and the parents' blood relations. The involvement of the widow in particular is considered a privilege and it is not uncommon for even the offspring to be ill-treated. For large sections of the population, the ultimate humiliation that could be suffered is posthumous shame: a ridiculed corpse, one whose funeral appears unfitting, one whose corpse has no home in which to be laid in state (see Brokensha, 1966, p. 202).
It has become quite common in modern times for wealthy offspring of a deceased person - in association with widow - to challenge the authority of the larger family and to assume real control over these decisions. As previously indicated, recent legislation does indeed aid such nuclear family solidarity. Where a man is educated and has a privately-owned house, his more westernized nuclear family is more inclined a priori to loosen their ties with the wider lineage and sub-clan. This preferred weakening of ties but the elitist section of society is made possible mainly where the householder owns a home, thus offering some alternative place of congregation and avoiding posthumous shame in the inevitable event of death.
Still, however, an individual who commands little wealth - and indeed almost anyone who is not himself the owner of a private residence - has to assume that de facto control over his burial will remain in the hands of the lineage. The most affluent citizen and the body of practicing Christians may dare to defy the establishment over the manner in which a funeral is conducted, and their interests in family-house as funeral parlous may indeed be significantly less than is the case with most traditionalist. For the majority of Ghanaians who are so concerned not to be "disgraced" on death, however, the Class One family-house - faithfully serving as funeral parlous - is a resource which they will not easily allow to disappear.
d. In similar vein, there are those whose interests derive mainly from the living accommodation available. These are typically the poorest sections of the urban community but their numbers are large enough to constitute an effective pressure group. Gugler and Flanagan (1978) contend that the more disadvantaged members of society are unlikely to relinquish so readily their claims to kinship privileges (in class Two family-houses). As there are no credible formal sector social security provisions to cater for the needy, the poor majority and the elderly cannot but continue to rely on the "generosity" of kin and the authentic insurance which the family-house system provides against adversity.
e. Even though home-owners may themselves have little or no need for the residential accommodation offered by the more crowded Class One family-house, it would be naïve to conclude that this signifies no real interest in such houses. Whilst owners of private houses - as opposed to corporately-held houses - may have little need for the such accommodation., they nevertheless have interest of a more indirect nature. For these, the existing stock of Class One family- houses provides the only immediate hope of a reduced burden form poorer kin seeking free residential accommodation.
f. Linked to the generosity of Ghanaian owners is the sometimes critical role of family assistance in the supply equation. In the absence of a functional mortgage market, family assistance - mostly I n the form of credits - can be vital though the amounts involved are usually relatively small. In the survey under discussion, 17 per cent of housing construction had benefited from one form or other of cash assistance from non-nuclear relations. In sharp contrast, however, the tenure class comprising rent-free kinsmen was present in a dominant 71 per cent of all houses. It may seem surprising that the lineage and sub-clan are able to conjure up free accommodation so easily, but it also does suggest that despite prolonged interaction with other cultures, the age-old power of the "family" has suffered less erosion than might be expected. Whereas the mean number of rent-free residents in each compound was 1.5 overall, the figure was significantly higher (2.1) where owners had received assistance from non-nuclear kin.
Conclusion
Several issues of policy relevance have been raised in this paper; not least the family-house's equity dimension. Family-houses however comprise some of the most neglected housing in Ghanaian cities. This is not only because the occupants are typically poor, but also because - being jointly-owned - family-housing tends to behave as a public good, with each title-holder leaving the responsibility for improvement to others. Until policy-makers recognise the vast resource inherent in this form of housing however they cannot even begin to consider what positive intervention instruments could be designed to improve conditions therein.
It is appropriate, in concluding, to briefly recall a recent episode which encapsulates rather succinctly the value of family-housing. In 1983, when Nigerian oil fortunes had dwindled under the strain of plummeting world prices, over a million Ghanaian migrants were repatriated to Ghanaian migrants were repatriated to Ghana. Within weeks, this formidable contingent constituting over seven per cent of the national population had been housed. One may ask how this feat was accomplished. Was it that the usually stifled housing sector suddenly received an injection of major funding from the already straitened national exchequer: or was it perhaps that the response of the national donor community was particularly swift and generous? No, the answer was not provided through either of these: these vast numbers were absorbed mainly into family-housing, saving a situation of major homelessness.
There is indeed some indication of an increasing preference for individuation. It is not altogether surprising that relatively recent discussions tend to allude more to the dependency of the extended family whereas older texts are more descriptive of its supportiveness. There can be little doubt, however, that it takes more than modest "modernisation: to break free of the family-house system. Clearly too, the pattern of relationships between identifiable units of the extended family is not wholly unidirectional.
Notwithstanding the changes that modernisation and the urban economy continue to introduce, I posit that family-housing will remain active for another generation yet.
The entrenchment of nuclear family practices in "middle-class" homes may well result in reduced access for the more remote and poorer relation. However, as long as housing remains a non-marketed good - culturally and now implicitly endorsed by law also - inheritance will continue to result, in the medium term at least, in joint ownership and multi-nuclear occupancy of one form or other.
Eventually though, acculturation and the development of sustainable mortgage lending may succeed in introducing changes in attitudes towards house selling, leading to the ultimate demise of family-housing. If and when this does occur and family-housing is overrun by modernisation, policy-makers should not be faced with previously unknown homelessness.. except of course if we researchers have failed to sound the early note of warning.
The foregoing is not to suggest that the family-house is the panacea for all problems of housing in urban Ghana. Neither is it to deny the presence of major stress. The point being made is simply to indicate that the problems could be decidedly more acute without the contribution which family-house continues to make. Given the harsh economic climate this form of housing is too important to be ignored by researchers and policy-makers.
ReferencesAMOLE, B, KORBOE, D AND TIPPLE, G (1993) The Family-House in West Africa: A Forgotten Resource for Policy Makers? Third World Planning Review 15/4, pp.355-72
BENTSI-ENCHILL, K (1975) Some Implications of Our Laws of Marriage and Succession in: C. Oppong, (Ed) Changing Family Studies, Legon Family Research Papers 3
BROKENSHA, D (1966) Social Change at Larteh, Ghana. Oxford: Clarendon.
CLARK, G (1984) The position of Asante Women Traders in Kumasi Central Market. Unpublished Ph D Thesis. Cambridge: Cambridge University.
EKPENYONG, S and PEIL, M (1985) Old Age in Town: Implications for Housing and Research African Urban, Studies 21, pp. 15-24. ELLIS, J (Ed) (1975) West African Families in Britain. London: Routledge FORTES, M (1971) The Family: Bane or Blessing? Accra: Ghana Universities Press. GHANA (1985) Intestate Succession Law (PNDCL III)
GHANA (1986a) Rent Control Law (PNDCL 138)
GHANA (1986b) Rent Control (1st Schedule) (Amendment) Instrument (LI 1318) a Ghanaian Kingdom, African 58, pp. 291-313.
GOODY, J (Ed.) (1975) Changing Social Structure in Ghana: Essays in the comparative Sociology of a New State and an Old Tradition. London: Oxford University Press, for International African Institute
HAMNETT, C, HARMER, M AND WILLIAMS, P (1990) Safe as Houses: Housing Inheritance in Britain. London: Paul Chapman.
KORBOE, D (1992) Multihabitation: An Analysis of Low-Income Residence in a West African City, Open House International 17/1 (Special Issue on Sub-Saharan Africa).
MALPEZZI, S, TIPPLE G and WILLIS, K (1990) Costs and Benefits of Rent Control: A Case Study of Kumasi, Ghana. Washington DC: World Bank